Sports Betting Legend Turned Sports Market Maker Rufus Peabody Joins | Ep 137

I've I was burned out as hell and like unmotivated for like five years and just kind of coasting by on a nurse and you know There's still things to get me excited occasionally, but like this is I don't know this has been reinvigorating my uncle told me At Thanksgiving dinner that he hadn't seen me this motivated in a decade What's up everybody we got GPSP and RP Three the three P's now this is like a crossover episode. We got Rufus P body joining us. You may know him from but the process Twitter The prediction, you know the call sheet leaderboards, you know pick pick your poison, but yeah, thanks. Thanks for coming on the show. This is a This is an awesome. It's awesome to have you on Well, I love you guys his podcast and I'm like very honored to be invited on the show. I haven't been invited onto a podcast in a long time Well, I was telling you I think people might just be it just think you're gonna say no I don't know or just be nice No, I don't I don't think so, but um, yeah, I mean this is a crossover, you know crossover episode with with bet the process too But this was one where I was like looking at our outline and making it And I was like I was like damn there's a lot like you've you've been in this business. Maybe like someone like aunt Um, we had on similar amount of time, but the career has like been very very long So this could be a five-hour episode. We'll only keep it to two a short episode by our standards I mean, I've never been on a on a two-hour episode as it is and I mean, I'm not sure what I'm gonna do with like time to speak without somebody else interrupting me Well, yeah, that's that's kind of the value prop of this show because we just After like an hour and a half everyone gets tired and they start leaking alpha. So I see that's the main the main thing Well, Jeff always says that like basically he saves me for myself Like if I you know, he cuts me off when he knows I'm not going anywhere with something or just gonna confuse people so I'm sure people can tell tell us afterwards whether that's actually accurate or not What's your take us bit Jeff is too quick to cut him off or not? You know, there's there's probably a balance range He's probably kind of you off sometimes when there there is some good stuff in there, but um Yeah, no, I I mean, I've been listening to your your show your true sports betting podcast Pioneer I was I was explaining to my actually my wife didn't actually even believe your name going back to I was an actual name. I'm like, you know telling my wife who are in interviewing and she's like is that his real name? I'm like yes So yeah, we're really we're really excited to have you on I think you know when we were putting together the questions I think we're trying as best we'll try as best we can hit like new Sort of discussion and stuff that hopefully you haven't talked about As much at least um just because you can you can offer a lot of perspective so maybe let's just jump right in I think You know, it's well documented and there's actually a great It's probably great several podcaster episodes or whatever. I remember the most recent one is the the risk of ruin with you and And some of your partners early on and like how you got started into it to sort of talk about your upbringing So maybe to just like jump into to some of the more recent stuff, which maybe not us talked about quite as much like let's just start with um Like you jumping from LVSE um And and some of your early models because I know you you sort of talked about that that being your start And I think you're doing some baseball, but I'm curious like what the The landscape looked at at that time of what what types of stuff you're doing slash betting like early on as you moved on to your home Or in the group. Oh man. Okay, so the landscape was quite different. You didn't have mobile sports books You you had offshore still, but you're I remember driving at like after work at LVSE driving from casino to casino Placing like prop bets and like derivatives like I built derivative moneyline charts and first halves and first quarters for NFL and college football This is my first Here so that would have been like the fall of 2008 oh and teasers preseason round robinning teasers that was that was where the alpha was But You couldn't bring your compute you couldn't have a computer in the sports book. You couldn't use your phone in the sports book So I had to look I looked at the LVSE odd screen to see these prices Across different books and then I had to drive to the casino and hope the price hadn't moved um and at the time like This is pre having an iPhone as well. So I had like some little flip phone or something Um, I couldn't really hotspot. I don't think I could hotspot I might have one of those points. I bought like some sort of mobile broadband thing so on road trips I could do that um or frozen my car, but like it was it's funny how It's How it is so different now and it kind of happened like it feels like it happened gradually But in terms of like my really models and what I was doing Like I was betting I think the first thing I bet was majorly baseball So I did this internship at LVSE for a summer and I was betting correlated parlays um, so sought run line total side total like Basically, I figured out that if you just parlayed every home underdog with the under that had been like a nice return for the price like For the past 10 years like at the pinnacle line. So you know, I would price shop and and especially with higher totals higher totals to the to the under. I mean home underdog obviously uh home team wins. You're generally only getting eight and a half innings And so there was you know and books took that I don't know. I don't know. I still take those um I doubt it. Not anymore. They certainly didn't take like like The road team minus one and a half with the over on low totals, you know same type of thing But so, you know, I think by the the things I did at the beginning were very much just kind of I guess you'd call it top down or Using a source of truth and pricing out derivatives off of that and then props, you know, I read Stanford Wong's book I used Passon distribution for player props I Like there were a bunch of game props that weren't priced well. For example, you know, things like will there be three answered scores Will there be a score in the last two minutes of the first half like all these types like a bunch of stuff like that um that You could you could find alpha literally just by running logistic regressions And having a data set built to be able to answer those questions and I mean now it was tough to acquire data back then And I think I had a nice advantage because I had data from LVSE like they had data going a long way back I don't know how accurate Uh, it it it is or was but Like I had football points spreads from 1980 which honestly wow or not useful at all now but So so I was just kind of like doing this top-down stuff like driving around Vegas um And then and then When I started with the group um that I worked with for a long time That's like Zach white um Arcturosa who's ed teach on Twitter and this other guy train We we we started doing live majorly baseball and so I was sitting in the LVSE office This is gonna sound like janky's fuck, but I was literally running regressions between innings on like Would it would does team a win by Or have a score differential for the remainder of the game of greater than a half if innings is five and it's the bot like and top bottom is bottom and like In the regression has like you know the the closing lie in total like Is it Do DH rules apply something with like batting order whether starting picture was in and like current score And actually with baseball you have a big and robust enough data set because there's so many games that you could actually Say like run the regression if game is tied in some cases and literally I'm just like It's it's almost like an intelligent query, but I'm spitting out stuff. I'm messaging Zach who's at the M Resort which was Run by cantor gaming and he cantor gaming who had just bought LVSE by the way He had a I was Triple A. That's what I was referred to as For a while there because they didn't want to out if somebody saw his computer who he's messaging like you know I had my like my coded name you think Rufus would work for a good name disguise, but But I did work LVSE so you know he so he was sitting at one of these terminals at the M Resort just betting the betting these between innings and and we were just kind of cleaning up Against their quote Midest model they had built and so So they were pricing it live it wasn't like So they had like their own guess as to what the live number was supposed to be it wasn't like supers. It wasn't like a stale No, no, no, this was literally like me my model of Running regressions on the fly against whatever model they built It's so it's so funny like you hear you describe this because Um at least my a lot of like you know the way the industry is now in the prediction market and all these things It's like you have like milliseconds on some of this stuff It feels like and you were literally back then like getting in your car driving Hoping the odds didn't change and getting the bet and it's just such a such a difference to to what it is today Well, that that's what I was doing before I joined this group But what we had with the group was we basically had we had four full-time runners that would Be driving around Vegas from casino to casino like one person would go to the win every morning at 8 a.m Right or like and we're we're literally we had somebody piloting them every day And so it's it's all oh this line popped up can you can you drive over to Caesars or a Caesars property or the back then it was known as Harris rather than Caesars, but In that situation like we're your runners just getting backed off a good bid or like well Yeah, so stations was the one book that would limit people to like one trip to the counter per day That was kind of their move and eventually they did cut some people but It was a very different game like I feel like there was more honor in sports betting back then Like it there was no such thing as being limited or You know being kicked out of a of a place unless you actually broke Broke the law or broke their rules You know you take a bet you move an umper and that was kind of the way I learned about that that's what I learned from The old guys that I worked with at LVSC and that's just kind of how the industry was This is a topic, but I'm curious if you think like What hypotheses do you think like why that changed? Do you think like that is still a viable business model under like The current situation tax situation like legal all this stuff or it was that A remnant of the maybe lack of of sophisticated betters lack of tools lack of people act actually trying to win Curious your take on that. I mean there were sophisticated bettors like the Billy Walters is of the world I think it was just a different approach to it. It was less corporate And I think less greedy And I also think that sportsbooks at the time were considered like they weren't profit drivers for a casino They were kind of like the lost leaders. They weren't hoping to make a lot of money at least Most of them weren't they were just trying to get people inside the casino Yeah, that makes sense. It's interesting at work. I mean like I know you'll prime has had trouble I mean, I think the big thing is just getting enough Like it's it's all about marketing and If you had a book that did that sort of had that sharp model and didn't limit bettors And had the same marketing budgets that draft kings or fan dual like um I don't know. I mean if you're not offering all the SGPs and all that other stuff that's Like I think it is hard to be competitive just Because of that Yeah, I would agree like it's just the menu right the SGPs are the most important thing But I think just as far as like rec volume goes and what actually like rec's like to bet so Yeah, I mean it was probably just like a part of the menu too back then like I'm sure like you said There's a lot of props where there's no SGPs or I guess it's yeah Or these teasers or whatever so there's like a couple spots where like you can get that structural edge But now it's like you can find these These glitches or whatever like in so many different spots right and I do think the doubt like sports books really have to protect They have a lot like it's like more of an existential threat like any mistake in their Code and you know, they're basically facing a PR situation if like they have all these sharp bettors I mean they build you know as as Louis who is way more eloquent and polished than I am said You know, they're they're building this whole infrastructure based around a source of truth It's like a mainline and that can be manipulated and it's It's a very fragile situation but I did want to say that I actually enjoyed this I thought I liked the sports betting industry so much more 15 years ago than I do now it was It was just I don't know it was more of an adventure. It was different. It was like Not it was more niche. I don't know. There was just something about it It you know being in Las Vegas during that time that was just like Special romantic school. Yeah I think that what about poke sense but like the Mark Twain or something poetic Yeah, I think that about poker. I think about like when I started watching poker before I started playing or right when I started playing you had like Phil Ivy and Doyle like playing and you know and everyone was in Vegas and they played these high six games and you know They're all characters. They all had some gamble like and now we're all kind of just like dorks You know running businesses or whatever But um well, it feels like any other industry now. That's the problem All right, like I thought it was it was this sort of completely different foreign world and I love that Well, it is interesting and I think like that's where this podcast is going to be The most interesting is like you won back then and you're still like winning at a super high level now Which I don't think a lot of your peers from 15 years ago Are and especially when we're talking about like this year prediction markets and stuff um So I think let's go to when you broke off onto your own From you said zack ed teach and train so when you decide to go Going your own