[ QUOTE ]
If you're actually a winner at 2/4nl (and all the stakes below) and only have $50 for some reason, the 20 buyin rule for you is terrible. You shouldn't keep yourself stuck in .05/.10 for a huge number of hands waiting to build your roll to $500 for .10/.25, you're earning severely below what your hourly rate would be at higher stakes.
[/ QUOTE ]
While I agree that the 20 buy-in guess is terrible, you don't have the option of winning your normal hourly rate. That's part of the penalty of falling to $50. Not only have you lost money, but you have lost earning power.
Many people overlook the fact that your win rate affects your bankroll requirements. Lower stakes games are much, much softer than higher stakes games. If 20 buy-ins is right for you at NL $50, then it is probably ridiculously conservative at NL $2, and far too aggressive at NL $400. (The SSNL FAQ incorrectly suggests that you should have 20 buyins no matter what. See
this thread. I am working on a replacement.)
See this
educated guess at a conservative schedule for moving up which reflects win rates achievable at different levels.
[ QUOTE ]
See
http://www.twoplustwo.com/magazine/i...kratz0707.html for a good treatment on aggressive bankroll management, which should be most applicable in a situation of a good player whose roll forces him several levels in stakes below his normal game.
[/ QUOTE ]
There are quite a few problems with that article.
[img]/images/graemlins/frown.gif[/img] The author performed a complicated miscalculation instead of using a simple Kelly formula: bankroll = standard deviation^2 / win rate. SD is about 1.7 buy-ins or 170% for 9 or 10 player SNGs.
[img]/images/graemlins/frown.gif[/img] Second, the swings with the Kelly criterion are too sickening for almostt everyone. Most people would prefer to use a fractional Kelly system such as bankroll = comfort * standard deviation^2 / win rate, where most people would choose a comfort level between 2 and 4. Comfort is the reciprocal of the Kelly fraction k.
[img]/images/graemlins/frown.gif[/img] "Growth versus value" is a tradeoff, not a dichotomy.
[img]/images/graemlins/frown.gif[/img] The ROI assumed was 30%, which is a lot higher than many people believe is sustainable in games of moderate stakes. The STT FAQ actually suggests 20% isn't sustainable in $6 turbo SNGs, which I think is wrong, but it shows the problem.
You can find some good discussion in the magazine discussion forum in
this thread, as well as some flames after the OP recommended a get-rich-quick site.