bart implies a laffer curve, and the numbers say you need to start playing 3x higher

sometimes there will be times to play above one's bankroll, but others are short buyers

I just want to reiterate that these comments were about playing professionally. The majority of players are not professional and therefore don't need to be as careful with a bankroll.

Optimal bankroll can not be determined in a vacuum.

And just like with gold rush, the real winners on aggregate were those who were selling supplies.

Actually if you have an unlimited bankroll then martingale works.

True but in that case you could also just hop through your wormhole and start over.

Hi Everyone:

There's a lot to this statement and I thought as a former professional statistician my comments might be interesting to some of you. Also, I haven't read the thread, so some of what I'm about to say may have already been covered by others.

It's always been my belief that if you're capable of making a good living at poker you're probably smart enough to better at something else. Of course, this assumes that you can deal reasonably well with people which not all poker players can.

This is certainly the case with some people, and is more likely to be present, in my opinion, among those playing big games where most of the players play well (thus producing a low relative win rate). But it's also not true with everyone, and it's not difficult to estimate a good bankroll size.

It isn't even clear to me what he's saying here. My guess is Hansen seems to think the variance of live poker will dominate your results so much since you can't play enough hands compared to the Internet (and you have to play at higher stakes to have the same long term income) that it's not worth playing at all. I don't agree except with the possible exception of high stakes games where you are not game selective and are frequently in games where everyone plays well.

No it's not. The easiest way to do this that I know of is to use the maximum likelihood estimator as described in my book Gambling Theory and Other Topics which was first published in 1987 and which many people on here occasionally reference.

I assume he's estimating the variance for his blackjack play without the comps and the exact same technique mentioned above works in the exact same way.

I never did this. Again using the techniques outlined in my book, I estimated my required bankroll at three standard deviations, increased it by approximately 10 percent to account for what I call the non-self weighting effect of not having all games you play in being identical, and never had a bankroll issue. (And for those who don't know, the standard deviation is the square root of the variance, and it has some nice properties which are useful for these type of calculations.)

However, many people discover that when using this method, the required bankroll is larger than what they intuitively think it should be. So perhaps his "multiply your bankroll (liquid cash to play with) by 3!" isn't as off as it first appears to be.

Best wishes,
Mason

Required bankroll is based on the relationship between your win rate and standard deviation. For no-limit experts this relationship may actually be too good if they're against relatively weak players. Thus this is a reason why no-limit games tend to get tougher over time while games like limit hold 'em and stud don't change as much. But that's another issue.

Best wishes,
Mason

I assume he's talking about a game with six to nine players. As for heads-up, I'm not knowledgeable enough to comment.

I agree, but they also have a lower win rate and your required bankroll has something to do with the relationship between winrate and variance.

Just to give one example, and some of you might disagree. But it's my opinion that expert players have a similar win rate in $2-$5 no-limit hold 'em and $20-$40 limit hold 'em. Yet the fluctuations in the limit game are far larger.

Maybe. I think this is probably a function of your specific opponents assuming you play each game well. But I really don't know.

Best wishes,
Mason

Hi DGAF:

I don't think you're using the term edge correctly. If you used win rate instead, then what you're saying makes sense. Specifically, PLO tends to have larger pots than NL, thus if would be possible to have a higher win rate but a lower edge at the same time. Do you see the difference?

Best wishes,
Mason

Hi BustoPro:

I don't quite agree. As I argued in my book Gambling Theory and Other Topics many years ago, those strategies that increase your win rate often come with an increase in variance. (There are exceptions, with the ability to read hands being most notable). So those players with a lower win rate may not need as much bankroll as you indicate since their variance may be lower than the player with a higher win rate. You're assuming the variance remains the same and that may not be the case.

Best wishes,
Mason

I agree.

Best wishes,
Mason

Hi gman06:

If you're playing a very small number of hands per hour, your 77% win rate may not be statistically accurate yet. One way of telling is knowing how often other top players win. If others who you think play well are no where close to 77 percent, then your number is probably too high. But if others are doing this well, then your estimate is probably correct.

I also agree that your HU variance could be lower than at ring games if your opponents are making certain errors. Most notably folding too much, especially on the river.

Best wishes,
Mason

Hi dank:

There's another thing that's missing here, and it's are you willing to move down if you've been losing? For instance, if you lose half your bankroll, are you willing to move to a smaller game? If you are, then you probably need a bankroll roughly two-thirds as large as what would be required if you refuse to move down.

Best wishes,
Mason

No. Your required blackjack bankroll for a given win rate has a lot to do with the specific rules offered, including the number of decks in play and what percentage of the cards dealt out. Based on my experience from years ago, blackjack swings can be far worse than poker swings.

Best wishes,
Mason

Required bankroll to assure survival is somewhat counter-intuitive to many players and the amount they really need does seem to be too much. That's because they need to account for the very large swing which only rarely happens.

MM

Hi Mason Malmuth:

Go home. You're drunk.

Best wishes,
Nutsfl0pper

And just as those real winners were on the supply side, many continued on to become some extremely wealthy people in later years...
This rush hasn't yet "panned out" for a lot of people.

Can you elaborate on this? Standarddeviation over how many hands? The unit of standard deviation most often used seems to be BB/100 hands, but I somehow doubt you would advise to play a game with a 20BB/100 stddev on a 66BB bankroll.

Again, I want to point out that I pulled a few quotes that stand out and Bart shouldn't be held to any of this until you hear the whole interview in context.

As far as I understand it - he's saying many live poker players underestimate the swings that come with poker and play on too small a bankroll. This definitely rings true for me and I'd imagine anyone else who has played regularly/seriously.

it might help if someone put some real life, honest to God numbers out there, anonymously or otherwise.
Theoretical formulas attempt to apply to everyone but like any one-size-fits-all thing they apply to no one.

I can confirm playing on a five buy in roll live is a terrible idea*






*Which I talked about in my newly published book.

I think you underestimate what happens when the fish get targeted by the pros. Many of you think that while you are coasting around in the card rooms making some money that you aren't targeted for the big haul in by the players who has been living of poker for close to a lifetime. I see many who come and play who has a big win when they start out. I would argue that the ev you have the first 5 tournaments and first month of cash probably never will be higher ever again. Once they get to know you the game will become tougher and they will target you. Making 10BB's/hour could become 7BB's/hour once they identify the biggest leaks, after that they pick at the small stuff and you could end up at 3BB's/hour.

I say most blame to much on variance and use this as some devil figure, when I was playing a bit more I will admit that I had some bad streaks and some punishment by this great variance but I will say that most of my "bad luck" was a bad game. I simply didn't play my A-game and let a few important pots slip out of my hands or got involved in some pots I didn't belong in. I know this was my fault and not some god of variance deciding to hassle with me. Given that I have taken some notice to how my game has played out over the years I will admit that variance probably has helped me out more then ruining it for me. I know a few hands where I was far behind but hit to win tournaments but can really only say that there are two tournaments where variance has kicked me in the balls.

The reason why you need so much money in a liquidity market is that this is what you are working with and also making a living of. If you have savings of $1m or a house that you live in for free this should by all means be included in your BR. You can't say that you have a BR of $10k playing games and never need to pay any bills with your winnings. The BRM theories are based on you providing for yourself and other from poker, not having it as a casual hobby if you don't have anything better to do.

Lol at this.