and you know In whatever. I don't know what you're better than was it golf or or whatnot, but like what was that Step to take that jump on your own like what was that decision process like and how did you kind of spin up your your business there? I mean it wasn't a huge huge change So I worked with that first group. Let's see from 2009. I think we Think it was 2016 that I sort of split off um So for me at that point like I was Doing about I was doing like 95% plus of our modeling and originating and At that point I had been able I had started sourcing some accounts for us as well and and so I still was only getting 25% of the profit and the I guess it was back in november november of 2015 something like that. I might have my dates a little bit off, but I get a call. I'm on an airplane. I get a call from my old boss at LVSC Kenny white who says Mickey Richardson wants to talk to you. I was like I said who the fuck is Mickey Richardson? I didn't have no idea. I don't know who that is He's the guy that runs backress I was like how does he know anything about me? He's like well He figured out that you're this for to account that's been like beating them very badly and he wants to meet with you And so I talked and like talked to them. They I flew down to Costa Rica um That's where I met Matt David how actually? Wow. He was well, you know, I saw the chris offices like um I and they they tried to hire me and and the amount of money they would have given me was substantially more than I was making betting and with with Crazy amount of upside Even though I'm not sure what my job description was but You know, I I said no because They were There was a lot of legal risk with taking that in I enjoyed what I was doing at the time but I think it made me realize that I would I felt undervalued I didn't feel like I should be splitting the profit equally given the fact that I was generating like My skills. I think had more market value That makes sense But it was a really tough conversation to have with the guys because Like we were all like a family And we all really liked each other. We were you know Everybody was in each other's weddings Subbed me because I'm not married but And and when we had this cut when it when I brought it up it was like it was a really emotionally charged Conversation and there were a lot of hurt feelings and I remember Zach saying you think you're better than us And I tried to say something like no but like it's not about who's better, but I just think that I I deserve more. I think it would be fair It's about what's fair and what feels fair to me and so that's why we ended up kind of like I didn't want to leave the group You know, I liked the fact that There were people that I didn't have to do everything that I didn't have to do all the trading um The collecting the sourcing of accounts, but I felt like I I felt like I wasn't getting um I wasn't getting the same upside. I wanted and and so I ended up. Yeah splitting off What one of the guys like train? I kept work he He kind of afterwards was like hey, you know if you If you want to work like I would still be open to working with you um under different arrangement and And then we but it was basically the same thing that we brought on You know a trader who This guy who lived down and he lived he was a friend of trains from college who was an econ guy um Also econometrics You know built some models was working on building models, but he lived down in Costa Rica and interestingly he Kind of got well, I don't want to say got he put us in a in a tricky position at one point when he became romantically involved with someone who was also romantically involved with one of the top guys at chris and it was like oh no We have seven figure balance with with them like There are an unregulated offshore book like we do not want to anger the people the control our money And so that part of sports betting is certainly dead like you wouldn't have that at the regulated fandom like that type of situation These people who are on sports betting twitter. It's not really in play for ton of uh ton of the better his now Yeah, but it was you know, it was a smaller, but we still We're doing the same thing what sports like yes golf NFL college football Majority baseball that that's the big one that I don't do anymore that that I was doing and I was certainly betting more NFL um like big park and stuff than I am now So I want it I want to talk about get into some of the sports and one of the things that we that we have I want to do a decent amount of discussion on is the modeling because one of the things To me that separates you and it was something gp just said is like To me there's I don't know many people are don't see many people Who've been in the industry this long who've have like a modeling or originating bent I see people who have like um, you know like edge seeking or whatever like more of an ap style bent Who've been in you know, gambling a long time that makes sense to me the The ability to stay like on top of this and at the sort of like cutting edge in a very high level and the the modeling the The originating is is much braver. So like walk me through I know you know like when I was first listening to your podcast years ago Like I think it was primarily golf at that point was was what you were uh betting a lot of um walk me through like how you got into a sport like that um and what What that looked like in the beginning? It was kind of an accident At the time this is back when we were I was trying to at the time I was actually trying to build a real Live baseball model after like I was like I think I can do better than just running these regressions on the fly um They ended up being like three months and the three month d2 or that didn't really pay off But but train had gotten his hands on some uh a golf data set and he he'd he was like He was betting on two guys like they were equal in skill based off of their scoring average adjusted for course difficulty field strength right but But one guy is a long hitter and another guy's a really short hitter and they're playing in quail holla Which is a really like a bomber's course and the market wasn't adjusting for that And you could like it wasn't pricing that at all it was pricing You know course history if one guy had like one at this particular course He was probably gonna be overvalued, but like there was some There was a lot of alpha there, but like on those matchups and I said hey like can I see that data and I Built out. I basically first thing I did was found how to properly like time decay the different skills um and then Figured out what was more important what was less important Um because obviously there's there's a lot of randomness in just scoring and and if you had two guys like That had the same scoring average and one guy that was just a really good putter and the other guy was a shitty putter like You expect the shitty putter is gonna be better We have a lot more data now that we we can actually assess a lot more nuance than then I did then but You know, I'd then I add it layered on some some course So of course fitted like course fed adjustments course experience and history and and ran simulations at the round level and produce numbers and we just Started betting it immediately and like it You know, I had never backtested it. It wasn't like baseball where I'd done extensive backtesting But we were I think based on what trained was betting at the beginning like it was clear that there were edges there And we and we had success from the jump um Though No, you you have down periods of course and I still remember there's back then we had uh we had By betting partners were called agents Um spanky hadn't rebranded that term yet Spanky is the branding master vibe. I know everything like top down betting and betting partners happen like in now that is just common lingo Right, but yeah, we still haven't made anything happen yet. That's one day I I certainly haven't But I'm still trying to make fetch happen I'm just trying to make Bayesian update happen But we had I still remember we had one agent that One you're legitimately told us Like he's like I don't want your golf in October november and september fine, but not October because we'd add to Yeah This is the kind of these are the kinds of people that are moving moving for you But at that point with golf what we would do we would have to It was it was a big operation to kind of get down on these matchups. It would be like Zack this point like mad dog um and Mark would be all in They they had different agents on the phone and there was like a countdown to for when they they were gonna hit all these books at exactly the same time and so we were able to get like you know 30 grand on on a matchup and just completely new to screen but but yeah That's pretty good. Yeah, I was well. I was gonna ask how much you were getting down At on golf at that point. Well, I was only it was only a quarter of it remember Right, right, right. So you were getting. Yeah, you were getting down like 7.5 K but the Yeah, I've I've always wondered like it kind of feels like the right way to start something where We we had this discussion about backtesting and whatnot and like Sometimes it's just like you're like wow, there is an edge like you just see like yeah, I don't know. I'm betting this guy Who they're pricing the same but it's bubble lots inverse Brian Harman at quail hallow and they think they're the same golfer You know whatever. I'm just gonna bet it and then you're like well like Okay, that's winning now how can I kind of build on top of that instead of being like I have no clue what wins. I'm just gonna like try and model this sport like right from the rip. I don't know if that's if you've approached Sports in both of those ways, but like have you gone into I think that's like a common misconception or it's just like Yeah, I'm just gonna like model basketball and now I have a basketball sim and you know It just spits out numbers and I bet it where instead like sometimes it's like well You know, I I have a rotation space that true whatever so when you're exploring new sports like how are you going Into it? Are you always going in with like an angle or something that wins or are you sometimes just curious and like doing a big project? I haven't explored a new sport in years nor do I have the desire to but I would say I'm more the person that Models the sport without thinking about where the edge is necessarily going to be and I think by edge It you know, it's not an angle. It's I think about doing a lot of little things well and being kind of creative and how I look at things and being sort of logical and each step of the process like contextualizing well, but My first model was major league baseball and I did my senior thesis on psychological inefficiencies in baseball betting market And in that process like it was actually that model that wasn't the actual model. That was a model that Was or at least the first step of that was Building a model to identify market inefficiencies and so I saw that strikeout percentage for pictures was undervalued walk percentage is undervalued like batting averages and balls in play was way overvalued, right? And so I had an idea that like okay, these saber metric concepts were not being properly priced into the market And so like it captured that but over time there were times when You know towards the end of my Baseball betting where we were almost shorting the like Like it was almost like the saber metric stuff was overvalued at a certain point so um I You know Jeff always says that you should be able to know where your alpha comes from But I think it's really hard in something like baseball where there are so many different moving parts and The alpha on one game might be coming from something completely different than another And when I did my like extended ESPN it was it was really hard to try to Give a narrative around a pick like why why do I have an edge here like honestly in most games? I have no idea. I couldn't tell you why I have an edge in most golf matchups either It's I feel like you could though like you could I used to be able to a long time ago when markets were much less efficient and and those inefficiencies were a lot more obvious Like it was a lot more obvious what the source that edge was I think maybe with some of the new things I've done with golf like there there are areas where I can say Oh, this is the reason I like this golfer I'm doing this. I don't think anybody else in the markets looking at this this this particular way But I think Golf is a I don't want to say it's simpler, but it is simpler. It is you're modeling a player. Yeah One guy you're not having to model all these different team interactions and Yeah, or opponent interactions Well, it's just doing course interactions. Right. That's the opponent You had mentioned getting into any any new sports in a long time like am I am I wrong in saying you've really ramped up Like your college basketball presence No, no, that's not wrong in college, but I like to be fair. Like that's mostly telemakis and that I that was what starting in 2022 I guess and We're not for meeting him and partnering with him. I don't think I would be betting college basketball You know It was just it was one of those it was just really good luck that that we ended up meeting and and He's a brilliant developer and had built this like incredible Incredibly complex nuance like for situate like a simulation framework for college basketball and I think it was It was a very Very well our skill sets are very complimentary Um in the sense that I was able to sort of help build out projections that fed into that sim and and Actually since since then he's kind of run with it and made this projections way better But so he that that that's mostly his department. I mean that is his department. I'm not gonna take any credit there Well, I am curious like on that front because Like you don't just have innate you don't you don't come out of one with innate like college basketball Domain expertise and maybe you needed that maybe you didn't but I would think like to To dive into like like walk me through how you did like he obviously came to you because he thought you could help You know, and it sounds like you did so walk me through like what you think you're your strong suits there Where maybe in a little bit like more detail or what you could you were able to help with? Okay, so first off. I mean he came to me. He had