Poker & Psychology (long)

    • GrimsDad
      Joined: 10.04.2010 Posts: 1
      Preamble: I've been playing online poker for about 10 years now and twice as long live. The following is an article I wrote for another website that might be helpful for beginning -> intermediate players (hence why I'm posting it in this forum) and deals primarily with the psychological aspect of the game. Hopefully it turns out to be a useful read to some of you.

      Confirmation Bias & Sample Sizes
      ... or why your pet theories are wrong.

      So, I keep hearing variations of the same story over and over again:

      • My account is rigged on poker site X because I win more on site Y!
      • Bigstacks always win flips against shortstacks in tournaments!
      • Ax always beats my KK/QQ / I always lose with AA! / Donkeys always make their draws on the turn/river!
      • No way that would have happened live!!

      The list could go on, and I'm sure you've heard similar variants yourself if you've been around the online poker world for a while. This is where the related concepts of confirmation bias and sample sizes enter the picture.

      First, let's deal with confirmation bias:

      Confirmation bias (or myside bias[1]) is a tendency for people to confirm their preconceptions or hypotheses, independently of whether or not they are true. People can reinforce their existing attitudes by selectively collecting new evidence, by interpreting evidence in a biased way or by selectively recalling information from memory.[2] Some psychologists use "confirmation bias" for any of these three cognitive biases, while others restrict the term to selective collection of evidence, using assimilation bias for biased interpretation.[3]
      (Wikipedia quote)

      Just a bunch of fancy pants words you say? Well, let's try to relate the above concept to how we perceive things happening when we play poker. Let's say you've been on a bad run the last few days, and some of the theories I listed start sneaking into your head. Why can't I win any flips? Why are the donkeys always rewarded when they make boneheaded plays? At this point, it's likely you're affected by confirmation bias; every time you see anyone busting out at a table in all-ins with a better hand, your mind will start to take special note, and worse, your own game might start to be affected due to a subconscious fear of negative outcomes. This typically has two pretty specific consequences:

      • You start playing "scared poker", which very rarely leads to an improvement of ones game. You start narrowing down your preflop range and become more passive post-flop due to fear of badbeats (going all-in with deep stacks on the flop with a good hand to chase away any takers should also be considered a passive/scared play in this context.)
      • You are more susceptible to busting out as a favorite in tournaments, simply because playing a narrower range will lead you to being the favorite preflop more often. This, in turn, helps reinforce the confirmation bias you've already exposed yourself to.

      Combine the above two factors with increasing frustration and it starts to get harder to maintain concentration and application of sound poker principles, and you might find yourself on a downward spiral and your bankroll dwindling.

      The phenomena of confirmation bias might affect you negatively if you've had a short term hot run as well, as you might start to expect that it's "normal" to always win hands where you're the favorite or get saved on the river if you're behind.

      Here is a fun little experiment to perform if you're still a believer in any of the above theories: Pick a random tournament and follow a table for a while. Use a program like PokerStove and simply track the outcomes of preflop allins and see whether you note any statistical anomalies (ie. bigstacks winning against smallstacks significantly more often than they should.)

      There's a chance you will find anomalies just by following a single table in a tournament, but more often than not you'll find that results are about as predictable as they can be in a game where it's impossible to be more than an 80% favorite preflop.

      If you do happen to find anomalies, we have to look at the other important concept to keep in mind, namely sample size:

      Being on a bad run for a couple of days can feel tough and demotivating (and maybe taking a break at such a point isn't such a bad idea if you feel your game becomes affected), but it's always important to realize that there is little statistical significance to a sample size consisting of hands played in 2 days (not even if you're a 10-tabling maniac!) Generally speaking, you need to log around 100,000 hands to gain a significant sample size for statistical analysis, and even 100,000 hands is on the very low end and susceptible to deviations.

      That's why it's always a good idea to use poker software like Pokertracker, Pokeroffice or Holdem Manager to log all of your hands. You will note that over time as your database increases and you log more hands, that statistical deviations will become less and less pronounced, and there's a few additional bonuses too: 1) By methodically logging all your hands you might even find leaks in your game by looking at the data. Perhaps you were unlucky for those two days of running bad, but perhaps you overplayed cards in the range of 77-TT? Tracking software should give you the answer provided a sufficient sample size. 2) You'll find plenty of fish to hunt down on the tables!

      Remember, winning in poker, is a long term endeavor...
  • 6 replies