Variance

[Abbaddabba]

When i hear about these things, the first thing i think is that someone just isnt as big a winner as they think they are.

Not necessarily that they're losing.

Low winrates are going to produce large downswings pretty frequently.

Just because you've got 100k hands worth of data as a 2BB/100 winner doesnt mean you are. And if for the next 100K you break even, the new weighted average is 1BB/100. In which case, the large-ish downswing isnt so out of the question. But people have the tendency to give more consideration to the database that puts them in a positive light.

I think that everyone is guilty of these sorts of things on some level. Even if you're trying to be honest with yourself, it can affect your behavior. An example ive found is that if i play on my laptop and have a great session, i am excited to import hands to my PC. If i have a poor session, i usually put it off and im sure at least a couple of times i have forgotten (not consciously). I seldom play on my laptop so it doesnt add up to anything significant, but things like these add up. Another exammple is choosing to start a database when you've had a favorable stretch, but saying '[censored] this' if you hit a rough stretch. When i asked if my friend if he wanted my excel poker log spreadsheet, he said "i cant win lately.. why would i want to track my results?" Which makes sense. You are more likely to track these things in good times. Which biases the results that you have. SO when he does start, the beginning is almost for sure going to consist of a high point, which means it'll be ever so slightly biased.



The other most likely culprite is SD. A lot of you have thrown out numbers like 20 or 30 SD/100. Does that actually exist? My SD is historically less than 15 in small stakes, and 16 - 17 in short handed mid stakes. I cant imagine why the SD would be so dramatically higher in high stakes games. The AP mid stakes games are usually pretty aggressive as it is... i cant see them getting too much crazier.

[ggbman]

Yes, standard deviations can be much much higher as the stakes increase, especially since games often run 2-3 handed instead of 6-10 handed.

[Trix]

Think mine is 200ish...bahahahaha [img]/images/graemlins/tongue.gif[/img]
Dont think I played much more than 150-200K hands of limit though..

[The DaveR]

Why doesn't someone just reverse the problem and estimate, given say 200 players with winrates of 1 BB/100 the likelihood of 1 or more having a 450 BB downswing. I suspect the probability is uncomfortably large.

[NLSoldier]

[ QUOTE ]
Why doesn't someone just reverse the problem and estimate, given say 200 players with winrates of 1 BB/100 the likelihood of 1 or more having a 450 BB downswing. I suspect the probability is uncomfortably large.

[/ QUOTE ]

pretty sure this is pretty much what GoT did a while back, i think he did it for various winrates though, and yes the numbers were scary as hell.

Im sure someone could dig up a link.

[Victor]

[ QUOTE ]
[ QUOTE ]
Why doesn't someone just reverse the problem and estimate, given say 200 players with winrates of 1 BB/100 the likelihood of 1 or more having a 450 BB downswing. I suspect the probability is uncomfortably large.

[/ QUOTE ]

pretty sure this is pretty much what GoT did a while back, i think he did it for various winrates though, and yes the numbers were scary as hell.

Im sure someone could dig up a link.

[/ QUOTE ]

joshthief did it too with similiarly scary results.

[sweetjazz]

[ QUOTE ]
Why doesn't someone just reverse the problem and estimate, given say 200 players with winrates of 1 BB/100 the likelihood of 1 or more having a 450 BB downswing. I suspect the probability is uncomfortably large.

[/ QUOTE ]

You also need to specify how many hands each player plays. If they each play 10K hands, it's probably not a certainty (though still quite likely). If each plays 100K hands, then it is just about an absolute certainty.

[bigbootch]

[ QUOTE ]
this has been an excellent discussion and one that everyone should read carefully if for no other reason than to learn how to properly manage their bankroll. just to throw out a hypothetical question -- what would anyone consider to be the largest downswing someone could have and still be within the consideration of being a "good" player. (now i know that it depends on what level you play and what people consider "good" to be. let's say for arguements sake that the levels played at are in the 10/20 to 50/100 range and that a good player is one capable of being expected to have a true winrate of 0.75bb/100+.) is a 1500 bb downswing as large as imaginable? how about 2000 bb? 2500 bb? are we in the suicidal range yet?

[/ QUOTE ]

The answer to your question completely depends on what confidence level you personally would need in order to say "I'm willing to assume X." 95%? 99%? 99.8%? (Obviously, there doesn't exist an X such that you can ever be 100% sure.)

I.e. your question is like asking "Let's say we flip a coin and it keeps coming up heads, after how many times would you consider it a weighted coin?" Somebody who is willing to accept a lower confidence rate might say 5. Someone else who requires a much (much much) higher confidence rate might say 5000.

-bigbootch

[Flintoff]

[ QUOTE ]
[ QUOTE ]
My biggest is 300BB (though only at mid-stakes) but I am certain bigger can happen. After my 93BB in 140 hands yesterday in which I played great, I'm open to anything!

If you can have a 300BB ds, then of course you CAN have 400, 500 etc. I remember Stox had over 400BB recently. Just hope this evil never points it's finger in your direction!

[/ QUOTE ]

I think you are confusing me with someone else.

[/ QUOTE ]

Sorry bud.

I was sure I'd read that from somewhere reliable. Anyway, I'm glad for you that it's not true!! [img]/images/graemlins/smile.gif[/img]

[aba20]

I think people use variance to rationalize why they are only small winners or losing. I always take the mentality that my results are based on my play. People act like there is some outside force beyond their control. Well there isn't your results are based off of your actions and nothing else. People need to take responsibility for there losses and realize they are to blame for losing and not some statistical anomaly. When I am losing or on a downswing I don't blame variance or bad luck I look to myself, because I know that I am probably off my game. I take responsibility for the losses and look to improve my game. I know this is kind of a rant but my one point that I think can help 95% of players on this board is this; Take responsibility for your play and results. Both are a direct cause from your actions.