a he wrote an article for unabated in the summer of 2022 And instead of payment he asked for an hour of my time Was this the baseball angle? Yeah, well, he's such a good article you wrote about baseball angle and this is such a good article He had a question for me of a data science thing around Something he was looking at it with baseball SGPs and We got on a call for it was over an hour and then I ended up actually spending like five hours writing some code for him and then Literally the next day he had already put this into production and where and was like we would have had this many bets It would it like it was incredible the what what he was doing from a development perspective like a dev perspective and so I actually I went down he was living in Virginia at the time and I went down and and met him and I was interested in Bringing on a develop someone for to help with our develop like the dev side for golf I had no idea he was even he'd built anything for college basketball And so that it came up in in that time I spent with him and And it ended up being I mean it's it's our number one sport now But yeah, so you're not even a golfer. No, well the thing is golf you can't scale golf in the same way And like this year we had we had a really good college football year and we were able to really scale that That up with like predation markets, but there isn't really the same opportunities with with golf at least there hasn't been but in terms of what I did for the the college basketball sim I I think it was It was essentially the same perspective the same process I kind of go through with other sports and and I think my strength is always been in sort of Not necessarily like running the great sims or anything like that, but but sort of building good projections Like understanding regression at the me and time decay contextualizing performance all that stuff and so I wrote some code that Taylor has since like expanded upon to Essentially build good projections for these different things that fed into the sim So but I I have not done anything on college basketball modeling aside from like A few just I mean Looking up a few things or whatever I haven't done anything on that in Over two years So I think one thing to it's interesting like You have these two different situations which were both like you're both working with and both of them you're working with Experienced like talented better But in one situation it was a very like complementary mix and the other situation, you know I think this is something we talked about sp on our like group you know betting group episode where it's like You don't want to over you don't want to have like too many mouths to feed right away because when you You know dilute the pie you have to make a lot of money for like everyone to be happy and people might not need exactly the same amount And whatnot um and I think it's just like goes to show You know how You should think about partnering with other other people in the space because we get that all the time if you ask us this all the time You have to remember at the beginning, you know I wouldn't have been a professional better without this without these guys that took a chance on me like I met I met market at casino. He almost maced me in the parking lot because he thought I was I thought he thought I was trying to mug him or something because you know, I'm a very intimidating looking guy Yes, I just seen him ask the ask the ticket writer or the supervisor at the palms casino about NBA props and I was kind of a little too shy to ask myself and I'd seen this guy before and I was like I wonder what his deal is so I followed him out in the parking lot But without meeting him and like getting introduced to to that group Like, you know, I don't know where I'd be. I don't know if I'd still be doing this because at the time I had a bank role of In the tens of thousands of dollars You know as I started after college, you know, I was in a job that paid $25,000 a year I um And I didn't really I had a bank role to begin with of like $10,000 which I'd save from like jobs summer jobs and during college And so it's hard to to To scale like that and and you know, they gave me a 20% free role from from the get go and like You know the first month I had made Like my share was $25,000 and it was just for baseball and then they they convinced me to quit and the question is like how much do I Do I owe them and how long? Because if I done that if if I let's say I had one You know, I did that we made a bunch of money for a year and then I'm like, okay, now I have the money. I'm gonna I'm I'm gonna I'm gonna go Right like I feel like I'm gonna take all the connections I've made through you and I'm gonna just Go my own way like I think that would be Not in sort of the spirit of the agreement they hired me in it, but at the same time I'm not an indentured servant and I think like I think that's what was so tough when I left the group because I felt like I owed those guys so much for the chance They gave me but at the same time I felt like it it had been long enough and I didn't think the deal for me right at that point was was fair I didn't think I thought I deserved a greater percentage and and so I do wonder like how Like how do you think about something like that? Like how much or if you if because if you bring somebody on right you're taking a risk Like a they could not be very good or be they could be really good, but then decide to like leave Outgrow you Go on then go their separate way Yeah, I mean in that situation I think you did the right thing by saying like Hey, I would want a bigger cut like you can ask I mean we had this situation with one of my partners Gabriel like it was me and my partner Luke to start we brought Gabriel on and he'd just like absolutely crush it and he had a smaller percent than us And eventually he was like hey Basically like could I get a bigger percent and we were like Of course, you know, but in that situation it because like he added way way way More value than his cut at the time, but it's a tricky thing with groups And it's a tricky thing with with you know setting up the percentages and stuff It's not like trivial. I think so many people get into this and just say okay We're all gonna split it like this way and like don't really kind of look Either don't leave room to adjust because like adjusting is a tough conversation like you said Like they probably thought you deserve more than that at the time, but You know there's emotions involved in ego involved too and I don't know like it becomes very tricky when you ask to like change a percent So that's why I always say to people like be very Very careful with how you set up and like what everyone's percentages are and you know make sure that When this grows if it grows how you want to grow everything still makes sense in two and three years I think one thing we do well now is we We revisit our percent our percentage splits every single year at the same time and it's And that's actually something that I learned a little well. I don't want to say learned but but Matt David out I remember him tell I was kind of astonished that like they would one year he'd be like 25% the next year He'd be 75% of ice first. They're just based on the work They were putting in that at that time and they were like totally fine with that like we don't Move things that much, but but it is I think the goal like fairness is isn't has always really been important to me and I don't want I don't want someone to be resentful I don't want to be resentful and I don't want someone else to be resentful and feel like they're not they're getting their fair share and I think we have you know We're we're doing yeah, we want it to be fair. I want to keep everybody happy and So far we haven't had a situation where We're far off on what we think is fair um, I don't know what would happen if that occurred But I think the key is I think the most important thing is working with Good people that you actually like and respect and trust 100% those are the things You know someone can be brilliant, but if you can't trust them If you don't like working them, it's It's not gonna last Yeah, no, I I feel like it's a nib at me where it's like on one side of the thing It's like work with the work with people you like and then like in the middle It's like you have to work with the person who has the exact right skills and the other end It's like work with people you like or like work with people you trust I think of things only in midway means basically that's my view of processing We got to hit you some more modeling questions Though these people I feel like people see you I ever I use you as an example of one type of better like the Originator. I don't know if you've heard those episodes, but I'll say like you could be like What's like a originator or edge, you know more of an angles guy I want to ask about mr. peanut that it was question because it kind of ties into that I think it was really good question. He asked um For college football and college basketball primarily Oh, okay for those two sports primarily how much of your edge you feel is gained from evaluating teams versus evaluating spots slash situations So like by weeks, home field of vantage, etc And does this change by sport? I think that's a great question because Where's it like we're talking about like what is your edge like Kind of is this question in one way or another No, I agree. I think it is a great question and for me my edge I think has always been more from evaluating teams so I would say That's at least 95% of it and so You know the by weeks the home field advantage that clearly matters But that's almost window dressing on top and I don't know how much alpha. I'm actually getting from that like I think it really is about understanding like the fundamentals of a team and When there is some sort of regime change or like where there is signal as you know a football team a basketball team is very dynamic in nature and You know, it's they're not the same team now like The Eagles aren't the same team now that they were were in week two for example and it's about being able to sort of I don't want to say like You want to have less lag to changes in a team quality than the market. You want to try to get there first And if you can pick up on some signals on like faster in the market I think there's there's a lot of value there and so that's that's I mean I think that's the way I approach it mostly because that's what I I learned from Kenny white who built power ratings and that was his like use my boss LVSC the guy that I call a call to brought me on for an internship and and he had like no bus full of this and this was his whole approach power rate everything build team ratings And so I was like okay the way to bet is to build team ratings and like and obviously You know play ratings that comprise the team ratings too, but I think that spots I I'm probably leaking some alpha by not doing more with that But I would call myself maybe a skeptical empiricist and the fact that I feel like there's all these narratives that you hear and I am skeptical of a narrative unless I can sort of quantify it or I have a really good reason to believe that There is signal to it And so I probably don't dive enough down that sort of rabbit hole, but I think it's also hard From a quantitative perspective at least when you don't it's hard to have the data to answer some of the questions Like earlier this year UNC I don't even remember who they were playing, but they People thought that that Bella checks existence were all about to quit and then maybe Bella check was about to quit And the market way undervalued them and I was like locker room chatter and I was like I don't know how to value that information But I'm gonna assume it seems clearly that it's being priced into the line a lot and I I think it's You know I didn't think I had any way of valet like verifying that information and I tend to think those narratives like Or overvalued just based on my experience but You know there's things like letdown spots right and I've tried to quantify this stuff I've gone through data in the past and been like okay if you came off of a win against like or a close loss Well, I guess it let down spots after well you you beat a top team and then you play a crappy team But there's so much more nuance to it right was it your rival was it like what's the state of that particular program and so just Doing it it's sort of like Grammatically where I identify like criteria to sort of test for like letdown spots or I don't know like a hangover off a close loss Things like that like I haven't found much, but I Think that that stuff very like very well could be there and there's stuff I'm missing, but That's I don't know that's that's not my source of alpha really So I did want to ask a question about that, but before I just had a follow-up question on Mr. peanut better better's question Like when you were talking about your your primary strength is like coming up with team rankings And like team quality maybe it's a better way to say it or player quality or whatever like what do you think you do? What do you think drives that skill so like when I when I think of this problem? And maybe this is a narrow view to think it through I think of you can have you know better data than Then like your counter parties here, but ever like in a distinct data advantage You can contextualize or like feature engineer whatever term you want to use that word by the way I hate that word feature engineer I hate that word feature. I hate the word feature engineering Yeah, I agree Just programs. I'll teach feature engineering. Oh, God feature. What was it? What was it called back when you were doing econometrics? It wasn't feature engineering? No Now feature engineering is like a machine learning word by the way analytics wasn't really a word then data science certainly wasn't a word Okay, I'm dating myself here, but Getting your variables from the dirt. Yeah. Yeah, you put like they're not features. They're they're fucking variables I agree. All right. That's what I call you. I'll use the layer layman's turn made layman's turn. I'm uh, yeah, okay, so like Coming up with the variables to use or what's like actually predictive contextualizing it I would put in like a secondary bucket sort of altogether and then the third Would be like actual like modeled choice and I don't know if you would view this even as a third But like there's