Bankroll Formula?

[caseclosed326]

Is there some generally accepted formula for what bankroll to have at a certain limit? i.e. 100times big blind per table for multitableing. i know it would be different for multitabling and 6max tables, i am mainly interested in 10max limit tables but would be interested in the others as well.

thanks!

[Rudbaeck]

Rule of thumb is 300bb. This is plenty if you play recreationally. If you play poker for a living something like 6 months expenses in a savings account and a 1000bb playing roll is more appropriate.

After you're settled in a limit you can use BruceZ's formula for bankroll. Notice that the winrate is the amount re-invested in the bankroll, not your actual win rate.

I haven't seen anyone do a statistical analysis of the risk of ruin of a non-growing bankroll, which is kind of interesting as many authors seem to believe the 300bb bankroll is a non-growing bankroll. (It's a starting point for a rapidly growing bankroll belonging to a highly skilled person playing in soft games with little variation actually.)

[caseclosed326]

Thanks. I knew it would be high. So id need $5400 to 3table 3/6 safely, yikes.

[Belok]

its 300BB standard, not 300BB per table.
1800 is what you need.

Even if you were playing 10 tables 1800 would be your BR.

[grouchie]

I disagree with this statement...
I think that with all the hands you see online playing multiple tables you need to go with about 300bb for EACH table you are playing.

When the concept was first developed it was developed for live games I belive. Live you can only play one table at a time, so everything fit.

Now that you can play online and multiple tables at once you need a different type of bankroll requirement. While it may not be 300bb per table, it is still going to be higher than 300bb total.

At least this is my opinion, not sure how others feel about it.

[Yads]

BR = -(SD^2 * Ln R)/(2 * WR)

BR - Bank Roll
SD - Standard Deviation
R - Your accepted Risk of Ruin
WR - Win Rate

eg. WR = 2 BB/hour, SD = 16 BB/hour R = 0.001 (0.1%)
BR = -(16^2 * Ln 0.001)/(2 * 2)
= 442 BB

[moondogg]

[ QUOTE ]
I disagree with this statement...
I think that with all the hands you see online playing multiple tables you need to go with about 300bb for EACH table you are playing.

When the concept was first developed it was developed for live games I belive. Live you can only play one table at a time, so everything fit.

Now that you can play online and multiple tables at once you need a different type of bankroll requirement. While it may not be 300bb per table, it is still going to be higher than 300bb total.

At least this is my opinion, not sure how others feel about it.

[/ QUOTE ]

As long as you have the necessary bankroll to play the game in the first place, you don't need more for multiple tables. Your swings SD should be no bigger on one table than on 4 tables, as all four tables are independent events, regardless of whether they happen simultaneously. Granted you have more money in play at a give time, but that money is not wagered on a single event, rather it is wagered on 4 distinct independent events. Those 4 hands are 4 hands, and it makes no difference if you were to make those 4 wagers simultaneously or sequentially, your EV and SD are the same.

You should expect to have the same results if you play 200 hands at a single table in four hours than if you were to play 200 hands spread out over 4 tables in one hour.

You will see more swings at more tables, but you would have seen those swings eventually on one table, so if you were not properlu bankrolled to handle them when they occur on multiple table then you were not properly bankrolled for one table.

Additionally, your "long-run" is a function of how many hands you play, not how long you play. By playing more hands per hour you get to the long-run faster, which decreases your per-hour variance (while your per-hand variance remains the same).

One caveat, however: if you play suffers as a result of multitabling and your win rate goes down, this can increase your variance (even if you are still a winning player). But the point is that your bankroll requirements do not grow in direct proportion to the number of tables you play, only minor adjustments may need to be made for your weakened hourly rate.

[Patrick del Poker Grande]

[ QUOTE ]
BR = -(SD^2 * Ln R)/(2 * WR)

BR - Bank Roll
SD - Standard Deviation
R - Your accepted Risk of Ruin
WR - Win Rate

eg. WR = 2 BB/hour, SD = 16 BB/hour R = 0.001 (0.1%)
BR = -(16^2 * Ln 0.001)/(2 * 2)
= 442 BB

[/ QUOTE ]
I have a hard time believing that the units in this formula should be BB/hr instead of BB/100. Is this a typo, or does someone disagree with me?

[moondogg]

[ QUOTE ]
[ QUOTE ]
BR = -(SD^2 * Ln R)/(2 * WR)

BR - Bank Roll
SD - Standard Deviation
R - Your accepted Risk of Ruin
WR - Win Rate

eg. WR = 2 BB/hour, SD = 16 BB/hour R = 0.001 (0.1%)
BR = -(16^2 * Ln 0.001)/(2 * 2)
= 442 BB

[/ QUOTE ]
I have a hard time believing that the units in this formula should be BB/hr instead of BB/100. Is this a typo, or does someone disagree with me?

[/ QUOTE ]

I believe this is the B&M formula, which assumed you could not multi-table or count specific the number of hands played, and therefore calculated BB/hr as roundabout way to get the info represented by BB/100. Just my guess.

[Patrick del Poker Grande]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
BR = -(SD^2 * Ln R)/(2 * WR)

BR - Bank Roll
SD - Standard Deviation
R - Your accepted Risk of Ruin
WR - Win Rate

eg. WR = 2 BB/hour, SD = 16 BB/hour R = 0.001 (0.1%)
BR = -(16^2 * Ln 0.001)/(2 * 2)
= 442 BB

[/ QUOTE ]
I have a hard time believing that the units in this formula should be BB/hr instead of BB/100. Is this a typo, or does someone disagree with me?

[/ QUOTE ]

I believe this is the B&M formula, which assumed you could not multi-table or count specific the number of hands played, and therefore calculated BB/hr as roundabout way to get the info represented by BB/100. Just my guess.

[/ QUOTE ]
I understand that, but I would think that in order for this to be a good formula, you'd still have to calculate using BB/100 and that you'd have to convert from BB/hr to BB/100 before making this calculation. If you assume something like, for example, k=2 for (BB/100)=k*(BB/hr), then when you do the calculations with the same WR, you'll get different answers depending on if you used BB/100 or BB/hr.

Example using the numbers from above:
BR = -((16*2)^2 * Ln 0.001)/(2 * (2*2))
= 884BB

Essentially, it's off by a factor of k because of the SD^2 term over the WR term.

Now, which is right - BB/100 or BB/hr? It seems to me that it should be BB/100. Although, if you think about it, if these two come out differently, that must mean something is wrong with the formula itself, since the units don't work out. You have BB on the left side of the equation and BB/hr or BB/100 on the right. You should be able to do it with BB/hr or with BB/100 and it shouldn't matter as long as it's consistent across the whole formula. The only way I'll drop this argument is if the person who came up with this is just doing it empirically and it's some sort of rule of thumb/guideline and he didn't put in some factor to square up the units.

I must be missing something crazy simple, no? It's been a while since stats class. Someone please show me where I'm being an idiot.

Also, what's a typical range for SD?