On a 180 BB downswing, is the following analysis correct?
 

I've been playing online for a while now, and have recently made the jump to the $10/20 game on Party. For my first 11,000 hands, I was absolutely killing the game at over a 3 BB/100 clip. I don't doubt that this was actually higher than my expected win rate, but since then, I've completely tanked and am on the biggest downswing of my career. As a good poker player, once the downswing started, I took a look at my play and considered the fact that I might not be playing well. However, I feel like I've been playing quite well, so I decided to look at the numbers to see if I could find anything. Just for reference, here are my $10/20 numbers from before my downswing:


BB/100: 3.21
Total Hands 11,311
VP$IP: 16.08
Went To Showdown: 31.45
Won $ At Showdown: 54.85
Win %: 7.93

Here are my numbers during the downswing:


BB/100: -2.10
Total Hands 8,645
VP$IP: 14.71
Went To Showdown: 31.22
Won $ At Showdown: 52.48
Win %: 6.78

By contrast, here are my numbers from when I was playing $3/6 (I only have a limited number of $5/10 hands in my DB):

BB/100: 2.35
Total Hands 35,410
VP$IP: 14.49
Went To Showdown: 30.17
Won $ At Showdown: 55.55
Win %: 7.63

For my analysis, I'm going to basically ignore the $10/20 numbers from before my downswing, since there aren't all that many hands and just compare with $3/6.

The one thing that really jumped out at me when looking at the numbers was how low my winning percentage has been during the downswing. I've condisered several things:

1) I could be playing too tight

I sort of threw this one out of the window right away when I looked at it. Both my VP$IP and WtSD numbers are actually higher than they were at $3/6, so I'm playing more hands and going to the river more, while winning less.

2) The Players around me are calling down more

This also doesn't make much sense, since the players at $10/20 are generally tighter. If I'm playing more hands here, I'd expect my actual win percentage to be higher, since I won't be getting sucked out on by terrible draws nearly as much. I also doubt I'm getting called down my more than one person regularly, which could also lower this number.

3) Bad Luck

I can't really think of any other reasons that my win % would be so low, so I've decided that it's quite likely that I'm just experiencing a terrible run of cards.


With this in mind, I assumed that all other things being equal, playing the same number of hands and going to showdown about the same ammount, I should win the same percentage of hands.

Thus, the difference between my actual winning percentage and my "expected" winning percentage (I'm just assuming here that my $3/6 win % is more or less accurate) is about 0.85%. Over 8,645 hands, this corresponds to about 73 hands which I nomrally would win that I've lost over this stretch. If we assume an average pot size of 5 BB (an underestimate since I hardly ever play in a game where the average pot <$100), this corresponds to about 365 BB that I would normally win during this time that I didn't win. Had I won these hands, I would have been +185 BB over this span rather than -180, which corresponds to about a 2.12 BB/100 win rate.

Even if I'm slightly off about my true win %, it seems like I would normally be a strong winning player had I won these extra hands. So, I have a few questions about this analysis.

<ul type="square">[*]Is there any other reason that I've neglected that could cause my win % to be so low over this span?[*]What other numbers are useful to look at in assessing your play over a downswing?[*]Is my logic in assuming that my win % should be higher flawed in any way? (I'm not concerned so much with my guess of 7.63% being accurate, but more of the general principle I'm using)[*]Based on the information presented here, would you say that I'm a strong winning player (2 BB/100 or so) at this level? I'm still averaging 0.91 BB/100 over almost 20,000 total hands now.[/list]

Re: On a 180 BB downswing, is the following analysis correct?
 

Check out the number of AAs, KKs and QQs you've been dealt in that time.

You should also consider you're just playing against better players, so they may be calling you down LESS and not giving you action.

Re: On a 180 BB downswing, is the following analysis correct?
 

Insufficient sample size.

-Brian

Re: On a 180 BB downswing, is the following analysis correct?
 

[ QUOTE ]
Check out the number of AAs, KKs and QQs you've been dealt in that time.

You should also consider you're just playing against better players, so they may be calling you down LESS and not giving you action.

[/ QUOTE ]

I don't think that getting dealt AA,KK,or QQ less is what causes downswings greater than +150BB. I think it's from whiffing on the flop over and over and over again. And making second best hands continously.