different especially today, you know, there's just an ever increasing like number of ways to actually put those All those ingredients in in in the sauce and come up with something. So like what It's probably a blend of of of all three to some extent, but like what how do you think through that? Wait, can you read can you summarize what those three work is our contextualization is to what was one Like the data quality like unique data or proprietary data or stuff that you don't think people are looking at It could be public, but it's just like you don't think people are really considering. Okay, so data quality Contextualization and third was like the actual model structure or or like the setting up or You know like the framework of the the problem like the questions you're asking of the data almost right because you can contextualize like I don't know like It doesn't really actually matter that much and you can do a great job at it But it doesn't actually impact the model because it's something fairly useless like Like you could spend a lot of time working on fumble recovery predict like predicted fumble Recovery rates for teams and you find that it's very different off of a sack than off of a completed pass Or then from an aborted step, but like Overall, they're so little signal there that that time ends up really being mostly waste. I think that's what I think a process filtering like Time man. That's more like time management. Right, but I would say you could be doing a great job contextualizing there and have great data But if but but you're going down it's not something that actually And so how I was thinking through the third one would be like You could come up with all these great variables or whatever to include in your model There are decisions you can make that will make that model better Maybe it look like to give examples like maybe it requires like a non-linear method Maybe it requires to break it into smaller parts to predict this that predicts that or something like that just sort of like the actual nuts and bolts of Of the the model not like so you to me. It's like I mean this is not yeah Yeah, maybe it's not the way you think about it But like to me, I think about as like the starting point like what what am I picking off the tree? What am I doing cutting up the ingredients and then what am I throwing it into and those would be like that? I like that three steps now I love I haven't picked things off trees lot I love that. It's a lost start lost start Yeah, it's fees and the Apple picking back yard apple picking season Make some cider with from it, right? Correct We've really gone off the route. So do you have any What is in the number four is is a tension span? So the data quality I think you obviously need to have good data like you don't want to be a data in a data Disadvantage, but I don't think that's my main source of alpha. I feel like If I'm if I have the same Quality data as everybody else. I think My I think I'm happy because I think my advantage comes from um, I would say numbers two and three. I think Context Contextualizing the data is like super important for me. My process is trying to figure out like What is and isn't random and how random it is? Right, like I want to try to feature engineer The most predictive features possible like things that are That are the least random that are the most sort of directly controllable and And have a set of variables that are like the most like sort of orthogonal to the to measure different things and then But I think a lot of and then Putting it all together is You know, I think I'm pretty good at that too, but I would say contextualization maybe is better is the best but I think the biggest thing is Really, I think there's a fourth thing and that's kind of Asking the right questions of the data. It's not like what process it's not what regression model you use what technical technical technique you use It's not how you contextualize But it's about like what questions you ask of the data if you have some theory that maybe other people have That you you believe other people haven't had and and you know I know you want me to start leaking alpha here, but All right, we're back Reef is just leaked all of the alpha We had to cut it. This is how espn i pay the bills and how this podcast keeps running No, it was just getting too comfortable with his with his mic leaning back. He was ready to to let let it rip and then So finish up what I was saying yeah, and that like I'm not the best I'm not the best statistician like I have a lowly bachelors of arts in economics. Okay. I thought for years that I was winning because there weren't other smart people in this industry basically that like The smart people go to Wall Street and you know, that I was kind of competing against people that weren't I don't know Big fish and a small pond base But then when I tried to hire I've tried to hire people over the years Including as Kansi mentioned I think him. Yeah You actually explored doing something with basketball. This is like 2019 But but before that I remember it's slow in the air. We put out One of these post job postings and interviewed people and I had like I gave a A sample problem Like I don't want to say it's like a problem set, but like I wanted to know how people would approach A problem like for example. I think there was a tennis one I don't remember what the tennis one was, but then there was a basketball one where I just was like How would you given a play-by-play data set with score differential time remaining and like pre-game expectation? What would your approach be to build a live model off of that? I just kind of wanted to see how they think about it like this guy that had a statistics PhD from MIT was saying that he would like use You know half of the season to like As a test period and then and then he would look and see like how the lakers have done when they're up double digits And I was like you have this big brain clearly You could run laps around me in terms of your stats knowledge Yet you're not using any of it here the way you're thinking about it and and so I think that made me realize so much of it It's just like how you approach a problem and like that how you ask questions of the data and and thinking creatively And I think that is my biggest source of alpha more than anything technical like There there are people out there doing way more than I am technically. I think I'm just a good I'm good at finding hacky solutions to Problems that aren't really clean and simple. I mean I remember Because I've worked with Cade Massey. We have the Massey Peabody ratings for He was my senior thesis advisor sort of those in 2010 It's been it's hard to believe it's been that minute that many years by the way, but wow but I've You know, he's a he's a professor at Wharton now and he's good friends with the head of the stats department there The sky who's on his podcast Wharton Moneyball as well But I asked Adi once about something like it was a a Stats problem. I was like using some sort of like Ordered logistic regression or multinomial logistic regression to kind of like Try to essentially figure out the distribution in a very hacky way Might have been baseball related like number of runs or something and And I was like what what should I be doing like how can I do this better and the first thing he said is like oh This is so wrong like you can't do it that way But then he was like but he's like there's no distribution that fits this like like actually what you're doing is incorrect But it's probably the closest it's probably it's the least wrong way to do it or something So I think that's kind of my strength like the hackiness and just kind of like Finding a solution that isn't Elegent isn't like quote unquote correct or pretty, but it like it gets me most of the way there like getting the 95% of the way there So like small data problems I kind of see it as like your super power small data problems Like you're trying to beat the power ratings you're trying to be ahead of the curve on that team You know you're trying to get the faster Yeah, and I think Right with with enough data if you have huge data like you can just brute force things Right the PhDs would do better PhDs when they do yeah, right That's why like when I see these data sets now with NFL and and all the I don't know computer vision etc. I'm like Like what the hell am I gonna do with that like that's like he just like that's not my my strength But it could be I don't know I mean, I don't know that I haven't I haven't looked at that I've looked at some NFL play by play data for like some really basic problems But nothing like computer vision what actually I want to ask one more Question kind of on this is like I think you've taken kind of a stand against Machine learning into a greater extent like AI and I like to use Conzi's quote is like what the fuck even is AI But but like I think that that that came up You know you've talked about it and how like you're still mainly sticking to your your guns with the regressions like I assume you've dabbled into some stuff that's been more black box But I just wanted to pick your brain here, and you know, maybe the listeners would like to hear your thoughts on like why you if you still are like are Mostly rejecting black box solutions and like still winning. Well, I'm not smart enough to do all the AIML stuff So I kind of have to like just dismiss it, right? But I think that I don't um So I've been told that what I do actually is machine learning depending on how you define it. So I think I don't know like I mean I you know, I've used things like principal components analysis and I mean I do some you know things that are But but it's all I would say it's generally in the regression family um I just think that there's a lot of There's a lot of noise that you're gonna overfit with the small with small data especially data that is especially when you're dealing with the sport that is changing like Football now is not the same as it was 10 years ago um, you know with rules changes and stuff like that and if you if it's black box you aren't really able to sort of post well to able to sort of understand The weaknesses of the model and I think understanding what a model is and isn't covering is pretty important And so I To be fair, I just also am not as interested in that like I'm actually not as interested in building models anymore as I As I was it just doesn't motivate me Um, which is probably why like the prediction market stuff has been so refreshing because it's like a new problem Because I've been doing the same thing for so long not modeling and it's like had enough of it Well, let's let's talk about some of the the prediction Yeah, you could tell you he's a professional podcaster first and better second with that with that transition um, so Walk maybe maybe to just tee it off like You know the prediction market scene. I'm sure you're you know always looking for any and in all opportunities But the prediction market seems sort of at its surface Maybe like a little bit counter beyond just getting more more liquidity and in what not for your rich name But I think you've you've the way you've talked about is you're you're taking a more active role in just like making and and that side of it as well Instead of just like expressing your opinions. So walk me through like Why the prediction markets are interesting to you how you got started on what it looked like when you first started and what Whatever you could say about what sorts of things you're you're doing today Well, I think I first started with some golf stuff um and Like seeding some golf and competing against GPD I was gonna say it's always beating me to those laying the 98 cents on the guy's 175 to 1 odds Rufus it it takes a hell of a modeler to figure out 98 vs 99 and That's where that edge comes from. Yeah, but my process. I can't like You know without without my numbers. I'm not gonna just know right off the bat I mean, I guess I could just look at what you could run your numbers on Sunday night. No um So the problem is my code right now takes like that to process the previous week's data and to generate get everything done takes like eight hours basically. So oh wow You know, it's it used to take longer. I've made it more efficient But actually I got it down to like four, but now it's increased again just because of like more and more data A lot of the shot links. Yeah, we're doing more we're doing more and using more nuanced data. So more columns Well, also it's it's in is in stata and so yes lots of people are gonna laugh me like it and I'm not the best coded like I'm Somebody like chat GPT or somebody smarter than me could probably make my code so much more efficient but I understand it and You know, it doesn't I don't know. It's got me as far But so that linear regression mindset right there. Yeah, it works And it why yeah, logistic regression Okay, anyway, sorry. I'm trying to remember what like the specific trigger was for this, but I don't know. I think I think just seeing some of the liquidity on college football early in September um and and NFL and like looking I remember Looking at an NFL game at halftime and or in commercial breaks and just being like seeing the liquidity and the How like price inefficient prices seem to be And I just like I thought it was a big opportunity. I I didn't I thought that There's a dearth of good market makers. I think my experience um with Sagan golf um kind of made well like I don't they do a lot of things and I'm sure they do a lot of things really well But I think before that maybe I'd assume that these Guys had everything figured out And I didn't realize like there's a lot of I mean, it's it's a really tough thing to do There's a Everybody's gonna make mistakes um, but I just realized that like I think that there is an opportunity there market making especially with an opinion Because I think if you're just strictly top down I mean, you can be manipulated really easily like I've seen times in like call sheet where I'm I get hit for something like for big size and Then somebody goes and moves the market after right if you're just top that because like there's more liquidity on call sheet And there is on regular sportsbooks, you know, so if you're top down I think there's always going to be issues with that especially if you're trying to scale But if you have an opinion and you can you know integrate your opinion with the market um And you can get a lot