Week In Review #52
Last week, I was on the Double Pivot Podcast with Sunday Specialist Mike Goodman and his co-host Michael Caley. They took a break from talking about soccer during an international window where nothing of interest happened (apologies to Italian POTD subscribers) to talk about me about Punt of the Day and Canadian soccer, but mostly we had a wide-ranging conversation about optimization culture and how it has trickled into all areas of life, including sports and poker.
One specific thing we talked about was the rise of sports gambling, and as a sports fan and gambler, I have many thoughts on how sports gambling has become more prominent in society, which I talk about on the pod, but I have one nitty-gritty technical gripe that I’d like to delve into further detail here for POTD readers: There is a difference between a bet being unlikely to pay off and a bet being bad. You can enter a poker tournament with a 50% ROI and expect to lose money ~85% of the time. It is a good bet, because some of the time you make money, you might make 100x your money. Even if the tournament were, say, a 1,000-entry winner-take-all tournament, you could lose money 99% of the time and still have a healthy ROI.
In mainstream media coverage of the sports betting boom, bad articles like McKay Coppins’s Atlantic piece1 and good articles like David Hill’s Rolling Stone article have been written about parlays being the ultimate sucker bet that are giant profit centers for sportsbooks, which they are, but not for the reason that I think the public thinks they are. I think people tend to view all bets with long odds as bad bets because they will lose most of the time, but that is square logic. If the payout is large enough, you should be willing to lose most of the time; your success is measured in money won, not number of bets won. Again, you look at tournament poker. If parlays were simply bad bets because they rarely win, then sharp bettors could beat sports books at their own game and bet on big favourites. The reason you can’t win at sports betting is because all sports bets are vigged, and the reason parlays are profit centres for sportsbooks is because when you stack losing bets on top of each other, the other house’s edge grows exponentially.
The math of parlays is simple: Edges, whether they are negative or positive, compound. Let’s say I offer you even money on a fair coinflip, that is a break-even bet for both sides. If that coin is unfair and it’s 55% tails2 and 45% heads, and I pick the favoured side, you would have a -10% ROI and I have a +10% ROI. So if we flipped the coin 4 times, bet $10 on it each time, you’d lose $4 on average. Instead of doing that, let’s say you bet $40 on a parlay that paid out $640 that it landed on heads four straight times. If the coin were fair, the bet would be break-even for both players. When the coin is unfair, the heads bettor would not lose $4 of his total $40 wagered on average, but $13.76. The issue with parlays is that each additional vigged bet you stack on top of the initial one causes the house edge to compound positively and your edge to compound negatively. This is all a little technical and beyond the scope of gambling writing in the mainstream media, but the simple “parlay = bad because you always lose” fallacy does irk me, and I wanted, in my small way, to correct this fallacy.
As I worked my way through this parlay math in my own head and talked to Caley and Goodman about it, I was led to another thought, which was prompted in part by the ups and downs of playing tournament poker. Is the problem people have with sports gambling that the game is rigged (which it is), or is there a more puritanical valence that any sort of risk taking is bad because losing money gambling is bad? This led to me a silly thought experiment that I’ll ask you, dear reader, to humour.3 Let’s say that someone with unimaginable riches who we all agree is bad (pick whoever you want) started running +EV lotteries. Let’s say once a week they sold $1,000 tickets to 10 million people (who we all agree are good) and gave the 100 winners $1billion each. From an EV standpoint, this would be good. Every week 90 billion would go from bad people to good people. From a societal perspective, this might be disastrous. You would create millions of people who would lose $52,000 year gambling; would that be worth creating 5,200 new billionaires? How much would you need in savings or annual income to play the +EV lottery every week?
Obviously as a gambler, I welcome this new +EV billionaire lottery, even if the famously aggressive Kelly Criterion suggests that you would already need to be a billionaire to consider participating. If anyone wants to start swapping action with me in this lottery, let me know. However, if you are a non-gambler or gambling agnostic person reading this4, or more likely, someone like me, a gambler trying to think of the psychology of a non-gambler, I think it’s worth interrogating where your personal line is. An EV maximalist might say that any bet that has positive expected value or positive expected bankroll growth is good for society because it takes money from casinos and puts it in the pockets of citizens. Someone who cares about preventing societal harm might say that any bet that carries a significant amount of financial risk that truly can damage lives is just a bridge too far, no matter what. If that is your concern, then parlays are worse for society than straight bets, because in the short term they will create more losers. But to me, the real scourge of parlays is not the idea of making a bet with long odds, but making a bet with long odds that do not have a payout commensurate with their odds. Parlays are bad for all the reasons people say they are, but as someone who has spent a lot of their life trying to hit a score in a big +EV parlay (what is a poker tournament if not a parlay of hundreds of poker hands you play in succession?), I wish people focused less on the long odds and more on the unfair odds.
Additional Sims For Premium Subscribers
This is the first three-punt week, but rest assured I did still write five punts, which means once my second child is born, we will still have plenty of punts for you to read. Premium Subscribers are given access to a Google Drive folder where they will also be able to download the raw files of sims I used to write my POTDs, sims that are more accurate and appropriate than equivalent sims in the big public libraries. This week I uploaded
An HRC sim with FGS-3 for POTD #261
A PIO ICM sim using custom preflop ranges for POTD #262
A PIO ICM sim using custom preflop ranges for POTD #263
Additional Analysis for Premium Subscribers
Everyday Premium Subscribers get an extra bit of analysis not included on Substack. Today, I’ll share #onemorething from POTD #261, where I wrote about why I think it was important to use FGS when analyzing this hand
POTD #261 onemorething
I think it’s important to use FGS in this sim because it gets the correct UTG shoving range … and that’s about the only reason why. The MTT Database Review sim shoves 10% UTG compared to my 17.5%. It’s possible my 17.5% range is too loose because it doesn’t see that once you post the blinds and the button, Ketzer will get free hands, but I am confident it’s a better shoving range than the MTT Database Review. Both in theory and exploitatively because I don’t think humans call that shove nearly as wide as the solver does. I’ll also note there is a very minor disincentive for Elton to bust Ketzer because it means he loses a free hand UTG. However, once I got the UTG shoving range and just ran a vanilla Malmuth-Harville with UTG node-locked to the FGS-3 ranges, the outputs for the rest of the hand are more or less unchanged.
Adding FGS does not always make a sim “better” there needs to be a reason to use it. If Ketzer shoved the HJ here, I wouldn’t feel the need to run an FGS sim for the sake of thoroughness, but it was a key detail in this hand that helped me create a sim that I think is more accurate than anything that has been posted publicly. Solvers are tools and are only as good as the person using them, using them efficiently is an important skill and a bigger game tree isn’t always a better one, but in today’s hand it is.
Media
I was on the inaugural episode of the podcast of POTD copy editor Nath Pizzolatto, the aptly titled Cards Speak. The first episode is on YouTube and you can watch it here. [Nath here. If you want an audio-only podcast, you can listen here, or it’s on most major music services. We had a pretty wide-ranging conversation, including the good old days and things in life outside of poker. If you enjoyed it and want to support the show, and would like advance access to episodes and other perks, you can subscribe to the show Patreon here.]
As always, I can be reached on
Substack
Instagram
Twitter
Bluesky
which never fails.
In memory of the late David Sklansky, we can call it a Special Greenwood Forum question.
There are thousands of readers. Odds are a couple fit this bill.
"let’s say you bet $40 on a parlay that paid out $160" Typo, should be $640 of course. Even the sportsbooks would blanch at offering 3-1 on a 15-1 shot (well, some of them).