Bare in mind that nobody you have in your database has anything like a representative BB/100 figure, if the player with the most hands has only 3000. 10k hands is next to nothing, but might give a vague indication, 20k is a more accurate representation, but I'd think you'd want more than 30k to have a really accurate figure. I have a player on 40 PTOBB/100 for PLO in my database with 2k hands, he's a good player but the figure is meaningless. I've been losing on the tables in the last month and I suspect that I've probably played 5k hands+; long-term my PLO BB/100 figure is about 35 or so (which'd be about 17 or 18 PTO BB/100). However, it's true that the $25 buyins are very weak so I'd say perhaps 20-25 PTO BB/100 might be possible for a good player who makes few mistakes, but I have no evidence for that figure.
If you want a real suggestion about when numbers of hands are significant, here's an interesting point as to just how long term it can be: I'm a long term winner at Cryptologic and Pokerstars (a loose site and a tight-ish one) at PLO, yet (although I don't have the exact figures) I believe I may be, in my lifetime, a losing player (not including bonuses) at Party/Empire (another fairly loose site), which is over about 18 months. I've no idea how many hands I've played there but I suspect it's maybe 15k hands or so (most of which has been this year). I've got 8k Party/Empire hands in my database this year (that's just the ones I've been able to download, don't think it's everything) and I've lost $400 in that period. It's something I've been trying to figure out for so long, although I suspect the real answer may just be variance - including NL, i guess i've played 25-30k hands there and am maybe somewhere around break even, yet I win everywhere else, tighter and looser sites, in all the different games I play. I guess because those hands are spread out over 18 months or so it feels like I can't win there and never will, but perhaps it's just that it's possible (if extremely rare) for even this number of hands to not give a representative sample of play. Given I win at tighter and looser sites there's no real reason for it, so I guess there's a slim chance that the long term is a lot longer than you or I might think.
Suffice to say, 3k hands is nothing, especially on a crazy, variable player who sometimes does odd things - I guess he has a high-variance game and you'd want 4 or 5 times that number of hands before you can read any significance at all on his stats, and maybe 5-10 times the number of hands for a probable accurate winrate.
Monk
xxxxxStatistics: Posted by Felonius_Monk — Sun Jul 24, 2005 11:13 am
]]>