more volume on Games that you wouldn't otherwise even maybe be betting like I make something minus 110 I'm not gonna be betting it maybe but if I can get it at even money like Clearly that's great and then especially with And with all the different markets and derivatives and stuff like that I feel like you can kind of edge stack a little bit as well Just from and and I think some of the like the derivative tools that I have are um, I think a strength of like a strength for me and so I've spent a lot of time doing distributions on some of these sports and and You know, it's not the same not every game that's minus 342 is the same either so So I think that there's some alpha there as well and but I think the the the thing that's most motivating Is just like yet is is that it's a new game and and the chance to kind of like build something Okay, I think for a long time. I feel like I've been missing like gold posts When I got started in the industry. I I mean how when I was working at LVSC When I got to a point where I was like, you know what? I think if I didn't have a job, I could actually scrape by I was like I think I can make 40,000 a year grinding blackjack and sports betting with my small bankroll um And then it was you know trying to like kind of rise to the top um and I don't know how But then at a certain point like you kind of it's There isn't like it's not like there's like a numerical goal or something like that right it's like I didn't get into this For the money. It's been really it's very nice, but But when you you know, I don't know. I just haven't had anything that I'm like working towards and that's the that's been the hardest part And and by the way, that's why that's why I started on a baited and I also had a A failed venture called G-street analytics that was Gonna be a soccer betting hedge fund Meet myself in another group of um people from academics and people from a different from horse betting We're all kind of got together and we're gonna try to try we were like gonna try to scale Try to win betting the one market that really you can scale But it was just kind of like I need new challenge and and I think I kind of was fitting uh Square pegging around whole with some of these because it was basically more of the same You know a like I think on a bait. It's great, but I don't think a consumer facing Company is like what I want to be doing It's so hard to define success there too like his success depends on Marketing and all this other stuff that I don't think frankly on a baited is done that is been our strength probably um And I you know you look at site like odds jam and and you know they have all these predator like Slimey emails and stuff like that and they've been very financially successful and like Like I thought it was cool to build tools that I would find useful that I found useful and I thought other people might and but but then like I don't know success depends on how that's marketed and and The stuff that I really liked building actually was fairly niche because it was for people more like me um, and so This is actually market-making on these prediction markets is actually a whole new different problem It's like and I've never been like a a trader, you know, and it's like you're thinking through all these scenarios You're trying to reverse engineer bot logic you're creating flowcharts. It's like it's so much fun and and building something like We're still very basic in what we're doing. I mean, we're integrating our numbers. You know, we under You know, we regress to the market and all that but you know, we're trying to get the best price execution you can and But there's all these games like but we're building this foundation. Sorry this foundation that I think is going to be the foundation of something Could be really really really great and that's what excites me The fact that like I'm learning You know every mistake is Is that you learn something from and you don't worth something right? You're like well at least that made me some money kind of Yeah, losing is an investment in future winning. I I very much believe that What uh What are some of the things that You didn't that didn't really matter for sports better or some skills or whatnot that now you care a lot about in prediction markets Um I don't know because I I would say Like I've never been as involved on the trading side Back in with the old group I was kind of building the models and the other guys were doing the actual bet execution for the most part Except like Super Bowl week and occasionally I'd have my like one day a month where I'd go run vets but And even now it's kind of the same thing like I do you know, I have some accounts obviously, but I'm not Um, I've been and I think we probably Leaked little alpha by I I probably have over the years by not maybe focusing as much on the game of getting down Um, we certainly do but I haven't I certainly haven't been the one as focused on that and So I I don't know how to answer the question except to say it's like It's a completely different thing It's it's right. I mean like understanding I mean, there's the there's the part about understanding your price and and all that but like there's also the part about I don't say Like disguising your order um There's figuring out q logic kind of planning ahead to figure out if you get filled here At this price like if you know, do you want two-way action the other side to get you an effectively better price Um And I'm just I'm such a beginner here but I'm learning Learning and getting paid hands in life If you're listening to this episode and somehow don't realize Where roof is sits on the call sheet leader board it's it's quite quite high up there Um Yeah, higher than me that it or not displayed. They're all hidden men show yourself. Come on now. It should probably just high I mean, I think I may have myself at some point honestly They're maybe upside Yeah, I kind of agree with that but um I think like Operating the open order book system like prediction markets is probably the the number one Difference because like you said like q logic is super important and we talk about tick size and and all of this All this stuff, but when you're when you started where you it seems like you're mainly Just going out and trying to get an opinion filled passively, but now you're Seemingly, you know, you're a bit more automated and will take both sides is that what do you mean when you say filled passively? Because you you've mentioned this word. Yeah, you called me a passive market maker or something I was like I don't know if I should be insulted or not No, no, no, like I don't think you're taking liquidity. It's all right I'm not take yeah, I'm not except for like occasionally you'll take It's just a big deck, but If there's something good then After yes, yeah, no, what what what I'm saying is like you're You have the you're not going in hitting the other side like you know you can get better price But it's your opinion and you'll get filled one direction on the game Almost always as opposed to like We are directional sometimes like But a lot of times we'll have both teams filled On a directional do you mean you actually have liability on one? Mostly are you saying you're okay? You're only seating one side No, no, you could be seating both sides, but like if you're only getting filled on one side in most of your games I'd say like you're mainly directional whereas like we'll often get both sides Yeah, no for sure directional, but also like Both sides it all depends on the sport and yeah, like if it's like Yeah, I don't want to get into more detail on that, but it all depends on the market like There are I mean if you're top down you'd be more if there's something you're top down and you'd be more Open to both sides basically and if you have your own number you'd be going to be directional But but there's some of I mean it's a blend and even if you're top down and just You know you're large you're gonna have big Physicians on one side often Yeah, no that that that's for sure, but the thing is even if you want to be directional You still Can benefit at least in a liquid market from being on both sides Yeah, I was gonna ask about that like have you changed your mindset towards Like paying what you think is a technically a bad price to your number to kind of get out some on the other side if it's liquid I mean, I've always thought that you should be willing to take negative VV things because like you know if if If you have Let's say you want 200,000 contracts on something, right? And maybe you make the fair 60% and you can get filled 200,000 contracts at 55 cents That's amazing, but what if you could fill a million contracts at 55 cents and then take 800,000 on the other side at 42 cents That's negative VV for you, right? You're buying but you're buying 55 and selling 58 and you effectively have a much better price than So you effectively at the end have 200,000 contracts of exposure at like 52 cents or whatever it ends up being Exactly. Yeah, like like the way I think about that is like this is the max would take an exposure But we're willing to be active and trade like eight times that in volume as long as the rest of it's all balanced out by the end like we'll take this much exposure But that doesn't mean like that's the max contracts you'll trade right but I think it's also important in Disguising your position a little bit too probably right if you were only putting up something on one side for liquidity Like eventually someone's going to see oh, you know There's 80,000 contracts here and there's like nothing else on the other side in this market like clearly this is Not your regular market maker, right? But like thinking about golf like let's say If you I think Justin Rose is a good example because he's a golfer that I think the outright was overvalued in the outright market this year at times because he had had a few really top finishes And I probably undervalued him but maybe because because of that I mean and I think data golf certainly was Lower on him than the market and I think most people I think this is one of the situations where it was kind of People thought that he had Well, he was priced Based on being a much better golfer than he actually was overall and it's whether you believed in the fact that he could just String really good performances together Or if you thought that was more just he got lucky in that he did that but I'm getting off track My point is like let's say Justin Rose is trading at 50 to one on It wreck books and let's say he's at 70 to one on exchanges and if you you price him at Your fair is 200 to one. What would you offer on a book? Let's say with Smaller tick sizes. Let's think more like a bad fair If you mean like would you jump the 70 to one or no Basically, well, you make him 200 to one you definitely yeah, oh, but didn't You say on the exchange is 70 to one or that's like somewhere else. Well, let's say that's 70 he's 70 to one on the exchanges like 50 to one at wreck books I'm just but I have another exchange where I can be top of book I'd probably be like and let's say you're on a like let's say you're on a different exchange. Yeah, what would you 52 You know, okay, and then you'd offer would you offer the other side Mmm, it depends probably if you're making on an exchange and there's only offer up on one side Like I think that's kind of like scary if somebody I would offer I would be like sure I'll I'll give somebody I'll let someone lay one to 125 against him winning Sure, yeah, on bet if on bet fair they could lay seven and you five to one or something right like like I'll take that fill there Because even though I it's a good bet in my opinion Um There's you can lay it off so easily Well, and also who's going to be betting it really think about the amount of money you have to post up for that Right, but I think having being willing to take that like if you do it's going to get you in a bet as long as you're getting orders on the just in rose Yes, people are betting the the rose yes, you're going to be getting a much better effective price by doing that And I think you you disguise your true Your true opinion much better Because if you put up that that's the main if you said like if you said oh give Justin rose at 100 to one and I'll put the other side at one to 300 like If you put up for big money someone's gonna be like okay this guy like Markets too high on rose and yeah you leak information just like I'm doing now No, I think oh god, I was gonna ask like what do you think we're gonna see increased Sort of games being played around some of that stuff in the in the near term. Do you see some of that now? I mean, I'm guessing it varies a lot by Which actual prediction market your exchange or exchange you are on on some of these stuff Um, but do you think you know as I just imagine a world where you know I think we talked about a couple episodes ago like this odds jam sharp money tool where the idea was to like identify the markets to have A lot of money on one side and and you know use that to your benefit like so that would be like an example of it like do you think we're gonna see increased Games being played on these fronts to try and counteract some of that. I think market making is all games or at least it should it should be or It is maybe they're not always being played well though And I think there's just there's different levels to all of it and again, I'm I'm still just trying to master the basics here like I I'm But I'm enjoying It's fun to go on to like a call she manually and try to figure out like exactly what this bot is doing and what What about it is good that I want to take and then what are its limitations and then how can I exploit its limitations like If you can figure out exactly the logic and draw a flow chart and be like what like I and and know exactly what this This bot is gonna do based on what you do like that's really valuable and it shows to that bot um This I mean every bot's poorly designed to some extent but but I think just but A bot a good bot and a good sports book should have randomness A book shouldn't a sharp book shouldn't max move every time on a bad eye make Right if it's I think you're predictability predictability can always be exploited Yeah, I think there's like a lot you said there that is really helpful one is I don't I think people undervalue how Much you get from going into an order book on call she and like hand trading something for a little bit just to see