|
|
#1
05-09-2005, 03:50 PM
|
|||
|
|||
I just uploaded 13,000 hands into Pokertracker. Results
wow, I just uploaded 13,000 hands.
The hand that was the biggest loser for me over 13,000 hands was AK o. I'm like what the..... My winrate was 9.1 BB/100 which makes me wonder whether 30 BB/100 really is sustainable or maybe I've gotten better since this. Not sure. I have two different names on each site. Anyone know how to merge these ? Here are the worst hands money-wise in order 1. AKo 2. QTsuited 3. A6o 4. J6suited 5. KQsuited 6. 22 7. 87o 8. QJo 9. JJ 10. T3o Top money makers 1. KK 2. QQ 3. AA 4. AKsuited 5. 44 6. 55 7. AQo Maybe 13K isn't even enough to be statistically relevant. 
|
|
#3
05-09-2005, 04:53 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
13,000 sounds a lot but it isn't really when you come to consider individual poacket cards.
Since there are 169 possible starting hands you are talking about an average of 77 hands per starting hand. So, one really bad or really good hand can throw the values off.
|
|
#4
05-09-2005, 04:53 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
[ QUOTE ]
The hand that was the biggest loser for me over 13,000 hands was AK o. I'm like what the..... [/ QUOTE ] I dont know if I'd be too surprised at that; the thing that would surprise me the most about that stat is that it was AK and not AQ or AJ or something. I think that in almost all cases your biggest loser will be a marginal hand; your terrible hands won't lose you money except when you're forced to put in blind bets holding them, and your great hands won't lose you money because...well, because they're great. [img]/images/graemlins/smile.gif[/img] [ QUOTE ] My winrate was 9.1 BB/100 which makes me wonder whether 30 BB/100 really is sustainable or maybe I've gotten better since this. Not sure. [/ QUOTE ] I have no idea what that means so I'm just going to move on. :P [ QUOTE ] Here are the worst hands money-wise in order 1. AKo 2. QTsuited * 3. A6o 4. J6suited 5. KQsuited * 6. 22 7. 87o 8. QJo 9. JJ 10. T3o [/ QUOTE ] * just for the record, I hate hate HATE these hands. On a side note, wtf were you doing playing T3 offsuit and J6 suited?? [ QUOTE ] Top money makers 1. KK 2. QQ 3. AA 4. AKsuited 5. 44 6. 55 7. AQo [/ QUOTE ] I think it seems like you have more than enough hands to get a good picture of where you're at statistically (even though it appears that you flop a disproportionate number of sets with fours and fives for some reason [img]/images/graemlins/smile.gif[/img]) In any event, thank you very much for sharing - as a newbie who hasn't gotten to the 13k hand point yet, it was very enlightening.
|
|
#5
05-09-2005, 05:00 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
[ QUOTE ]
13,000 sounds a lot but it isn't really when you come to consider individual poacket cards. Since there are 169 possible starting hands you are talking about an average of 77 hands per starting hand. So, one really bad or really good hand can throw the values off. [/ QUOTE ] Whoa whoa whoa - that's an excellent point to put things in perspective, but there's a serious fallacy there. You're making the implicit assumption that each of the 169 hands is equally likely to be dealt, and that's not the case (AKo is twice as likely as AA and three times as likely as AKs) So as far as likelihood of the 169 hands goes, you have three types - suited hands, offsuit hands, and pocket pairs. XYs: 4 ways to make XYo: 12 ways to make (you don't count the fourth "Y" because that would give you XYs) XX: 6 ways to make so all things being equal, your data on offsuit hands should be the most statistically accurate since you get them far more often, followed by your pocket pairs, and finally the most statistically volatile hands in your sample will be your suited hands.
|
|
#6
05-09-2005, 05:14 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
In the interest of mathematical curiosity, here's roughly how many hands of each category (and specific hand) you should see:
169 hands - 13 pairs = 156 non-pairs / 2 states (suited and offsuit) = 78 "XY" combos XYs ~= 18.2% of your hands ~= 2,364 hands / 78 combinations ~= an average of 30 times seen per XYs hand. XYo ~= 54.5% of your hands ~= 7,091 hands / 78 combinations ~= an average of 91 times seen per XYo hand. XX ~= 27.3% ~= 3,545 hands / 13 ranks of pair ~= an average of 273 times seen per pocket pair. so after 13,000 hands, you should see AKo about 91 times, AK s about 30 times, and AA about 273 times, so my earlier post should be corrected - your data on specific pocket pairs will be the most statistically stable (I had forgotten to take into account that while its rarer to get a pair than XYo, youre much more likely to see a specific pocket pair before you see a specific XYo, since there are only 13 ranks of pair and 78 XY hands)
|
|
#7
05-09-2005, 05:21 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
[ QUOTE ]
On a side note, wtf were you doing playing T3 offsuit and J6 suited?? [/ QUOTE ] sample size. I have about 60k hands and my 4th best hand is 53 suited because I've won 4 gigantic pots with it, more than making up for the 95% of the time I never play it.
|
|
#8
05-09-2005, 05:28 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
[ QUOTE ]
XX ~= 27.3% ~= 3,545 hands / 13 ranks of pair ~= an average of 273 times seen per pocket pair. [/ QUOTE ] This isn't even REMOTELY close to correct. A pocket pair is 16:1 against. The Doc
|
|
#9
05-09-2005, 05:34 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
...and I screwed up my math again. *Sigh* someday I'll get good at this, I swear. :P
anyway, here's how it actually works if you have both common sense and a calculator (i've tried my best with only one of these :P) OK so there are 4 ways of making XYs, 12 ways of making XYo, and 6 ways of making XX. XYs = 4 ways * 78 XY combos = 312 total XYs combos XYo = 12 ways * 78 XY combos = 936 total XYo combos XX = 6 ways * 13 pairs = 78 total XX combos so for you math geeks keeping score at home, I used Bayes' Theorem, but incorrectly because they're not equally weighted (although there are fewer ways to make a specific pocket pair than there are to make AKo, there are fewer pocket pairs to begin with) which is why the pocket pair data in my calculations is very skewed. So your chances of being dealt any particular class of hand are (given 312+936+78 = 1,326 total possible hands) XYs = 312 / 1,326 = 23.52% XYo = 936 / 1,326 = 70.58% XX = 78 / 1,326 = 5.88% so then you will see XYs ~3,059 times, XYo ~9,176 times, and a pocket pair ~765 times. which means that any specific XYs will be dealt to you roughly 39 times, XYo ~118 times, and each specific pocket pair will be dealt to you roughly ~59 times. Sorry for the confusion, and I really didn't mean to threadjack (honest!) - just mucking around out of curiosity. Feel free to ignore (as I assume the majority already do). Again, thanks for the brain food, redpill.
|
|
#10
05-09-2005, 05:38 PM
|
|||
|
|||
|
Re: I just uploaded 13,000 hands into Pokertracker. Results
[ QUOTE ]
[ QUOTE ] XX ~= 27.3% ~= 3,545 hands / 13 ranks of pair ~= an average of 273 times seen per pocket pair. [/ QUOTE ] This isn't even REMOTELY close to correct. A pocket pair is 16:1 against. The Doc [/ QUOTE ] Yeah, I know - see my other post; I weighted the probabilities of each class of hand incorrectly for Bayes's theorem. But on a personal note, thanks for wording your post in such a way as to sound like a jerk when all you were doing is "correcting" what I did via actual work with a figure that I'm guessing you probably only came by because you read it in a book. I appreciate it.
|
 |
|
|
Linear Mode
