Friday, January 27, 2006


It's 50/50 - You Win Or I Win

Just had a moderately amusing hand in a Sit and Go on Stars. Three-handed, so in the money, quite high blinds. I find AK on the button and auto-push. The small blind calls. At this point I have to make a decision on another table. By the time I switch attention back to the first one, I simultaneously see that I have been called by AA, and the flop has come KKQ. There's no further drama, I double up and go on to win the S+G.

AA might well have kicked his cat half way across the room at this one, but let's calm down. Both players had an automatic play. Both players would have done exactly the same thing if the roles were reversed. AA can claim he was unlucky if he takes his starting point as the time he looked at his cards. Had I lost the hand though, I could just as easily claim that I was unlucky to run into Aces. Someone has to win the hand and the loser will, if he wants to, feel hard done by.

This is an extreme example but you see this all the time. It's the easiest thing in the world to blame a tournament exit on bad luck. Either you ran into a bigger hand or you got outdrawn, take your pick ! We can launch into a big analysis of when the money went in and who mis-played the hand if we want to claim the higher ground but the bottom line is why bother. Try to ignore the noise as best you can and get on with it. For example, ask me how the sit and goes are coming along. Go on. Well, after 100 they were going very well ; in 35 since then I haven't won too many but I don't know exactly how much I'm still ahead because I won't even look until I've played 200.

This is a trick from Taleb ; the more frequently you check your results, the higher the noise/signal ratio and the more likely you are to see a loss, even if your expectation is a healthy plus [1]. Seeing a loss is more psychologically damaging than seeing a win is beneficial, by a factor of 2.5 Taleb quotes but who knows how they worked that out, mice and pellets probably. So anyway, I tap my results into the spreadsheet but all the $ columns have zero length. When I've played 200 I'll un-zero them and see where we are. It's a nice trick if you have the discipline to stick to it, try it out. And incidentally, I managed to do a version of the same thing in the particular hand. By looking away I avoided the trauma of "Fuck ! He's got Aces ... Woo-hoo KKQ !" and instead experienced a simple "ah yes AK on the button stands up, as one would expect". Of course, this is very difficult to do consciously when the game is right in front of you !

[1] Of course if your expectation is negative, it's reversed. A losing player who only checked at long intervals would see continuous losses and probably give up. So it's a good job they don't :-). Volatility is the bad player's friend. Which is why, now I think about it, bad players should play satellites. It's their best chance of making any real money out of poker. So that's my final contribution to the debate. Yes, you should play satellites, but only because you suck :-)

So, it's a factor of 2.5, is it? That feels about right.

I've adopted a more rigorous "play for x amount of time, only move if game is bad, tend to quit a table if behind or hitting a bad streak after being in front". In other words, focus on playing properly.

I'd like to ignore how things are adding up, but I'm always looking to see what games are good and which are bad. I can feel that a game is not good but, as it transpires, it is good (just more volatile). If I hadn't been keeping a relatively constant check on results, I wouldn't have known this.

But once I reach a point where I know I am good enough to win, I'm not planning on moving up, or changing sites, I'm going to rejig the spreadsheets. it's a bit difficult with pokertracker, though, which likes to tell you at every turn how well you are doing.


(Originally posted by PeteB, I have re-posted because the blog was playing up somehow)
It is a balancing act. I am very confident that I should be a winner in these games, I'm just not sure how much, and it would be impossible to lose a lot without noticing (I'd have to re-deposit).

It's funny though, I'm playing on three sites. On one of them I can't lose, and on another I can't win an argument, over the last few days. But I do think that I'm best placed to decide on the best game at longer intervals, just so that the noise is filtered out to some degree.

“the more frequently you check your results, the higher the noise/signal ratio and the more likely you are to see a loss, even if your expectation is a healthy plus”

I’m not so sure I agree with this.
I look at my results constantly. Certainly on a daily basis, if not more frequently. I particularly like to look at a long term graph, the shows net result on a session-by-session basis.
I find it very helpful, after a loss, to see that the over all trend of the graph, over the last 12 months is a steady climb upwards, and that the loss I have just suffered, no matter how big, is but a small blip in the graph.

Therefore, by checking often, particularly after a loss, I am reminded that a single session is meaningless, and all I have to worry about is the long term.
Get it quietly Mr Shoreman.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?