Pluribus Poker Ranges – RFI LJ

Let’s us look at the opening ranges of Pluribus when UTG/LJ, basically first to speak pre-flop; he can fold, limp, or bet… and actually never limps.

So the big surprise, if any, in the range is the bet sizes, going from 2BB to 3.5BB. Some of the other pros had a bit of variability there but not that much. Pluribus opens 18.6% from the LJ, which is a bit wider than what we’ve been used to with Snowie and others.

The only “real” surprise in actual hands is probably suited Kings, playing 100% with K8s, maybe with K7s (only 2 hands), and 2/3 of K6s. I find very odd that once each, AKo and JJ were not opened. Then more suited connectors than we usually see, with apparently a dislike for 87s. Obviously in order to open a bit wider, there are a few more offsuit broadways mixed in, like KT/JT at low frequency, and KJ at 55%.

Another surprise for me is there is not that much mixing up the hands, some, but very reasonable. Sizes however are all over the place.

Displaying here ranges with some additional data on number of times Pluribus actually got the hand, percent open with it, and average, min, max bet sizes. Because the statistical significance of frequencies is limited (although not too bad for RFI), it is important to consider how many times Pluribus was in a given situation. For instance, he got AJo 17 times, and bet each time, we can be confident he opens at a high frequency with AJo. On the opposite with T8s, he got the hand 4 times at the LJ, and bet once, so we compute 25% but that could end up being 10% or 50%, all we know is sometimes he bets it, sometimes not.

So it is important to not take percentages for granted. Still the chart gives a very good idea of how Pluribus opens.

Pluribus AI poker range RFI LJ
Pluribus AI poker range RFI LJ

Bet sizings are best looked as a distribution with percentage per sizing, in number of BBs. In general I would say sizing lower than what we’re used to, but with a lot of variation with sizes going up to 3.5BB. Either the CFR algorithms never converge on opening sizes so it stays more or less random or there is a real reason. Can’t really see what this achieves, again some of the pros did vary their sizings slightly so there must be a good reason to do it. What do you think ?

Pluribus poker bet sizing RFI LJ
Pluribus poker bet sizing RFI LJ

6 thoughts on “Pluribus Poker Ranges – RFI LJ”

  1. Every action of Pluribus is sampled. This explains the variability in bet sizes and other actions. For a particular situation the strategy may be for example 30% bet 2x, 40% bet 2.5x, 25% bet 3x, 4% bet 4x and 1% limp or even fold. Other actions may have something like 99.99% fold and 0.001% raise for example. This is to be unpredictable, in fact human players were able to study the hands and the cards Pluribus had after each day to try to find holes in the strategy, they found none. It also does not adapt to other player strategies, if it did so it would deviate from its ‘quite optimal’ one and become less strong. It tries to play perfect in an statistical sense, not to exploit anyone. Also 10.000 hands is just a drop in the ocean in comparison with the billions of games it played agains itself to create the strategy, it may change a probability from 30% to 30.01%.

  2. I take the variations in pf bet sizing likely to mainly mean it effectively doesn’t matter as long as the mean raise is in the 2.25 to 2.5 range.
    The different bets are probably vestigial like many things in biological evolution: At earlier stages Pluribus had different RFI hand ranges for raise small, medium, and large that were ideal for those specific hands in each category–such as raise larger with AKo and JJ to protect them and raise smaller with AA, KK, and AKs to induce calls. Then, overtime, to protect more limited hand ranges in each category from being exploited, it had to keep moving some AAs and KKs into the big raise category and some AKo and JJ into the small raise category. In terms of exploitability and expected profit, it probably doesn’t matter to have always the same raise or always the same mean raise with random variation. It’s also possible some hands are still slightly favored for big and small raises, but only by tiny amounts.
    >>As a human (or even a computer investor) you don’t want to vary your raise amounts, bc it increases your variance. But Pluribus faced no penalty for higher variance based upon how it’s games were scored–i.e., based only on expected win rate. But with any bankroll other than infinite, you want to reduce variance in the real world. Likewise with any fixed amount to invest, you have place smaller bets or play in smaller poker games the higher your variance.

    1. Variance reduction is super important and not just for poker. Real life (well, non poker life) application of the sort of things Pluribus does without managing bankrupcy risk is crazy and can’t be implemented. Like increasing super long term EV by 1% by increasing short term bankrupcy risk by 50% is useless and application-less. The assumption of infinite resources is fun, but not real.

  3. I抦 now not certain where you’re getting your info, but good topic. I must spend some time studying more or understanding more. Thank you for magnificent information I was searching for this info for my mission.

Leave a Reply

Your email address will not be published. Required fields are marked *