Quote:
Originally Posted by meale
This whole thing about "10bb/hour" really irks me. Plenty of people are saying crushers can win at 2/5 at 10bb/hour which is so BS imo. @ 25 hands/hr, this gives us a 40bb/100 winrate. There is no way any player on the planet can attain such a winrate after any reasonable converged sample. Magorko (one of the 500z regs) recently played 20k hands at 2NL with a 20bb/100 winrate. 20bb/100 at $2NL. From one of the biggest crushers in online poker. This is likely unsustainable for 200k hands either. So how can we be saying 40bb/100 is attainable by a $500NL live "crusher"? Sure 10bb/hr is achievable over 100-500 hours, easy. But it's just not possible over a converged sample.
I would think at 2/5+, if you're winning at 5bb/hr over 2k hours or more, you're crushing.
the skill gap is much narrower in online poker. in live there are people who routinely sit down with little to no knowledge of the game at 1/2 and 2/5, obviously less often at 2/5, but it happens. many of the people playing 2/5 wouldnt be able to beat 50nl online, some might not be able to beat 20nl. you get a very knowledgable person playing against people like that and the only way the lose significant money is by being coolered.
Quote:
Originally Posted by Dopedupwalrus
Well... assuming that you are improving your game with those adaptations, or at least trending so that you will be improving over time, can you not still infer things? #Mostlikelyanignorantnaiveresponse
meh, not too much really.
win rate? 30k hands is a drop in the bucket. with an avg of 25 hands/hr, 30k hands = 1200 hours. say you have a 10bb/hour win rate. thats $12,000. say your next session you play 8 hours and you lose $900, 3 BIs for most places. that drops your win rate from 10bb/hr to 9.19bb/hr.
i'd say when your winrate doesnt fluctuate more than a maybe a tenth or 2 of a point when winning/losing multiple buy ins, you can start to actually infer things.
but then think about this => my current online database has 300k hands. all of my PPs have over 1k sample each. i have a + return on all of them except for 88/77. the sample size is still too small for them. so even once you get into huge practically unattainable numbers for live players (300k hands = ~12k hours, probably 99% of players never make it that long), you're still unable to really pull out any data past a win rate. i need probably 20x hands (so 6 million) to get anywhere near an accurate representation of how i play my PP.
Last edited by johnny_on_the_spot; 09-24-2015 at 09:49 AM.