Roy Cooke article in Card Player is wrong.

[Snoogins47]

[ QUOTE ]
[ QUOTE ]

150BB is a bit low for online play. You'll have one worse than that at some point if you keep playing.

[/ QUOTE ]

I will be very surprised if this happens. Alot of posts are made on variance and winrate, and the sample size needed to assess them. I think this is less of an issue at smaller stakes, and that most of these posters are mid stake players or higher. Dead money mitigates long term variance and thusly mitigates the need for bigger sample sizes to assess variance, winrate, ect...

[/ QUOTE ]

I think people tend to have similar mindsets until they have said downswing. I do realize that a hugely higher winrate, and a significantly smaller SD has to make the chances of a downswing of that magnitude significantly less likely. Not having a bunch of sims or anything makes my post here more speculation and guessing than anything else, but I have to imagine that -150BB is not a startlingly rare downswing for almost anybody across the board. Of course, much of this becomes arguing semantics, which I'm painfully fond of.

Now, if you're referring to 'going broke' after starting with a figure of 150BBs, this is a different story entirely. I would, though, like to try to hunt down simulations done with stats more in line of somebody who is crushing a micro-limit game, as opposed to the ones I've seen of somebody beating an aggressive online game for ~1 or ~1.5bb/100, just out of curiosity more than anything.

I also can't imagine that a game being weaker would 'actually' make a player's winrate converge in a shorter period of time, unless of course a weaker game significantly lowers your SD (And, as the last time I took a math class was 4 years ago, and really know very little in this realm... I am assuming a lower SD would mean that winrate would tend to converge in smaller samples... is this true? hehe)

[stigmata]

I meant that "presence of multi-tabling TAGs" affects the difficulty of the game (and hence need more bankroll), but multi-tabling itself should not have any direct effect on bankroll (although it will have an indirect effect due to decreased overall winrate).

[Mike Haven]

I understand what you are saying.

I was trying to point out that, (imo), you do need more bankroll the more tables you play simultaneously.

To give a more exaggerated, (if somewhat unlikely), example, if you were playing 12 tables at once and had only a total 300bbbr, each table could take a hit of "only" 25bb's and your br would be wiped out.

It's an unlikely occurrence, (what bad beat is not?), and unique to multi-tabling internet poker, but certainly possible within any "unlucky" hour or two.

[stigmata]

I was allways under the understanding that with multi-tabling you just hit the long run faster. E.g. everything else being equal, the chance of a 300BB downswing at 1-table is exactly the same chance as simulataneous 25BB downsings at 12 tables.

[pzhon]

[ QUOTE ]

I was trying to point out that, (imo), you do need more bankroll the more tables you play simultaneously.

[/ QUOTE ]
You are wrong. Your bankroll requirements do not increase just because you play multiple tables. It may be counterintuitive to you, but this is not just an opinion. I'm telling you this as a mathematician.

[ QUOTE ]
To give a more exaggerated, (if somewhat unlikely), example, if you were playing 12 tables at once and had only a total 300bbbr, each table could take a hit of "only" 25bb's and your br would be wiped out.

It's an unlikely occurrence, (what bad beat is not?), and unique to multi-tabling internet poker, but certainly possible within any "unlucky" hour or two.

[/ QUOTE ]
An acceptable bankroll does not mean you are certain to avoid going bankrupt. It means your risk of ruin should be very low. The probability that you lose 25 BB in a couple of hours on all 12 tables is much lower.

[Voltron87]

His idea is right. Since live games are generally much softer than online ones, generally your bankroll will not need to be as large for them. If you have two players playing in the same 5/10 game, a person winning at 5bb/100 does not need a bankroll as large as someone winning at 1bb/100.

[theghost]

2 obvious (but relevant) points I would have liked to see addressed by this article were:

1. Your winrate/100 may go down slightly at multiple tables, but your hourly will certainly go up if you are beating the limit. (I do agree that lower winrate/100 would mean you need more BR.)

2. If you have bankroll concerns, you can cut your br requirements almost in half by playing 2 or more tables of 10/20 instead of one live table of 20/40 (for example). I think Roy dropped the ball leaving this out of the article.

[Innocentius]

[ QUOTE ]
I was allways under the understanding that with multi-tabling you just hit the long run faster. E.g. everything else being equal, the chance of a 300BB downswing at 1-table is exactly the same chance as simulataneous 25BB downsings at 12 tables.

[/ QUOTE ]

If you disregard the drop in quality of your play due to multitabling, this is completely right.

[pudley4]

[ QUOTE ]
If you play one game of 10-20 let's agree you need a 6,000br, or a 300bbbr.

Whether you play this one game fast or slow is irrelevant; you still only need this 6,000br.

However, independently, each game you play needs this 6,000br, so if you played 4 tables at once and kept separate bankrolls for each game, for some strange reason, you would need a 24,000br in case each one of your games hit a major downslide at the same time.

Obviously, you don't really need to keep 4 separate bankrolls, because, in theory, you would have to reduce the number of games you were playing to 3 once your total br reduced to 18,000, and to 2 at 12,000, down to 1 at 6,000, to stay in line with standard br advice.

Working up the other way, as long as you know you should reduce to 1 game once your br has reduced to 6,000, then, in practice, it is alright to "push your luck" a little and play more than one game.

There is an exact mathematical way to work it out, but it is probably reasonable to say that a 100bb downturn on each table at once would be "unlucky", so, if you are playing four tables, with a view to reducing to 1 table in a "br emergency", a 4,000 + (4x2,000) = 12,000br should be safely sufficient.

That's a 600bbbr for a 10-20 4-tabler.

imo

[/ QUOTE ]

No. Not even close.

Your bankroll requirement is based on your win rate and standard deviation. These are commonly measured in BB/100 hands. It doesn't matter how fast or slow you play those hands; the more hands per hour, the higher your win rate per hour and the higher your SD per hour. Conversely the fewer hands you play per hour, the smaller your WR and SD are per hour. These changes in WR and SD offset exactly enough to keep your bankroll requirement the same.

The only difference in adding more tables is that your WR may go down - this change is the biggest reason for the increase in BR requirement when multi-tabling, NOT the increased number of hands/hr.

[Voltron87]

[ QUOTE ]
However, independently, each game you play needs this 6,000br, so if you played 4 tables at once and kept separate bankrolls for each game, for some strange reason, you would need a 24,000br in case each one of your games hit a major downslide at the same time.

[/ QUOTE ]

this is waaaaay off