Bankroll requirements for mult-tablers

[CaptainSubtext]

As recommended you should have at least 20 buy-ins for the level of cashgame you are playing, words said many times before.

But what is the bankroll requirement if you multi-table?
Is 20 buy-ins also enough when you 3-table? Or should I have 60 buy-ins for 3 tables? 20 buy-ins X number of tables = bankroll requirement?

[OrangeKing]

Not exactly - you're just playing your hands faster, meaning you're getting to the long run faster. In and of itself, it doesn't change your risk of ruin.

That said, if multitabling lowers your BB/100 winrate (which it almost certainly does to some level, big or small), you will need a somewhat larger bankroll to keep the same risk of ruin. I'm not sure if there's a "standard" for how much larger, but it's not nearly as much as multiplying the bankroll requirement by the number of tables you're playing.

[reno expat]

For multi-tabling, I like to have 10-15 (depending on the game and limit) buy-ins per table. For example, if was I was 4-tabling 50NL, I would want about 40 buy-ins. If I was 3-tabling 100NL, I would probably want somewhere around 36.

[lucky_mf]

[quoteThat said, if multitabling lowers your BB/100 winrate (which it almost certainly does to some level, big or small), you will need a somewhat larger bankroll to keep the same risk of ruin.

[/ QUOTE ]

Additionally, the potential for tilt to do serious damage to your bankroll is much greater when when playing multiple tables.

Lucky

[pzhon]

[ QUOTE ]
For multi-tabling, I like to have 10-15 (depending on the game and limit) buy-ins per table.

[/ QUOTE ]
If this gives you a fixed risk of ruin, that means your win rate in $/hour does not depend on the number of tables you play, that playing 4 tables instead of 2 cuts your win rate per hand in half. In that case, why are you playing more tables?

Some people want a security blanket. If you need to have 1000 buy-ins in the bank to feel like you can play something other than AA, then you better have 1000 buy-ins. However, rational bankroll management is about letting you weather downswings without having to add more money. In that case, you can use

bankroll = comfort * standard deviation^2 / win rate.

The minimum comfort level you accept depends on your personal risk tolerance and ability/willingness to move down when you hit a bad streak. Most people are happy with a value from 2 (aggressive) to 4 (conservative). If you stay at the same level without withdrawing, your risk of ruin is about e^-(2*comfort).

If your win rate and standard deviation per hand do not change when you multitable, then your bankroll requirements do not change, as you can see from the formula.

[lucky_mf]

[ QUOTE ]
If your win rate and standard deviation per hand do not change when you multitable, then your bankroll requirements do not change, as you can see from the formula.

[/ QUOTE ]

Win rate and standard deviation may not change globally, but multi-tabling might make your play more state dependent than it would if you were just playing one table in the sense that spells of good or bad play last a longer number of hands. Tilt, bouts of sub-par play, and bouts of brilliant play are usually governed by time, not the number of hands. Four hours of tilt playing 6 tables is much worse than 4 hours of tilt playing 6 tables with respect to its potential repercussions for your bankroll.

Lucky

[Albert Moulton]

20 buy-ins is fine for multi-tabling.

But if you run so badly that you end up with about 10 buy-ins at your current level, then you'll want to move down to the next lower level where you will have 20 buy-ins again in order to work your way back.

[pzhon]

[ QUOTE ]
[ QUOTE ]
If your win rate and standard deviation per hand do not change when you multitable, then your bankroll requirements do not change, as you can see from the formula.

[/ QUOTE ]

Win rate and standard deviation may not change globally, but multi-tabling might make your play more state dependent than it would if you were just playing one table in the sense that spells of good or bad play last a longer number of hands.

[/ QUOTE ]
That's true. This is a phenomenon not captured by the models I use, or the bankroll guidelines based on the Kelly criterion.

[ QUOTE ]
Tilt, bouts of sub-par play, and bouts of brilliant play are usually governed by time, not the number of hands. Four hours of tilt playing 6 tables ...

[/ QUOTE ]
Tilt isn't rational, and it is hard to model. I'm not sure I agree that it is based on time regardless of the number of hands. If you go on screaming monkey tilt and push preflop several hands in a row, you might do this until you get called once or twice, not once or twice per table.

There are other phenomena which are similarly outside the basic model. For example, games are softer on Friday nights than they are on Monday afternoon, since more casual gamblers can play on Friday night, and they don't have to work the next day so they may play while inebriated. If you play one table, you might play a few hours at softer times, then a few hours at tougher times. If you multitable, you might play 30 table-hours at softer times, then 30 table-hours at tougher times, etc.

While this increases the chance of busting out in the short run, if you are properely bankrolled, your probability of busting out in the short run is very low. Most of the ways to bust out should come from long dismal stretches, on a larger scale than a few hours of tilt or playing at a bad time. Because of this, the unmeasured state-based variance should have a relatively low effect on your risk of ruin, even though it may greatly increase the sizes of some small swings.

[lucky_mf]

[ QUOTE ]
[ QUOTE ]
Tilt, bouts of sub-par play, and bouts of brilliant play are usually governed by time, not the number of hands. Four hours of tilt playing 6 tables ...

[/ QUOTE ]
Tilt isn't rational, and it is hard to model. I'm not sure I agree that it is based on time regardless of the number of hands. If you go on screaming monkey tilt and push preflop several hands in a row, you might do this until you get called once or twice, not once or twice per table.

[/ QUOTE ]

Some people tilt and start open pushing. Clearly you can only open push so many hands before you are broke (or severely damaged). The open pushing type of screaming monkey tilt you refer to is less common the the subtler forms that just amount to not playing well without all the dramatics. For me the repercussions of the subtler tilt are larger when playing multiple tables. This is true because the tilt last a longer number of hands with playing multiple tables and I am more likely to tilt playing multiple tables. This increased propensity to tilt is most likely caused by larger absolute (downswings) per-unit of time. With multiple tables you are more likely to have multiple horrendous beats shortish period of time via the increased number of hands.

Lucky