Circles Off - Ep 161 - Matthew Trenhaile Part 2

Podcast Release Date:
Podcast Link:
hagrinBPR Review Date: 2024-08-01 08:46:20
Rating: 8.25

Notes:
Bias - I consume all Trenhaile content. This is 3 hours 14 min long, I'm not sure I will be able to fit all my notes into a single tweet so might condense my notes for this one. Pretty sure a Trenhaile + hagrin podcast would cause the heat death of the universe as neither of us would ever stop talking.

For those who don't know, Trenhaile resume - former bookmaker, syndicate trader, forex trader, Anubis Trading Limited (consulting). Part 1 was rated extremely high, maybe the best CO episode which was 2+ years ago.

8:00 - Crypto Exchange Arbitrage: Tren talking about the Kimchi premium and the arb existing due to capital control laws in the foreign country. Sports betting equivalent - Tren when approached likes to be able to explain the mechanics behind those types of situations (non-payers, hold your money, etc). For Anubis, getting approached by both sides of the counter, operator skewed. 11:00 - Johnny asks a great q about "how does it work interacting with both sides?". Covers his answer here, but then an interesting point about much of his rec book consulting has been around app dev and no real questions about player limiting, etc (interesting).

16:30 Trading Systems: Recent advancements - scaling of bet processing, backend engineers, cloud scaling, new products requires more computational power, live requires computational repricing. Ex - soccer live > player prop updating + correlation between them + in-run SGPs + Cash Out + Cash Out only 1 leg of a live SGP = you see this gigantic computational need. Tren brings up a great point - on operator side, this creates a tech and computation moat - not may ops can do this. Conversely, many of the trader interfaces, bet "tape" hasn't changed much.

23:30 - Consumer Expectations: This is an interesting topic because let's say you do believe in some of the longevity stuff espoused online - how exactly are your normie consumers acting now? Tren brings up parlaying live on "game script flips". Rob brings up the old NFL Sunday chasing 1pm losses at 4, now they are chasing live "the chase has come much sooner". Bunch of the typical SGP talk, Rob brings up the SGP marketing issues. 40:25 - "SGPs have hit Live volume" - Tren. This is a very interesting observation especially if you're trying to hide your action within the masses. Tren mentions there was some operator worry that live volume was going down, but then the margins were healthy so it's just a consumer switch to SGPs.

50:00 - "I feel like that GRP video was one of the significant Twitter events of the year?" "Which video?" "When he bought his New Balance sneakers?" Very funny Johnny lol.
51:30 - User Retention: Tren thinks the SGP provides a more high/low experience that can be shared with many people. (This is completely outside of my knowledge area, but this rings true. I would point to some of these X/TikTok guys giving away free picks/content scamming and the posts are highly traffic with very emotional comment sections geared around parlays and SGPs so rings true.)
56:30 - "You don't have to wrap up any live data. The player who places the pre-event SGP is now the most valuable player in the book aparat from the super whale." Wow, so this was the first light bulb moment for me this podcast. To put words in Tren's mouth, as these huge computational businesses continue to require more compute for these new applications and features, valuing or measuring the value of the customer by the amount of compute needed to service the customer. Absolutely have never thought of it this way, fascinating extrapolation out of measuring apps or features based on a compute needed/hold or $ earned ratio. Brilliant. "The hyperspecialization to lose at the required rate." Great line, Tren is cooking now. Rob - "The good player (from op perspective) - highly engaged with the product, lot of play days, multiproduct / casino".

1:02:30 - Sportsbook Weaknesses: 1:04:20 - "The observability of these liabilities can sometimes be hard to display a meaningful real-time way to trading teams." Great quote. Better watchdog apps for traders. Ex - "How do you know you're seeing an uptick in an event where maybe your data feed is slow?" No real answer other than a generic liability limit. Tren replies to people like me that you're making so much money, you don't put the effort into closing these small holes. Data warehousing - massive databases (player data, real-time event data, etc.). Data feed aggregation.

1:11:00 - AI: Feeding player data into machine learning which in turn can drive marketing, timing, etc. "AI personalization". 1:14:45 - Funny stories about using a soccer model for ice hockey, then another story that old Bet365 used to use the rugby model for NFL.