like how everything works Well, I don't know the good stuff has come from that You know jelko told me that I mean he still I remember once he actually called me um This was a few years ago. It was just out of the blue called me to ask about why Matthew Fitzpatrick was trading at Like this matchup was trading with him is like a minus 150 favorite against This other golfer and I don't know is the US open or something despite the fact that they had the same odds And he's like this trading this is very like the way this is trading with the volume on this side is like he he couldn't He had a he didn't understand something about it and he asked but But he's like he's actually you know, he's like the most successful better in the world and he still is Going on exchanges and and figuring out stuff manually and he says the best Like You're never going to have a bot that's better than the really good trader Absolutely, but so important but what yeah, but you can't scale having a really good trader Right manually right and so the key is to try to find The key is to try to Program a bot to be like that trader to understand exactly what to do and so it's it's hard like I'm I'm sometimes like the other day I went in and was Literally taking screenshots of the market before I did some action and literally I wrote what ended up being a 20-page document that I shared my team basically with it was like my investigation of Of a certain bot like to figure out I want to figure out exactly why it was doing what it was doing what and and And if I could actually exploit it and get it to put up a bad price This is like Charlie Day meme vibes, you know with the board and all the things It's like when this bot did this you know, but I should say I'm not like putting up orders that I don't want But it's like I'm just moving like I'm putting You know what I mean? I'm Trying to get it. I'm trying to see how far it will go I mean, I'm sure you have you seen there's bots that'll like in illiquid markets We'll just continue if you put up an order they'll always just jump you I mean if you have a certain thing. I mean they don't put enough to make it worth it There's like this penny jumping bot that's doing like 10 shares. It's just like can we just get rid of this person? There's so many of these but so what are you doing 10 shares? Who cares? It makes the book yeah, just join them, but then they jump their tent like they actually jump on joins a lot of time Yeah, at least like this one I was dealing with and you figure out how to make it pay so it doesn't do that anymore Yeah, that's the goal You have to teach it a lesson GP Just like you teach teach me a lesson when you bet San Verance Well, the lesson was you mean you won the money I can't that's right. So far I like to do when I've taken taking years Yeah, I don't know against riva's not it's let's just gloss over the fact that it's you know a guy to win a golf tournament The odds are like 51 but Um now I think what you said about like the best Trader will always be better than the best bot is something that like I Really had to even like learn from from my team like That there's going to be spots where you can be manual or you want to at least learn the logic man And that's why Circling back to like when you were saying when you lose you learn something I think being willing to lose a little bit like right now has never been more valuable Because in prediction markets like so much is new and you're really like fighting on the ground floor and figuring things out Like losing and learning like losing just has a premium right now like obviously not like dumping and whatnot But being willing to stick your neck out and like Have something fuck off or like you know be on the wrong price or or whatever Right now I I think is so valuable I couldn't agree more. I mean, I've made a lot of dumb mistakes and um I mean Sam Well, so some make some epic ones some Some of the mistakes you don't actually learn from you just learn like Shit like This is called human error and it can happen like a fat finger or clicking the wrong thing and Creating a massive arb with yourself Been watching a lot of money just disappear in a microsecond like you know that kind of thing It's like well, what are the safeguards? Well You know this yeah, I had a thing running this uh this weekend and it didn't turn off before the game started I didn't learn anything from that I had that kind of idea but but not really so I mean there are I think like The games part of it that you you talked about is really really interesting I think it's really important because You know, I think when we talk about it, you know people can get a false sense of like oh I'm just gonna like fire up some some bot and like connect to the API and print money and You know the points the points you made about like being predictable and whatnot Um eventually people with well, you know, it might not be happening today to the full extent But as people get better and better and more like experience like if you don't have like As at least if you're not at least like trying to think through these things like you will be You will be taken right so yeah Yeah, it's yeah, I think it's so tough to design a like a really good bot and every bot has a flaw And I think the more aggressive you're trying to be as a bot the more likely it is that the Well, the more Penal a flaw can be Are you saying if you're taking as opposed to make like well, yeah for sure that like the taking Bot take it all no Is I was like should it it's so dangerous like we found something the other day where like something that was supposed to like replenish in order like an iceberg thing like ended up um Somehow the logic was flipping it So instead of of taking instead of Replenishing in order at 43 cents that would replenish at 57 or is trying to but thank God I can't because you would have to take yeah Not taking his like I mean because there's so many mistakes that have happened yeah and then but but By not taking like we've eliminated We we've saved us ourselves hundreds of thousands of dollars for sure Yeah, yeah, absolutely like That that's something that Gabriel says like a taking bot can can just can blow up a company like if it's not properly You know if there's a bug like if you're passive like there will be mistakes, but it won't be as bad And like you can only test so much in a demo environment, but I would say sure I mean My inclation is to say let's if we want to do something taking we need a position limit of like test it with the position Limit of five contracts or something like that and and and just because let it go for that Yeah, to find the error because there will be that the there's all there's an edge case that you won't have anticipated that'll blow you up But a funny one in hockey there is this there's two players on the the conucks with the same exact name In first-hand last name yep And one of them's like a star center and the other is like a shitty defenseman and let's just say we offered quite a good price for that star center to Score call was on Kashi. They're like they don't have now they have the jersey number in there Like I don't think they have robust player ID thing Yeah, no, but they have the number in there and it it was just It just done for and if you're listening to this and you got that should be it should me a DM You're welcome for the 10 grand you can be on the pod in no Yeah, in two weeks. Okay. Yeah We have I mean that's the type of edge case like it just takes one edge case to just melt your bank girl Yeah, and as Peter your point we had this the same kind of thing happened with like an NFL well It pulling the start time for like the next week's games because that point we just built in like our for our our thing of it was like Okay, this is San Francisco against this team, but somehow it pulled it wasn't searching for date It was searching and it pulled in like the San Francisco start time for the next week's games and Let's just say actually might have been it was like Cincinnati. I think it was Cincinnati against Buffalo or something like that And this was that game that Cincinnati had That big Like jumped out to a big lead and Buffalo came back in one miraculously, but like You know our our Buffalo minus 14 and a half's were Taken quite happily by somebody once they once Cincinnati was up seven nothing or and so we Yeah, well, they informed you of the truth of the situation. Yeah Now like and now I'm always basically going to rest I'm I'm like gonna go into resting orders before things start and check like I was for a little while to make sure that that was Was going and then and we also also improved our architecture to make sure that won't happen Yeah I was gonna say we have like so many areas here We have a lot of questions. I'll probably get to the same thing So maybe we should we should hit the questions. I think we should hit the questions. We'll try and rattle through in about You know 2030 minutes for these these questions. Okay, so you had the record definitely for questions asked I don't think it was particularly close um Let's start with the with our friend Bunchu Because we're on the topic. What are the most impactful changes to your process both psychologically and operationally uh Since moving from they're moving a lot of your business to market-making. Well first off like We're still like it's that's not a We I want it to become a huge part of our business, but it isn't yet. I think it it we're we're the early stage early stages We're still doing most of things like the traditional way um but I'd say Psychologically it's been it's been a lot better because I feel like I'm building something and so if it in I'm learning and I think I wasn't doing that quite as much with sports betting I mean I've I was burned out as hell and like unmotivated for like five years And just kind of coasting by on inertia and you know, there's still things to get me excited occasionally, but like this is I don't know. This has been reinvigorating. My uncle told me At Thanksgiving dinner that he hadn't seen me this motivated in a decade Um, you just see it I went on my face and the way I was talking about things So like I think it's been great for me psychologically in that way. I don't think that was maybe the way he meant To ask the question, but um, but I do think every mistake As we said every mistake or things that go wrong. There's a lesson there. So it it you know and we're building a bet And we're using that to build a better foundation for something that that I hope to be like really really good in the long in the long run. So um Yeah operationally, of course, it's a completely different thing I think I think the biggest thing though is that It's forced me to think more strategically and like long term Kind of in a way which is kind of similar to building a good model Where where you're kind of it's not a quick process You're starting you kind of have this idea you're starting with this granular data and building it up And you kind of have this idea where you're going to go and and that's kind of what we're doing here And so I think that's um Yeah, I think the strategic thinking Yeah, it makes total sense. I mean, I think of a big reason That I'm attracted to it is it's you know, it's all a game sports bangs a game like I totally understand this but this you know coming from like being coming originally from DFS And some of the more like actual game like what your opponents are doing this is more interesting and and And just fun frankly for me so much more interesting and fun, isn't it? Yeah, it's just like not Not as much just like you know could make a number bet the number right and I know sports bang there's more to that But like That's been my life mostly make a number bet I remember or make a number right someone else especially because you've not been The other aspects Right All right, let's go to BA analytics question How much would you give to less variance in total dollars one on the year? So for example two people win one million dollars But one has a three percent ROI and one has a 14 percent ROI How much better or is the four percent tea or 14 percent ROI even better at all and the addendum was Please don't use the argument that the 14 percent ROI guy should increase volume and lower and have a lower ROI to make more money Assume it's necessary to have a higher ROI Okay, you can somebody tell me where it would be more necessary to have a high I'm guessing I'm not positive, but I'm guessing his argument because I've seen this type of argument before would be like You're gonna you're gonna piss off betting partners if you have too low of an ROI would be my guess of what the argument would be Which is fair Yeah, okay, okay, really if that's true though that there's that guy can scale a little bit better or that guy has there's more meat on the bone right I feel like what the 14 percent guy But is this is this is Porter asking this so that he can I'm trying to figure out his angle here Well, isn't it just it's just like kind of a sharp ratio right like there's just like a value to be had for having A lower fluctuation of returns You know as far as like if that's what he's asking like I would say probably just like calculate like a monthly sharp or something would would be my answer Obviously the sharp's sharp ratio. Yes the sharp sharp ratio and like the 14% person would Maybe not have a better sharp like maybe their 14% came from like banking like three huge scp's or something Right, and if one if the 3% ROI guy is in a big liquid market or something and he just isn't betting as much money Like yeah, I have no idea if that's like clearly you want less variance if possible, but I feel like it's hard to Compare these two like I mean, it's why I'm willing to hedge something at neutrally V always You know So yes, I think less variance is is good, but I think that It's hard to separate I think his come the two things he wants me to compare or very hard to compare but Yeah Give for them and then yeah, check out a sharp ratio maybe Big buck hunter a friend of both pods, you know, we love Henry Yeah, thoughts on the systematic long short article the I don't know it was like the prison of financial We talked about this action in the last podcast. I assume you read that or yeah, yeah, I did I really enjoyed it. I thought it was I thought some of the I thought the examples were kind of I don't know. I think the numbers weren't really super believable on