Regarding your avatar: How can anyone consider a picture of a cute lil' poker playing gal like yourself to be gay? Clarkmeister is gay. Don't listen to him.

Re: On a 180 BB downswing, is the following analysis correct?
 

[ QUOTE ]
Insufficient sample size.

-Brian

[/ QUOTE ]

That is such total BS. He has more than enough hands to analyze his play.

I get so frustrated when people constantly beat the "insufficient sample size" drum when they have nothing to say. It's garbage.

Re: On a 180 BB downswing, is the following analysis correct?
 

Matador - I'm not flaming you this time.

To show how sample size needs to be HUGE to accurately calculate win rate, Homer made this post:

"How many hands do I need...." By Homer


It's truly eye-opening.

Re: On a 180 BB downswing, is the following analysis correct?
 

With the move to a higher limit, this is probably a combination of three factors. First off, you are most certainly facing, on average, better opponents who will not pay off as much as those at 3/6 did. Secondly, you are probably just running bad. Finally, because of the downswing combined with the jump in levels, you are probably not playing your best poker.

Keep in mind that overall you are still positive at 10/20. Also note that your downswing is much shorter than your upswing (in terms of # of hands). If it's messing your play up, drop down and get your confidence back. You could drop to just 5/10 if you wanted.

Re: On a 180 BB downswing, is the following analysis correct?
 

Don't feed the trolls. [img]/images/graemlins/grin.gif[/img]

[EDIT]: Over the course of almost 20,000 hands, the guy is up $3,630, or approximately 180 big bets. Not only is 20,000 hands an extremely tiny sample size in which almost anything can happen, but also he's actually up. He needs to put some more hours in and come back if he's still having a problem at 200,000 hands.

-Brian

Re: On a 180 BB downswing, is the following analysis correct?
 

While I agree with Homer's analysis and really found it useful... I am going to be siding with The-Matador (aka various other people) on this one (excuse me while I put my flame suit on).

Too many posts are the generic "sample size is too small". For example, in this reply there is no mention of which sample he is talking about. Okay, we can infer from reading comprehension that he may be talking about the 8k hands during the downswing... but what useful information is being given by "sample size is not big enough".

I know that in all online communities the veteren and long time posters get tired of answering the same quesitons and "use the Search noob" is a pre-determined phrase on their keyboards..., and I think in many situations this reponse is sufficient or required. Specifically when the OP asks a generic "Help me find the answer to generic question XXX". However, in this case it is obvious that the OP has put some tought and work into analyzing his statistics and is looking for 3rd party insight to his thoughts. Reading between the lines a little tells any reader that it is obvious that the reader is not looking for long-term win rate judgement, but more comments on how he is analyzing his own data and ideas on other things he can do to improve his analysis.

To the OP:
I would take a look at how your big hands have held up (AA, etc.). How many times have you been called on the river during both your up/downswing. Thinking players that have seen you might be less willing to pay off on the bigger streets after having some information on your aggression factors on each street. Learning when/whom to try to induce bluffs on the bigger streets is a new skill that I think can be added in the short handed game (I am hoping at least... because i just started dabbling in the 6-max and know for a fact I have no idea how to do this correctly).

Re: On a 180 BB downswing, is the following analysis correct?
 

You could look at it as winning 1BB/100 over 20000 hands?

Or is that just unacceptable to you?

Re: On a 180 BB downswing, is the following analysis correct?
 

[ QUOTE ]
I don't think that getting dealt AA,KK,or QQ less is what causes downswings greater than +150BB.

[/ QUOTE ]

They can contribute though - at 33,000 hands I figured I *should* have had them 150 times (hope my math is right). They were actually all at less than 130. Given they win an average of 2BB for me per hand, that's 120 BB I'm 'missing' right there.

Re: On a 180 BB downswing, is the following analysis correct?
 

[ QUOTE ]
You could look at it as winning 1BB/100 over 20000 hands?

Or is that just unacceptable to you?

[/ QUOTE ]

Exactly.


Also, to the OP: post hands you're having trouble with.

Re: On a 180 BB downswing, is the following analysis correct?
 

