Why 10,000 hands? Well....

[CrisBrown]

Hi All,

I don't usually post in this forum, but this topic isn't unique to any one game. I've noticed a common phenomenon, both in my own results and in comments I read from others. It goes something like this:

You're a reasonably competent player. That is, you have some idea of starting hand requirements, drawing odds, pot odds, and the like, and while you're hardly an expert, you at least have a good sense of what you're doing. You pick a new game, or a new stakes, and begin playing. And you win ... a lot ... a WHOLE lot. You may wonder why everyone else thinks this game is so difficult, because it's so damn easy!

And then ... suddenly ... you lose. A lot. A whole lot. Maybe half of what you've won. Maybe all of it. You now wonder if you're a complete idiot for ever having thought you knew how to play. At first it seems like the cards, but obviously as time goes on you realize that can't be all of it. You see other people win on marginal hands, and you see your opponents play marginal hands against each other, but when you get into a pot, they either fold and you win nothing, or they take you to the river with the nuts or something very close to it. You've gone from winning big pots and losing small ones to winning only the small pots and losing all of the big ones. Uggh....

I think this is a pretty common experience for many players, and I've thought about it at some length over the past few weeks. I think it's also the reason people say you need at least 10,000 hands of experience before you can begin to infer anything from your results. And I think I know why it happens.

Back in those early days, when you first decided to try this game (or these stakes), none of your opponents knew you. And many poker players share a common weakness: they assume any unknown player is a fish. So your opponents would call your opening raises with weak hands, or bet into you with marginal holdings, thinking you were playing junk like all the other fish. But since you are competent, you are going to win ... a lot ... at first....

Within a few sessions, however, the brighter opponents have notes on you. They know what starting hands you generally like to play, and have some idea of how you play them. But more to the point, they recognize that you're a competent player, and make a note that they're going to need a real hand to tangle with you in a big pot. So they stop calling your open raises with trash, and they stop betting marginal hands at you. They also learn your common bluffs, and pick them off. And you start to lose ... a lot....

It's not that your luck has changed, or that the cards have gone against you. It's not even that you're playing all that badly. It's simply that competent opponents have learned to respect your game, and they're not throwing money at you anymore. The golden age is over. Now you'll have to work for your money.

Of course, if you're reasonably competent, you've also been building data on your opponents. You know what hands they like to play, and have some idea of how they play them. You know who the consistent winners are, and you try not to get into a tangle with them unless you have the goods. You also start to study your own tendencies, and look for ways to mix up your game, to be less predictable, even for those opponents whom you see regularly. The more you do these things -- study your opponents, dodge the sharks, and mix up your play -- the more you see your bankroll start to build again.

But not like at first. The good news is, while you're not winning as much as you were before, what you're seeing now is perhaps more stable and predictable.

And by now, you've probably played about 10,000 hands at that game.

Cris

[LetsRock]

Your theory ("your opponents now know you") holds water to a point.

I don't think most players pay that much attention at most sites. Yes, if you play with the same crowd a lot of the time, it's more likely that the better players will show some respect, but often we play against different opponents on every session.

I find that my swings are 90% based on the cards. If I get a "high" proportion of good starters that don't get sucked out, I have a good session; if I get "my share" of decent cards, I usually book a slight to medium win; if I get nothing to work with and sucked out on the few decent hands I get, I (not surprisingly) usually book a loss.

Yes there are days when I get the "win small pots/lose big pots" syndrome, but I really think it has more to do with the fact that my opponents just couldn't call when I land a decent hand (sometimes when you hit a good hand, the deck is crippled - I had top set on a rainbow board and got no action; it's not likely that anyone has much that can compete with that).

The one place that I've noticed that I'm playing with the same opponents with any regularity is at True Poker. Of the 40 or so players that I see with regularity in my level, only 3 or 4 seem to have enough sense to be keeping notes on me, and they're not the ones who I'm making my profits from anyway, so I don't put too much weight on attributing tough sessions to being "known". I wish that was the case, because that's a very easy problem to fix. If someone is keeping a book on me, they'd have me in the rock corner (I'm a little more aggressive than a rock, but I would guess that my image is rock-like.) If I sense that the table is reacting to tht "image", it's real easy to run a couple of bluffs through (once every couple orbits) until I get busted: image adjusted and I get action again for a while.

[BigBaitsim (milo)]

I don't know how much people actually attend to your style of play. I use PT and export notes on anyone I've played more than 25 hands with. Even with 35K hands in my current database I rarely sit down with more than 2 or 3 people I have notes on. I play under the same name I post, so some 2+2ers probably recognize me and don't call me down with junk, but I don't think this has much of an impact.

[lostinthought]

nice post.

I think you may be more right at higher limits or at smaller sites, but really - there are 30000 players at party poker at prime time. Now I know a bulk may be play money players, but this is too large of a pool that 10,000 hands puts you on the common blacklist. On top of that, my hunch is that alot of people aren't taking notes.

Furthermore - In 10,000 hands you can still statistically have a good or bad run. Nothing to do with notes..


hey, despite this, isn't there always a party at partypoker?

[DcifrThs]

I think you have a point but it really only applies to x% of the poker population. how big "x" is really an interesting question because as x increases, your theories impact also increases, possibly proportionally, possibly exponentially...but whor eally knows? if 1-x (those players who are the partiers, have no notes, don't even KNOW they have a "playing style") =2or 3x, meaning that x=.25 or .33 then your theory may hold more water than it initially seems.

but i think 1-x is MUCH> x. so much so that your theory almost does not apply at the tables to which you refer.

for instance. is it at all possible that those draws you play late come in alot at first and then later you end up making expensive investments and a run of "cold cards" causes those draws to cost much more than they make in the short run? further, in the back of your head, is it possible you start equating flopping a draw to it completing (as it did at first a lot due to fluctuations) and start putting MORE money in at first as a dog than you normally would? wouldn't that decrease your expectation if you're not doing it corectly?

other things like this occur to the beginner and have occured with me. there was a brief period of time where i won a LOT of money and then lost a LOT of money. thankfully i realized what had happened in my mind before i lost enough to cripple my bankroll and built it back up, and up and up, as did most of you, but that initial learning curve most likely proved very expensive in terms of profits (from the first run of luck) that you lost back due to less good, or even POOR play. i like the term, "less good" play btw ;-)

also, 10000 hands doesn't necessarily NEED to be significant in terms of the measurements you're attempting to make with the degree of confidence you're attmeting to make them. at 70 hands/hour thats only 142.xyz hours. not very many hours indeed. i saved my relatively firm conclusions about being a winning player until i had about 500 hours, or according to you, 35,000 hands.

its hard to place conclusions on data that vary like this at all, so the # of hands also assumes you play the same in every hand you play...which as we all know isn't true. so similar to chaos analysis, it may be necessary to DISCARD the first however many data points you need to in order to arrive at the true trajectory of your system, here being your win rate with respect to time in a pseudo phase space of W(your win rate) against W-1 and W-2, or +1 and +2.

in any case, you see its hard analyzing this stuff because of what statistics entails.

but your post has a good point although i think it relates more to different types of games...

take care
-Barron

[cornell2005]

solution: change your screen name if possible on your server