everything, but um I You know ship or vehemently hated it. Yeah, I was gonna say there. There's there's a shipper camp which which SP is and and uh all the smart foreigners and then I I agree like There's a lot wrong with the article, but like I also thought it was interesting probably like Direction I correct yeah, I like that. I think it was directionally correct to and it it was an interesting perspective on it And it's something I hadn't really thought about so like in the fact that the poverty line is indexed to food costs Assume and assumes that food costs are make up the same percentage of people's total like expenditure in a year like Seems completely insane given how much housing cost of risen and health care cost of risen right like so I think we can I would hope you could agree that maybe that should be at least defined in a different way Yeah, yeah, I don't know. I know I think the For I get I'll just speak for myself like the aspects. I thought we're insane. We're like That everybody is like hypergeographic and when in the reality like That's such a small niche of this corner of the internet and if like you to me I go talk to normal people It's like they don't feel like they need to bet their entire life savings on like Crypto like they're just happy to like work their job come home You know have a beer like but like obviously gambling is increasing in popularity I just think the way it was like oversold in the article. I like the idea you know directly Directionally correct article with um with which maybe missed the uh the true the true median or whatever We got to give ship or something to y'all at us for so we'll call it directionally correct It was the it was the article about the gambleification of things If I'm thinking of a different article that I'd even call it like No, you're think no you're thinking the right one. It's not gambler. It's it's because there's different things that hit on Yes, so one is just like people are more inclined to gamble because they're They're you know costs are arising and there's this gap between like Being come I thought like oh, it's bad that people are like comfortable That was like one part of the article. I really disagreed with not to like really get this whole article But he was like yeah like well now like everybody's needs are taking care of so now like everybody is chasing like this Like higher level meta goal and to sp's point. I just like don't see that at all. I don't know. I do think people need to have goalposts Yes Right, like we have to have something we're working towards. I think these humans and that can be supporting your family or something like that Sure Working in the field working a far like on a farm or I don't know whatever it is like there's I feel like we need those little like external markers to sort of Well try and get the author on the on the party. See if he would support working on a farm. We need the out of boy right everybody needs an out of boy Next when speaking of shipper um if rufus could hire only one of sprauts better or conzi which would he hire and why Did you see kanzi's response? Yeah, that you didn't hire him So it's i feel like it's not sprauts. Yeah, I mean, no, I think I think they're re-kanzi honestly Because I think that I think I think they're both very smart. I think they're smart in different ways And that's why they work very well together, but I I think having a developer like Like that's what I would be hiring for more than You know, then mark um Yeah And I'd enjoy all the intellectually stimulating conversations with kanzi I like chatting with kanzi Yeah, I don't know sp do you have an answer for this? Um, I mean to me it really depends on like I don't know either of them personally at all it to me It seems like what skill set like like going back to discussion that That would work well, right? It seems like they complement each other well and Which when you indeed right, so yeah very different skill sets Uh both important both important so we do uh yeah I'm a lot of love and be a second one. How would you define the best these might be two separate questions How would you define the best originator slash modular if those are different things to you? You can answer them differently if they're the same Honestly, I don't know like what what Who makes the most money returns the most units the one who influences the market the most like I've never really thought about it that much So you definitely don't think it's the one thing I noticed you didn't say is the one who can predict the sport the best What is predict the sport the best mean though? Uh you can maybe that's what the question is is it like mean absolute error or something like that? I would say I don't think that should be the goal shouldn't be the most accurate model Like if you you can make any model more accurate by putting the line in it You know what I mean? Would you say accuracy without the line? Yeah, and and how much it adds that isn't in the line Right, I think that's that's Yeah, I think that's the most important thing Then probably I mean money's a good way to like filter for Competition to because you could have something that's like super accurate Yeah, I mean the sport that I mean certain people are more risk averse than others too. Yeah That's true. That's true. I mean like having somebody who's a greater originator I mean on an answer episode he talked about a guy betting $50 into pinnacle they'd move the line off of and like Italy or something, right? Yeah, yeah, so it's like so like that that exists like I know we like to all think like oh They would just move up if they're better, but something like you're right some people will just Are very academic, but they're not willing to really let it rip but they're still great originators now Yeah But who really cares I think that's What does it matter if someone is the best or is not the best We need a kalshi leaderboard for origination only I don't know how they'll quantify it or uh, no pennies. I sure is a lot We try to quantify too much That's fair and share and we all just be working towards our own personal Pictures okay, we we can get all move yeah, it's very like Mechanical and try to you know, but I'm very much. I live my life I'm more of a vibes guy in terms of how I actually live my life outside like personally and outside of like the fact that I build models So when a betting I you know, I love a good. There's a lot of depth there, you know, I do see that. Okay. This one is a tough one This is from brawner if you could design the regular Oh, okay But that's right like dr. Bronner's magic soap If you know what I'm talking about Is wait dr. Bronner isn't that like that medical mystery show? I don't know all I know is I got that shit in my Christmas stocking every year when I was a kid Oh, so I don't know the anti-clearly uses it Okay Bronner if you could design the regulatory framework for all gambling in the US from scratch and all states would have to go along with it What features would it have what things would you limit more than others? How would you split PMs and traditional sports books? So this question could be like an entire podcast right and I think you probably should do that at some point I think that would be a really interesting podcast I probably end up having to do it. We can never come up with topics so we appreciate the suggestion um So If I was like this are and could do that the first thing I would do would be I would outlaw outlaw mobile gambling altogether No, no betting on your phone No betting on the internet um I would make sports betting licenses easier to get and allow sports books to exist outside of casinos So you could like more like in England right you could have it but but I would still Well, I think there needs to be friction because I don't think sports betting I think sports betting is a net negative for society and in general things that are net negatives we put some friction and make them more difficult to To do so that like you have to really want to because there are a lot of negative externalities and so I think creating a friction that someone has to go someplace if they want to place a bet and they can't just like have a slot machine on their phone like that's I think That's what I would do. I know that's not In any of our best interests financially, but if I if If I was optimizing for society, that's what Yeah, I would do but again, I still do think people if you want to place bets should be able to but it shouldn't be That easy um In a world where we still did have mobile betting um What I would do would be create some sort of agency some sort of centralized um Company, I don't know if it would be a government agency and NGO or just a private company But that one company would handle all the responsible gaming the income verification the deposits the withdrawals They would see a ticker of every bet that comes in for me every account Link the different accounts from different sports books and they'd be able to make determinations on whether someone has Uh a gambling problem and if so they get cut off from from all books not just one because right now sports books all have a perverse incentive to cater towards degenerate problem gamblers because If they boot somebody out you know, it just helps their competitors who wants them right so I think right now the system is in setup In a way that with incentives that actually are or what I think we want As a society Like I think we value protecting vulnerable people um Even if most gamblers are not the most The easiest people to empathize with So we need uh we need the coal miner the SP always says that be our advocate Um, I absolutely am against icosino Full style yeah Um, how do we split PMs and traditional sports books I don't know. I mean prediction markets Like I mean the stock market is gambling too, but like prediction markets and events contracts are like I mean, I I like the fact that they exist right now from a personal perspective because it's fun but Um and potentially lucrative, but it's You know, it is gambling I don't think it's true think it's true Is gambling and I think that if if I don't think it's good Well, we got we got another question that was related to that so I'm gonna ask that now and then ask like a spin off of question I had so the question from Mookie was in an industry that is completely self-serving and overall Negatively negatively affect society. How does one rufus positively contribute to their their community and build self-worth in themselves? So, you know, you can ponder that the the thing I want to I'm also curious to get your perspective on is like How do you think of like gambling or sort of what are the reasons you think of a gambling as it's so negative to society? is it primarily like You know the addiction part of it for for a small, you know, subset of people is at the main drive because like, you know The if I'm playing devil's advocate the argument would be like it's an entertainment product Just like anything else and so is it is it is it the addiction part that's obviously like outsized as compared to something like You know movies or music or whatever or is it are there other aspects that you think are are like a reason you Because you you said it in the last answer like you view it negative. So yeah That's a great question by the way, and I think that the addiction is definitely a key part of it, but I also think And I know Isaac Rose vermin wrote about this in a recent Our sub-stack article I think but it takes people away from reality having this on your phone And you get you go down this like you're You spend less time with your family spend less time in the present. It's like social media in that regard That's addictive to and I think it makes people less happy overall. They're tethered to their phone. They're what they're they're they're It's the short-term dopamine hit that we think is gonna make us happy and overall it I think makes us less happy globally in long-term And so yeah, that's I wasn't actually I had never I hadn't actually Thought about mentioning that but you asked and I was because I do think the problem gambling is the like I mean, I'm not saying outlaw social media, but that is my opinion and social media that it's bad Well, yeah, I mean, I think that's the challenge is the only thing that really differentiates it from to me like a lot of other things is the the addiction aspect of it like You know, so like it's not that dissimilar from like video games Is that you know or social media or whatnot without that so it's You know, but the consequences of addiction are so much higher Than anything. Yes, then then alcohol like then drugs Alcohol financially you can and you can some like somebody can Completely like the bankrupt themselves super quickly Yeah, no Anybody here is like to debate that so but like that's why I asked is because You know it's a tricky question because it shares a lot of similarities with a lot of other industries beyond and there's levels of addiction right like you know People are addicted to other things too is just not the incidents right and not the certainly not the severity of like how messed up your life You can't mess up your life that bad if you're like that and you I mean, I guess there's stories about that candy crush addictions and why not very different right You don't go bankrupt, you know Are there you and far between you don't use you ever sure no with your house Yeah, yeah, so did you have an answer to to Mookie's question about contributing to to to community and building self-worth sorry for interrupting. I just feel like I'm just used to being on a podcast with Jeff You can interrupt. It's good. It's good. Yeah, I think that's a tough question. I think I Mean, I think I I don't I don't think I've done enough to positively impact my community. That's something I kind of hope to change But I think for me to it needs to be something that I find meaningful I guess you know and You know, I think the world is like pretty heft overall like I think I think humans have I think human stick and big I think humans have like 200 years left maybe and most I don't say it most. It's probably stick here like there's a world where they could live for millions more years but there's like You know, I think technology has made the world so much worse But at the same time, I think people are good overall Like and I think that betting has made me a much better person It's made me more process oriented and it's made