Breakeven stretches

[Psy_Mike]

Hey!

I usually don't have long consecuitive losing streaks. The biggest are like 3-4 games lost in a row so not bad. However I'm wondering what type of breakeven stretches you guys experience? Can they last for 30 games? 40? 50? Assuming you're a 62% winrate player, is it possible to calculate how long stretches you are likely to run into after say 100, 200, 500+ games etc?

Jay_Shark, do the math and someone can make a nice lil graph [img]/images/graemlins/wink.gif[/img]

[jukofyork]

[ QUOTE ]
Hey!

I usually don't have long consecuitive losing streaks. The biggest are like 3-4 games lost in a row so not bad. However I'm wondering what type of breakeven stretches you guys experience? Can they last for 30 games? 40? 50? Assuming you're a 62% winrate player, is it possible to calculate how long stretches you are likely to run into after say 100, 200, 500+ games etc?

Jay_Shark, do the math and someone can make a nice lil graph [img]/images/graemlins/wink.gif[/img]

[/ QUOTE ]
For a 62% winrate player the probability of losing N games in a row is simply: (1-0.62)^N = 0.38^N

To lose 10 games in a row should only happen 0.006% of the time, so you would have to be pretty damn unlucky for that to happen! [img]/images/graemlins/smile.gif[/img]

Juk [img]/images/graemlins/smile.gif[/img]

[i]EDIT: Also, if you want to run this as a simulation and see then output as a graph, then this post explains how to do this in excel.
EDIT2: Sorry half asleep here... I didn't read your post carefully enough, as you asked about breakeven stretches rather than losing streaks (this post might be helpful, but related to ring games).[i]

[Psy_Mike]

Thanks [img]/images/graemlins/smile.gif[/img]

But I actually don't mean consecutive losing matches, but breakeven stretches or downswings!

For example if you go 5-5, 4-6 and 6-4 over 3 batched of each 10 games you'll end up being breakeven minus the rake.

[acidca]

[ QUOTE ]
[ QUOTE ]
Hey!

I usually don't have long consecuitive losing streaks. The biggest are like 3-4 games lost in a row so not bad. However I'm wondering what type of breakeven stretches you guys experience? Can they last for 30 games? 40? 50? Assuming you're a 62% winrate player, is it possible to calculate how long stretches you are likely to run into after say 100, 200, 500+ games etc?

Jay_Shark, do the math and someone can make a nice lil graph [img]/images/graemlins/wink.gif[/img]

[/ QUOTE ]
For a 62% winrate player the probability of losing N games in a row is simply: (1-0.62)^N = 0.38^N

To lose 10 games in a row should only happen 0.006% of the time, so you would have to be pretty damn unlucky for that to happen! [img]/images/graemlins/smile.gif[/img]

Juk [img]/images/graemlins/smile.gif[/img]

EDIT: Also, if you want to run this as a simulation and see then output as a graph, then this post explains how to do this in excel.
EDIT2: Sorry half asleep here... I didn't read your post carefully enough, as you asked about breakeven stretches rather than losing streaks (this post might be helpful, but related to ring games).

[/ QUOTE ]

I don't know why everyone thinks winrate is a constant.

[omgwtfnoway]

what's the problem wiht thinking of it that way?

[Winenose]

There's a post on BBV or NVG forum where someone posted his cashgame grap of last 580 THOUSAND hands, played during last 18 months or so. His first 240 THOUSAND hands was breakeven.

Cashgames are a lot different than HUSNGs, but I guess when that guy was breakeven for SO LONG, everything is possible with HUSNGs also. I'm talking about not being up or down more than 20 buy-ins over couple thousand HUSNGs.

[frede89]

[ QUOTE ]
what's the problem wiht thinking of it that way?

[/ QUOTE ]

I guess it is because you winrate depends so much on your opponent. So you winrate will not be the same against 2 different opponents.

[Indiana]

long breakeven stretches are usually due to changes in playing style or life events that cause you to play differently. Thus, its very hard to quantify mathematically the chances of it happening. In a vacuum for a winning player, yes, its rare to have a stretch of 30 games like this.

Indy

[jay_shark]

Good question .

Ok suppose your true win rate is 62 % .

I will compute the probability you break even in your first 20 games .

There are 20C10 different ways to win 10 games from a string of 20 .

P= success of 0.62
q=failure of 0.38

P(w=10) = 20c10*0.62^10*0.38^10 = 0.097 or 9.7 %

Lets figure this out for 50 games .

50C25*0.62^25*0.38^25=0.025 or 2.5 % .

[Psy_Mike]

Ah very nice jay, thanks [img]/images/graemlins/smile.gif[/img]

How about the probability of a X long breakeven stretch happening any time in Y games?

So for example, a 50 game break even stretch any time in the period of 200 played games.

Thanks again!