|
|
#22
04-19-2007, 11:45 PM
|
|||
|
|||
Re: Heads-up tourneys and variance
Thats harsh. Its true that the amount of variance will always be the same but what they mean (I think) is that the swings will be much greater the closer you get to being a break even player and you will need a much larger bankroll.
|
|
#23
04-20-2007, 12:12 AM
|
|||
|
|||
|
Re: Heads-up tourneys and variance
[ QUOTE ]
Thats harsh. Its true that the amount of variance will always be the same but what they mean (I think) is that the swings will be much greater the closer you get to being a break even player and you will need a much larger bankroll. [/ QUOTE ] That is absolutely 100% true. There's a *huge* difference even in simulator runs between just a few percentage points in winrate. 58% is enough to play far fewer buyins, and 60% is enough that there's very little risk of going broke even playing as little as 5 buyins over the long run, but at 55%, playing even 20 buyins gives a very sizable chance (like 25%-ish) of failure over 2000 games. A player who's just slightly above break-even can't possibly be profitable over the long run no matter how conservatively they play with their bankroll. 55% is really the minimum. Anything less than that, and you're at least as likely to go broke as you are to end up with anything at all, and what you do end up with is pretty small on average. For example, feeding in a 54% winrate and a starting bankroll of $50 gives you basically even odds of surviving 2000 games, no matter how conservatively you play, even if you don't step up until you have 50 buyins for the next level. So yes, your winrate is absolutely the #1 key in how big the swings will be.
|
|
#25
04-20-2007, 04:43 AM
|
|||
|
|||
|
Re: Heads-up tourneys and variance
Well, yeah. And the chances of not improving over 2000 games has to be pretty damn close to zero, I would think. [img]/images/graemlins/smile.gif[/img]
But like I said, it's all just statistics, and numbers can only take you so far.
|
|
#29
04-20-2007, 02:40 PM
|
|||
|
|||
|
Re: Heads-up tourneys and variance
[ QUOTE ]
[ QUOTE ] Thats harsh. Its true that the amount of variance will always be the same but what they mean (I think) is that the swings will be much greater the closer you get to being a break even player and you will need a much larger bankroll. [/ QUOTE ] That is absolutely 100% true. There's a *huge* difference even in simulator runs between just a few percentage points in winrate. 58% is enough to play far fewer buyins, and 60% is enough that there's very little risk of going broke even playing as little as 5 buyins over the long run, but at 55%, playing even 20 buyins gives a very sizable chance (like 25%-ish) of failure over 2000 games. A player who's just slightly above break-even can't possibly be profitable over the long run no matter how conservatively they play with their bankroll. 55% is really the minimum. Anything less than that, and you're at least as likely to go broke as you are to end up with anything at all, and what you do end up with is pretty small on average. For example, feeding in a 54% winrate and a starting bankroll of $50 gives you basically even odds of surviving 2000 games, no matter how conservatively you play, even if you don't step up until you have 50 buyins for the next level. So yes, your winrate is absolutely the #1 key in how big the swings will be. [/ QUOTE ] Absolutely true, I would never disagree with any of that, but I think you are interpreting the original article too charitably. This sentence: [ QUOTE ] a really solid heads up player with a 70% win rate can get away with using numbers about twice the size of the regular SNG numbers in the chart above [/ QUOTE ] obviously indicates he thinks you need a LARGER bankroll for HUSNGs than STSNGs, which is simply ridiculous. A solid winning player of equivalent skill at both needs probably (just off the top of my head) about BR/3 for the former vs. the latter. HUSNGs are the lowest-variance form of poker, bar none. Your comments about winrate and variance apply more or less equally to HUSNGs, STSNGs, and cash games. Winrate is always the biggest factor in variance.
|
|
#30
04-20-2007, 02:51 PM
|
|||
|
|||
|
Re: Heads-up tourneys and variance
By the way, this
[ QUOTE ] A winrate of 58% gives good odds of survival even playing only 6 buyins, across 2000 games, and the equity penalty from there get pretty stiff, with minor increases in survivability paid for by large decreases in final bankroll. [/ QUOTE ] is why I think people who play the nosebleed HUSNGs are probably all making a mistake. Other than finding the occasional deep-sea fish, I can't imagine anyone's winrate at the $2k+ SNGs is sufficient to justify playing them. Personally, from stat analysis, I can see that almost all of my long-run profit comes from the weakest 20-30% of players at the stakes I play ($55 and $110 turbos). I rank players 1-10 on generic skill once I have played them at least 3 times. Of course, there may be some results-oriented bias in there; I'm probably more likely to give high rankings to people that beat me. I don't know a better way to do it though.
|
 |
|
|
Linear Mode
