Articles Poker Forums

cowpig : Downswings and the Information Game

published almost 5 years ago

About a year ago, during a crushing downswing, I was reading a high-stakes 2+2 thread in which Jman28 was arguing about stop-losses. He said that he didn't set a stop-loss, because he could tell if he was becoming tilted, and could maintain his A game even if he was down several buyins vs the same person. Many other successful high-stakes players argued that he should set a stop-loss regardless.

In this particular thread, Taylor Caby posted something which piqued my curiosity. He said, "I've always quit playing HU when I lost a couple buyins and I can't emphasize how much that has helped me. Simply put, it's just really hard to play anyone, much less a WCP, when he has the momentum in the match."

Taylor seemed to believe something that he couldn't quite express, and I became fixated on figuring out what, exactly, this metagame edge was. I too had intuitively felt that losing to a certain player put me at some kind of disadvantage, regardless of how much it rattled my spirit. But his “momentum” concept seemed vague and incomplete.

Upon a deep and lengthy reflection, it finally occurred to me that a heads-up player who is up a few buyins typically has a real tactical advantage that goes beyond the realm of poker psychology: he has an informational advantage.

Suppose I am playing someone heads-up, and I cannot hit a hand for the life of me. Every time I am involved in a pot, I miss my draws, hit TPGK when my opponent hits TPTK, etc. We are a few hundred hands into the match, and my opponent is up a few buyins against me, mostly from running hot.

In this scenario, my opponent has the tremendous advantage of knowing the lower end of my range in certain spots, what kind of plays I'm willing to make, the types of boards I won't play back at, etc. Each time I call down and fold to a small river bet, he can infer that I played a draw that way and adjust accordingly. Each time I fold to a turn check-raise, my opponent has gained information about my hand in that situation (that it is not strong enough to withstand a check-raise, that I am capable of bet/folding in those types of situations, etc), while I have gained very little information about his hand.

Another example of this informational edge is where I make a thin call on the river, and my opponent happens to have had the nuts. In this situation, I've gained very little information about his play (as he will of course be betting with strong hands), while he has gained a lot of information about my range in a certain spot, and what kinds of calldowns I am capable of making.

This informational dynamic can be most easily observed in a heads-up match. However, it applies to all aspects of poker, including the dreaded of all: swings.

If I run hot over x number of hands, I gain all kinds of information about players' tendencies. For example, if I flop a big hand vs someone and take a line which forces them to fold, I may gain some insight into future bluffing opportunities. If I happen to take a creative line which causes someone to stack off extremely thinly, I've added a powerful line into my arsenal which I will be able to use in the future vs the same type of player.

Running bad has the opposite effect. If I have a powerful +EV bluff which just happens to run into the nuts 5 times in a row, I will likely lose that line from my arsenal, and my winrate will consequently deteriorate. If a couple opponents river sets in spots where I successfully got two streets of thin value with middle pair vs their underpair and I end up having to fold, I'll probably think I incorrectly bloated pots with marginal holdings and miss out on that value in the future.

I often hear players say, "I hate that site. It's completely rigged. I can never win there." Perhaps those players simply ran bad at first on that site, and consequently all their regular villains lucksacked their way into exploitive information. Maybe they really can't win at that site, and it has nothing to do with the site itself!


or track by Email or RSS


clowntable

Avatar for clowntable

289 posts
Joined 02/2008

This observation as so many great ones seems really trivial and "doh that makes sense" once you read it but is probably something I wouldn't have thought of on my own (at least not explicitly, maybe blink style implicitly).
Very nice article.

Posted almost 5 years ago

jk3a

Avatar for jk3a

898 posts
Joined 01/2008

losfer

Avatar for losfer

57 posts
Joined 08/2008

Do you then never play that person again? If you do play them again, what makes it so that they lose this informational advantage or what makes it no longer important?

Posted almost 5 years ago

dkir

Avatar for dkir

3 posts
Joined 02/2008

aislephive

Avatar for aislephive

23 posts
Joined 07/2008

I don't play HU but this is really interesting and I agree completely with what you wrote. Nice job.

