300BB downswings are not normal. PERIOD!

[Jeff W]

[ QUOTE ]
I think this guy wins.

[/ QUOTE ]

I've wondered a lot about whether shorthanded games(and limit holdem in general) follow a normal distribution curve. The way to settle this is get a bunch of people(how many would be enough?) with large databases to post graphs of their results and analyze the resultant swings to determine the distribution they follow.

For now, I remain undecided.

[DeeJ]

[ QUOTE ]
I think this guy wins.

[/ QUOTE ]

I thought the guy with the graphs had won, but you may be right.

[ISF]

It is impossible that they could be normal and iid. The only question is how much the autocorrelation effects the probability of any size down swing. A simple test to figure out if these are autocorrelated is a variance ratio test.
If you variance over 100k hands is > then 10 time the average variance over 10k hands they are positively autocorrelated this would result in larger swings then mason's formula or these graphs would suggest that just use winrate and sd.
Like I said for this to not be the case you would have to be a robot playing against other robots whom all play the same way every time.

[peachy]

i TOTALLY disagree...those downswings can so happen without tilting or bad play...ive seen it happen to good players who were doing neither...ive even had it happen to myself

[steveyz]

I think there's two definitions of "normal" here. First, at given time, it is not "normal" to be in the midst of a 300BB downswing for a solid winning player. However, after a good solid player plays > X number of hands (X depending on his skill and game conditions), there will come a time where his mathematical likelihood of experiencing a 300BB downswing is > 50% (replace 50% with whatever % you would start regarding as normal). For example, if you've played a million hands in very aggressive online mid/high limit games, chances are probably pretty good that you have experienced a 300BB+ downswing. Of course, this depends on your true winrate and how easily you tilt, etc.

Unless you are a machine, you will probably not be playing optimally when you are in the midst of a large downswing. That increases the chance that you will experience a 300BB swing (usually increase both the length and severity of all significant downswings). But at somepoint, even if you are a machine, as long as you keep playing, your chances of experiencing such a downswing will grow > 50%. This actually translates to 400BB or 500BB or greater downswings. However, at that point, to expect a > 50% chance of experiencing such a swing would require so many hands that it would be basically be humanly impossible.

So a valid point would be that most posters here have not played the > X number of hands such that their odds of experiencing a 300BB downswing is > 50% (if they always played consistently, no matter how well they were running). In that sense, a 300BB downswing could be considered NOT 'normal'.

[bernie]

[ QUOTE ]
Just a year ago a 100BB downswing was considered a sure sign of bad play here and posters advised people to examine there game. Now they are just told that it is to be expected...


[/ QUOTE ]

I think many posters also advised that one should examine their game even if they are on an upswing. Giving just 'running good' or 'running bad' the last resort. Basically, keeping tabs on one's game, as best as you can the whole time.

I think anyone who says first off that it's probably just running bad is missing how to analyze his play to a degree. Basically, they don't want to analyze it. It should be the last thing crossed off the list.

100BB is not a sure sign of bad play. But it also shouldn't be the 'trigger' for causing one to check their play.

b