Profiling Opponents in the Absence of Statistical Convergence Pt. II

In my last column, we noted how we must constantly make playing decisions in poker without the benefit of statistical convergence, either because we simply do not have enough trials with a given opponent for any statistic to approach convergence (for example, we are facing players new to us) or because a given situation occurs so infrequently that a given statistic will essentially never approach convergence (for example, flush board check-raise stats).

At this point, we have three reasonable choices:

Default to basic strategy: With little or no information on an opponent, you should default to basic strategy, which involves making the best play – on average – against the universe of opponents.

Play the percentage directly: Take the statistic at face value and play the percentage as if it were statistically significant, or – and more to the point…
Profile the player: Use the statistic in the context of other statistics, the way the player has played other hands, your history with the opponent, and other profiling factors in order to improve the accuracy of your decision on a case-by-case basis.

Let’s discuss the first option.

Defaulting to Basic Strategy

With little or no information on an opponent, your play should be to default to basic strategy, which is non-player specific and non-exploitive. Your basic strategy for any given game involves the best play on average against the universe of opponents in any given situation.

In blackjack, for example, the default play under basic strategy is to hit 16 vs. a dealer 10, rather than stand or double down. What this means is that – in the absence of other information – if you hit 16 every time a dealer shows 10, you will lose less money than if you were to stand every time, or do anything other than hit every single time (for example, double down sometimes, or hit sometimes but stand other times).

Similarly, in poker, you should have a default basic strategy involving a standard play for any given situation. For example, let’s say it’s $0.50-$1 no-limit hold’em online with $100 stacks. You open with a raise to $3 on the button with the A-T offsuit, and only the small blind calls. The flop comes 8-6-2 rainbow, and your opponent checks.

In the absence of any information about your opponent, the standard play should be to follow through with a continuation bet. Of course, once you start to gather information about your opponents, your basic strategy can be improved upon on a case-by-case basis.

Adjusting Your Play: Deviating from Basic Strategy

Now you might get weird looks from the “experts” at the table if you were to stand on 16 vs. 10 in blackjack, but it turns out that 16 vs. 10 is actually a very close play.1 Anybody who has spent much time studying card counting knows that if you are using a balanced count, that any time the count is even slightly positive (that is, if one more 3, 4, 5, or 6 has been dealt than ten-value cards using a Hi-Opt I count in a given shoe, or one more 2-6 has been played than tens and aces using the Hi-Low count), the best play is actually to stand on 16 vs. dealer 10.2 And by tracking the cards – that is, by using other information available to us – we are able to improve our playing decisions.

That said, as in blackjack, we can improve our decision making by accounting for other information in poker as well.

Let’s go back to the A-T offsuit continuation example mentioned above. Given some time at the table and some information on your opponents, you can adjust your play to each player accordingly. In The Poker Blueprint (Daily Variance Publishing, 2010), Tri Nguyen and Aaron Davis recommend following through with a c-bet against both nits and TAG players, but adjusting to both loose-passive players and good aggressive players by checking behind. The reasons you check behind against the loose-passive player (calling station) is because you have showdown value, you are not likely to get bluffed out of the pot, and because you have little fold equity. On the other hand, you check behind against the good aggressive player because you have some equity in the pot and don’t want to get check-raised.

C-betting in PLO is similar. As prescribed in my book Advanced Pot-Limit Omaha Vol. II: LAG Play (Dimat Enterprises, 2010), you want to follow through with a c-bet as often as you can get away with it without becoming a check-raise magnet. For the most part, in pots contested heads up after the flop, you are betting your strong hands (hands that can stand a check-raise like top set or a 13-card nut wrap) and air against basically all opponents. But against tricky opponents who are apt to check-raise you with some frequency, you might check behind hands like bare top pair or open-ended straight draws, which are hands you might otherwise go ahead and bet against weaker, straightforward opponents who rarely check-raise (and to a lesser extent, rarely check-call).

And so, with regard to adjusting your c-bet frequency against a given opponent, one statistic you want to key in on would be his check-raise percentage.
This brings us to the second option given limited information, which we will discuss next time: Playing the percentage directly.

1. For just how close, I refer you to Appendix E of Stanford Wong’s Professional Blackjack (Huntington Press).

2. In fact, if you are counting cards, are varying your bet sizes according to the count, and had to choose one play or another as your standard play in an effort to avoid getting flagged as a card counter, you are probably better off standing on 16 vs. 10 every single time, because standing will be the correct play when the majority of your money is being wagered. See Burning the Tables in Las Vegas (Huntington Press) by Ian Andersen.

Card Player columnist Jeff Hwang is author of Pot-Limit Omaha Poker: The Big Play Strategy and the three-volume Advanced Pot-Limit Omaha series. Jeff is a consultant for the soon to be released PokerTracker 4, public beta available now.