[ QUOTE ]
I know that in all online communities the veteren and long time posters get tired of answering the same quesitons and "use the Search noob" is a pre-determined phrase on their keyboards..., and I think in many situations this reponse is sufficient or required. Specifically when the OP asks a generic "Help me find the answer to generic question XXX". However, in this case it is obvious that the OP has put some tought and work into analyzing his statistics and is looking for 3rd party insight to his thoughts. Reading between the lines a little tells any reader that it is obvious that the reader is not looking for long-term win rate judgement, but more comments on how he is analyzing his own data and ideas on other things he can do to improve his analysis.

[/ QUOTE ]

You pretty much hit the nail on the head here. I'm not looking for an actual "what is my true winrate?" here. I've done the math before, and even being the relatively tight player that I am, with small standard deviation (13 BB/100), it would take me over 6 million hands to know my true winrate +- 0.1 BB/100.

What I was really looking for was an critique on my analysis that the % of hands won is likely a source of the recent downswing. More or less, is it reasonable to assume that I will generally win more times when I play this many hands?

Obviously, it takes a good number of hands to know your true winrate, but it takes a significantly smaller sample size to determine other things, such as standard deviation and (I believe) % of hands won.

Regarding some of the questions.

I have considered the fact that I've been playing with generally better players at this level, but I thought this would be somewhat covered by the fact that I've actually gone to showdown more over this stretch (and more in general) than I did at $3/6. My thoughts on this were that if people weren't paying me off with weaker hands, my WtSD numbers would be lower as would my Won $ at SD.

As for my big pairs, they've held up fairly well, although I haven't gotten them as much as usual

AA: 86% (1 out of 233 hands)
KK: 69% (1 out of 298 hands)
QQ: 63% (1 out of 288 hands)
JJ: 43% (1 out of 210 hands) &lt;--I'm actually down slightly with JJ

My aggression factors are still on a par with what they normally are. The river is lower, usually it's around 1.6. This is probably due to less value betting since I'm against better players, the fact that I'm winning less often anyway, and possibly a little fear which has been put into my by the bad streak.

Flop: 3.11
Turn: 2.87
River: 1.25

Re: On a 180 BB downswing, is the following analysis correct?
 

Greke, it looks like you've done some good analysis, probably better than what you could hope to get by posting here.

I don't have anything to add, except that your own analysis is probably right - you're just running bad.

I liked what you did with the percentage of hands won analysis - it shows what a big impact a small number of hands can have.

Re: On a 180 BB downswing, is the following analysis correct?
 

Thanks Linus,

That's pretty much what I was looking for. Either someone (or multiple people) to confirm what seemed like a liklihood, or to point out anything I might have neglected to notice that could also be contributing.

Since I'm currently aiming to correct the problem, I thought I'd post here to see if anybody could see anything in my numbers that I had missed. Of course, if nobody sees anything, then it means that there probably is no problem, so I'll just continue on with what I've been doing and hope that things start to turn around soon.

Re: On a 180 BB downswing, is the following analysis correct?
 

From what I understand of ring play. Your sample size *is* too small to draw definite conclusions. However, that *does not* mean it is meaningless.

Also, no one has really addressed what can be expected as a win rate at 10/20. I don't really know but it seems 3BB/100 is most likely overly optimistic. At the same time I've seen some of your play and it's very solid. I doubt the bad run is due totally to tilt and I doubt the good run is due totally to skill.

All that being said. The ~+1BB/100 combined is nothign to sneeze at especially considering you sort of came straight from 3/6 (where your sample size is also too small for definite conclusions).

As for reasons for your stats. Variance is 99% of the explanation for them. If this is what you're calling 'bad luck' then so be it (although there really isn't anything such as bad or good luck).

My advice is that as long as you're not tilting or feeling that you're playing sub-optimally and can stomach swings in the 10/20 then continue to play them. If you don't think you can handle another 8,600 hands at -2BB/100 right now, then drop back down to the 3/6.

As far as plugging possible leaks -- go ahead and post hands that you felt were tricky. Also feel free to identify the most common situations where you *might* be making mistakes. Obviously those specific situations would be the 'leak' if you're not playing them very well.

If you need a support group then visit the following forum and post on it at least every day:

QLC

Yugoslav