1:16:00 - Attacking Sportsbooks: "First, don't project further than 10 years out." Brings up crypto (for some reason), but he's absolutely correct (although this becomes a very abstract conversation about "killing crypto" and the spectrum on which that exists. I will infer Tren means price appreciation or a full throated Operation Chokepoint 3.0.). "It has to be synergistic with the sportsbooks - so it has to be parlays". 1:23:05 - "Limits are undoubtedly wild for player props compared to the quality of pricing". "These were markets that in the UK were done for $200-300, here in the US it's $2k-$3k." (Rock Flag and Eagle Matt). Some talk about attacking political markets (Polymarket has been fun the last 3 weeks I would agree). After this there was some discussion about can you attack large liquid markets like the Euros. A lot of the discussion is a lot of the talking points you'll hear about tournament soccer, but Tren does bring up BTTS (both teams to score) around 1:34:30 and how pricing that market is a difficult problem. (As a soccer bettor, I have a lot to say on this topic especially on this Euro specifically, but Tren still has 97 minutes left. Just a reminder note for me - my modeling of Jude playing centrally, lack of Palmer inclusion, England pricing, DrunkenGoon's comment on pre-Euro pricing).

1:37:00 - Talking Publicly About Good Info: Some discussion about "even if you give out good info, people won't put in the work". A lot of this section is not exactly correct and it gets back to much of the same analysis for scaling betting tools/products. A thousand people might whine it's too much work, but if one Spanky or one Logan realizes the value in what you're saying, suddenly the info gets assimilated into the market. Basically, the variable that matters here is not the # of users who understand the info, it's the amount of liquidity that will be consumed by the user(s) who understand the value. There's then some "people just want picks" talk. I've found that even that is too hard for some people where they won't take +EV advice in search of higher ROIs (see the crypto channels in the RAS/RAS refugee discords).

1:46:10 - Ethics of Bearding: Trench distinguishes between placing a bet for me vs here's my acct info login for me. The 1st you can frame as touting (or "consultation" which was the word du jour when this issue happened) & is likely free and clear legally. 1:50:00 - interesting story about harvesting utility bills to open up accounts and Trench mentions that it ran dangerously close to criminal. 1:54:00 - "It's almost like a question of scale and organization level can kind of make the difference in perception." (which he counterargues himself maybe that perception is wrong, fascinating debate). Turns into an ethics conversation about taxing operators, the ease of taxing ops b/c of a segment of society's outlook on gambling, the costs of running a betting op due to these taxes/costs, etc. Discussion turns into sharps complaining & then Trench discussing the mechanics of automovers & laying off sharp action & how difficult that all is. 2:13:00 - Very brief Greek mention which was in reference into laying off sharp action & some laughs about Spiros. 2:15:10 - StarLizard, people competing for his bets which were so valuable that they would use to destroy unaware books, "at one point if you had a British accent, you were assumed sharp". 2:18:00 - Some discussion on NBA injuries and Johnny claims just pull it OTB, Trench correctly states "yes but this is a race condition" to which Johnny says no. (There's actually a way to get *some* NBA injuries faster than Twitter so Trench is right about it being a race condition. The very funny part is this whole process was discovered via ... a Betstamp user!). Wow @ 2:20:20 Matt nails the edge and says it out loud. Then a discussion about Johnny running a sportsbook where it's a little unclear the *exact* ruleset, but basically unlimited sides on major sports. Personally, I think theoreticals like this are extremely hard & they actually all bring up the rec bettor % of the customer base issue and I think the best way to approach this segment as a listener is try and think about what you would do in that hypothetical situation.

2:23:00 - Elite scraper comment. For me, thats AbnormallyDistributed. I've seen him host data that I have never dreamed of scraping. He's also made data freely available hosting it himself. He has data on sports that I actually, even now, don't know what the industry standard data feeds are. AD's resume in the public space on scraping data far outweighs what I have put in the public space. Now, speed is another issue, but just no way of judging speed. But the correct name drop here should have been AbDist. AD & I not being friends is one of my life's great tragedies considering how closely his stories in chat match my life. Great sense of humor too.

A lot of talk about the Be Better Bettors episode they can't remember the name of - it was the Danny Kramer, Kosher Boys episode which I HIGHLY recommend.

2:37:30 - Current Frustrations: Generally, the complaint was people pretending they have solved betting (well shockSTUN). "All the recs hitting the England price (during the Euros."
2:44:45 - Current betting: Predominantly golf.
2:51:55 - Funny line "Competitive eating full under North American sports at Pinnacle."
2:56:20 - +EV - Good convo about "wisdom of the crowds" vs "wisdom of experts". "Curating your information sources." Rob's +EV move was if your ice coffee dilutes as the ice melts, use some frozen grapes which is a wild suggestion which I have never heard I have to try this now (Johnny brings up whiskey stones, I don't drink so TIL here too). 3:10:05 - "Incredibly successful people are often the product of incredibly unique circumstance." "Do not outsource your thinking." Two very good quotes (as I have just outsourced my thinking). And with that, we're done.