Posted almost 5 years ago

durkadurka33

Avatar for durkadurka33

32 posts
Joined 02/2008

Chaostracize

Avatar for Chaostracize

62 posts
Joined 01/2008

Profane

Avatar for Profane

18 posts
Joined 05/2008

Hielko

Avatar for Hielko

4352 posts
Joined 07/2008

Nice article, but I don't believe in your final conclusion that some players can't win at a certain site because the regs lucksacked their way into exploitive information. If those regs would be capable of collection and using that information - assuming that the value of it is indeed so high to make a winning player a loser - they also should be able to gather this information during normal play, so that would mean that those players should slowly turn in losers, even when they start at a certain site while running hot.

Posted almost 5 years ago

durkadurka33

Avatar for durkadurka33

32 posts
Joined 02/2008

I don't think that that's what he's concluding. He's not saying that some people will never, in the medium/long run, be able to win on a given site because some people lucksacked him early. No, he's saying that some people, who play a few thousand hands (or less), get lucksacked on, and have this information effect against them in the short term, often are the ones claiming that a site is rigged and that they can't win on it.

Posted over 4 years ago

imnuts4u

Avatar for imnuts4u

149 posts
Joined 01/2008

Do you then never play that person again? If you do play them again, what makes it so that they lose this informational advantage or what makes it no longer important?



this question raises a reasonable counter-point that deserves a rebuttal

Posted over 4 years ago

cowpig

Avatar for cowpig

14 posts
Joined 01/2008

Do you then never play that person again? If you do play them again, what makes it so that they lose this informational advantage or what makes it no longer important?



In practice, I doubt anyone is going to remember much specific information about a player they played a week ago. A month later, I doubt anyone will remember much of anything.

During a particular heads-up match, you can gain information as subtle as a general impression of the rate at which you can expect to adjust, for example. In the long run, a bunch of these types of impressions form a greater concept which creates an overall/default strategy.

In a particular match, quitting when down a few buyins is a good idea. Longer-term, switching sites or moving down in limits after a prolonged period of running bad is a good idea.

Nice article, but I don't believe in your final conclusion that some players can't win at a certain site because the regs lucksacked their way into exploitive information. If those regs would be capable of collection and using that information - assuming that the value of it is indeed so high to make a winning player a loser - they also should be able to gather this information during normal play, so that would mean that those players should slowly turn in losers, even when they start at a certain site while running hot.



I didn't mean this quite as literally as you seem to have interpreted it. I meant more that this informational dynamic can be used to explain dynamics that occur on a wider scale. Suppose, for example, that you are BldSwtTrs, and are one of the pioneers of the preflop 3-bet at party 10/20 games. People don't realize how wide your range is, and consequently play with a grossly skewed concept of your hand range in bloated pots, giving you massive amount of +EV bluffing opportunities.

But a particular regular runs super hot against you, and notices that you're folding when he shoves the turn against you way too often. Or calls your shoves and sees you showing up with hands he didn't think were in your range. Etc. This player begins to adjust his strategy against you, and maybe considers trying out this 3-betting thing himself. Before long, he's discussed it with friends, who've discussed it with their friends and the whole game dynamics at Party 10/20 have changed, and not in your favor.

Obviously that's a very extreme example, but my point is that regulars don't have to be specifically and consciously adjusting to your tendencies for running bad to hurt your winrate. Just the compilation of several subtle changes in peoples' general approaches could be enough to do significant damage, in theory.

Posted over 4 years ago

robvleugel

Avatar for robvleugel

2 posts
Joined 08/2008

If a novice player first plays at a particular pokersite and loses like 5 BI's in a row, he might start to suspect the pokersite to be rigged. As soon as this has happened, this logic will be burned inside his head and he will only look for information that supports this idea. He wouldn't notice if he hits a flushdraw a little bit more often than he should be. He only looks at the bad beats, which are obviously caused by the doomswitch.

The same will happen if you start to belief things like above. Sure there can be a slight informational edge for the winning player, but this edge is so small and is nothing compared to the variance caused by the cards you are getting. It's realy not like you're going to notice it right away! However, as soon as this idea is inside your head, you might start to belief you are realy a big underdog because your opponent is on some winning streak. This could eventualy lead to a decrease in self confidence, causing you to get off your A-game which DOES decrease your overal result. It's just not +EV to worry about these things.

Posted over 4 years ago

czzarr

Avatar for czzarr

243 posts
Joined 02/2008

awesome article, this is actually going to help my HU game a lot imo

Posted over 4 years ago




HomePoker ForumsArticles → cowpig : Downswings and the Information Game