me see the world probabilistically and I think I I view people I view people in the cut I contextualize people I think without thinking about it now. So why did somebody You know like say something mean to me maybe it was Yeah, I don't Probably because I'm an asshole No, but I think people's actions are they come from like All these little things that are outside their control like people that were abused as children are more likely to abuse people right like there's And I think understanding That it's not like this person is good or bad, but or this per or some action That someone directed towards you isn't about you and there's all the this nuance that sort of determines Or all these life experience all these all this context that sort of determines Um, what you see from a person and so um I I think that has helped me be a better person and And Yeah He's just like what more empathetic can kind of or that that's exactly like Yeah, I think this built self-worth thing is a different question um and I mean, I would say like self-worth comes from um self-acceptance mostly Yeah, well, I could I don't want to get too deep here, but like I think there's if if your self-worth depends on Something external. I think it's going to be more fragile and it's it could be there, but it it it is um It's vulnerable to Externals right and so I think it comes down to if you like into or Accept yourself like and it's not Independent of whatever your job is or achievements or whatever then like I think that's where Self-worth like a lasting self-worth comes from but So basically Buddhism and therapy Well, there's a sort of tangentially related. It's not self-worth But it's it's what when you found fulfilling from from satisfy sports Like and you talked about this a little bit of how you you've tried different things or maybe I've thought different things would give you a sense of fulfillment and maybe That's been left left short or not there So the question is whichever you've found more fulfilled it fulfilling building tools or helping other betters with unabated or being a pro-better I the way this question is keeping the alpha for myself. Yeah, that's very To be honest to be honest being a pro-better and keeping the alpha for myself Thousand percent agree. I don't really enjoy People complaining Yeah, people complaining and and and with betting there is a direct correlation between my work and how I do like success, right? And with a company that's consumer facing there is all this other thing other shit that I mentioned earlier and so It's I understand betting more That's probably why I will say what if it was like I love Interesting conversations with other smart people in the space I think that's why you have a podcast Yeah So maybe that would have been the better comparison Helping mentor betters is in a game like they I like games and Mentoring betters is a great thing to do and like captain jack is Like that guy does so much that people don't see and I know he's kind of divisive but like Like if you saw he responds to every DM and it's helped like offers to help everybody that asks And like I don't know how he does it. I have so much respect for him for like what he does that nobody sees Shadow shadow captain jack Um, okay, let's quick fire some of these modeling ones. Okay. Let's do this one about there's a couple golf ones They kind of bundled them into one so basically Someone asked as an avigolf better. He's trying to create my own model I'm curious how one learns to move beyond merely Effectively replicating DG's model to actually building something That provides an edge without being almost entirely tethered to DG, which is data golf and then like the second part of the question is what markets would you start with? Honestly, I have no idea how to answer that question. I think data golf does some good things. There's things that I mean they do a lot of good things but There's things they don't cover but I think you kind of need to if you want to be better than them you first need to Build the foundation Just like with anything Um, if you're trying to build a model like I think you're probably going to have to do a lot of things that they did too So I don't know how one learns to move beyond Replic I mean how do you go beyond replicating it have good I have interesting ideas and see the work see things in a different way I ask questions that other people aren't asking think creatively. I think it's the stuff you can't really teach I think if you can replicate it just using shot data and like Yeah, obviously not um, you know you have to get that data So that that's one part of it But if you can replicate what data golf does without using any of their like data as inputs you're already there Because that would mean you have their data like they're like adjusted strokes gain or something like that Well, you know, I don't I don't know what's you know exactly what the person asking the question is But I'm just saying like if you can take your own if you can take shot data And have it spit out data golf's number then from there It's much that's the hard part now you can make tweaks because now you can understand everything that's going it right So like that's kind of like if you're replicating it well with your own data set I would say like you might not even Be at you might not even need to ask us this you're probably about to figure it out anyway I don't know Let's maybe do I think we touched on Gavin's question about one sport to out there when we talked about you know college basketball earlier I also think we hit on Andrew's question on on the mental model and some of that you know Aspects we just talked about so maybe let's hit Justin's and then and then close it out here So Justin's question was how much more is a your a ball knowledge question how much more do top golfers And college basketball coaches know about their sport than you maybe this is this is more relevant to everything based on what And how much you and how much of your model error comes from that knowledge gap so like if you if you had Maybe another way to say this or at least the way I was thinking about if you had the brain of like you know Tour pro or caddier whoever you thought scottie sheriff like what like what do you think the delta in sort of like I know it's hard to quantify but like how how much do you think you'd improve your process model whatever Um, I think I would I think they know way more than I do and I think it would Not help my my betting very much at all. I don't think it very contribute I don't think help much of my model error it all comes from That knowledge gap. I think it modeling sports is different than playing them and you have to have some understanding curiosity But I also think that people that are close to it oftentimes um I don't want to say I have a biased perspective. It's like you can't see you're not the most objective person to look at yourself right Um, you kind of have to be delusional. Yeah And so you tend to I think they would think things certain things matter way more than they actually do And um, but I think I think most model I mean there's there is always going to be model error because there's things that like That are unknowable right you have a decay function that fits the data well overall But you might not know that a player started seeing a sports psychologist or changed his clubs or like Had a divorce or something that right change or I don't know um You know took creatine powder instead of creatine gummies under the course It's a very niche reference Um, so But there there's there's always your reducible uncertainty. There's always stuff that you know, I love that word I think Kate introduced it to me your reducible uncertainty like there's The variance you'll never be able to get rid of Because it's just inherent in the process So That's my answer. I think that's that's a perfect way we have we have two questions to close you you preface that you're not sure whether you had an answer to them The first would be a hot take or opinion that you think your peers, you know, the the people of of sharp gambling twitter or or otherwise, you know You sharp gamblers you you that are in your circle would disagree with I think CLV is in as important as most people think if you're an originator I think that like I've always built models that I tested and had to beat the like against closing lines like I wanted models that would be the closing line and Yes, it's harder than being the opening line, but If you're looking at things in a fundamentally different way than somebody else is then you're going to come to different conclusions And at some point like and if you can do that in a way, it's directionally correct and you know Then You may not be getting CLV and like sub CLV that you create yourself and obviously that's it's hard to kind of To do that and I think most people I think most sharps are kind of doing using the same types of Techniques building the same types of models. There's a lot of overlap there And I think if you can do something differently You can win without getting amazing CLV And I've never I've never gotten great CLV um In well engulf you have to know I guess engulf I like but But I created myself kind of too Yeah, so I mean betting like Like I don't know overall overall overall over the course of my betting career I don't think my average bet on a major market has beat the no big or the big free close not the not the close at the Yeah, we're not sitting or not like you're we don't do sponsors on You know the market is in the market is more often moving towards me than against me, but but If you look at my actual ROI versus my what I guess you people call X ROI You people you people I call XR I call XROY I built my XROY is defined by the Game result. I actually had this idea. God, I keep touching my phone just as a I don't know not a nervous habit You know what I mean it'll and then it like starts talking to me accidentally so Technology's bad. I know I hate it Except I'm addicted to it but Basically I haven't My What would I expect what a top-down better would expect to earn based off of my bets is way less than what I've earned like Most of my edge has come from Like edge I still like I still have at the close. Yeah, so that makes sense. So I think markets I think for top-down better is the closing line is The end-all-be-all because it's all you have Right, it's more efficient than any other market at any other time But I think that the market isn't always it's not the end-all-be-all overall It all depends on your perspective I think that's really interesting being just be especially because I think lately there's been You know, just some of the discussions on gambling Twitter. I've been like You know moves that close. Do they move enough to they not move enough like as much like you know stuff around that and I so that I think that's interesting to hear that point about Assuming like you know the closest efficient like most of your edges actually come from like The true price being different than the close is set a different way. So yeah All right last one. We'll just a nice GP. You have a great one. I can do a quick one to set it up This sometimes yeah, so this is amazing up there. This is something you've changed your your mind on Recently so mine is keeping the theme of the podcast like it's okay to have Non-measurable goals so like I'm or if you may have been someone like or have someone like this talking about the goal post But I like to have things that are very very tangible if I can Um But I started with one of my good friends and he was like he's doing really well You know, he has he's married has a kid on the way a good job But he's like yeah, I don't really feel like There's anything like really specific. I'm working towards and I was like well I hear you but like on you know, everyone look at you and say like you're really successful So like there's got to be something there where you're like I'm doing well or I'm I'm accomplishing something like it's it's going well But I do so it's made me think about like am I too specific in my goals sometimes and are there more ambiguous types of things that I can be happy working towards I think yes That's really good SP said I can only go if it was good It was good. Do you do you want do you want I guess I can give a brief one that just came to me um spent To originally from the north spent three weeks back up there You know for the holidays. I don't think I saw the sun once I am reaffirmed the the benefit that the sun sun does to mood and and whatnot I knew it um, but man, oh man. That's that's depressing to go three weeks without seeing the sun damn I was I was up in Maine for for Christmas and New Year's and It was we had a white Christmas and I just it was really really cold. It was like 15 degrees every day. I felt like and Because that's what the temperature was actually but Actually felt colder because the wind Now or it's But it was amazingly serene and beautiful and maybe you know what maybe here's what I've I think Maine in the winter might be more beautiful than Maine in the summer. I mean the summer is amazing But it's just different. It's just a different beauty and I think I've When I was younger, I kind of wanted to spend this up from the DC area. I preferred spending our holidays in DCs a family and I mean we've been doing it in Maine in the last maybe decade or so and at first I didn't like it and I've come around now I can't imagine Anything else. I love so that's not a very big global Bayes you know, I think it doesn't have to be and it works on a hot take level to be so I think people would It's hot actually. Yeah, so summertime means amazing. But yeah, that that's what I was saying You're so good and peaceful about like walking in the woods in the snow I like it as fee and you're completely opposite on that I wouldn't want to live there in the winter. You get to visit the winter. That's the best you visit I went to I went to Iceland too. It was like one of my favorite trips ever You know like I've never been to Iceland or summer. I'm sure it's wonderful too But like I want a vacation to the cold, you know, so I think I like the perfect perfect way to close it out Perfect. I gotta get I gotta get do do this deer kiss And I felt like I'll cut and hopefully get you didn't me before that. So okay. All right. Well, thanks for coming on And everybody for listening and all the questions. Thank you We will see you on the next episode