Hold up, wait a minute
Ok sorry for that last whiny crybaby post. At the time I was in a dark asshole of a place.
It's amazing how much a night of playing Wii can cheer you up.
I decided to load up another round of SNGs once I was in a better mood. All signs pointed to me playing a standard, non tilty, +EV game. And the results, while not spectacular, were actually positive (+$204). So theres that.
But I was still very unsure about my game, and the state of the games in general. I needed a way to decipher why I was losing. And since I have a lot of spare time, I decided that I was going to tally up all my all in's from the last 3 days to see how I am doing in the luck department.
I manually went through the last 105 games, searching for every all-in confrontation that went to showdown. I divided them into categories based on probability of winning. I had 5 such categories: 80% favorite (pair over pair, pair vs undercards), 70% Favorite (dominating hands, pair vs 1 overcard), Coinflips (pair vs overcards, 2 unpaired undominated cards), 30% underdog (dominated hands, 1 overcard vs pair), 20% underdog (pair under pair, undercards vs pair)
I ended up getting pretty liberal in my definition for "coinflip". I would basically count any all in where two hands went against each other in which each card was live as a "coinflip". Therefore a confrontation like 8T vs AK ended up falling under the "coinflip" category, despite 8T being a 35%~ dog. While this seems like it would skew the results, what ended up happening is that the times that I was on the 35-45% side of the coinflip, offset with the times that I was on the 55-65% side. Obviously not a perfect system, but still a decent indication of true variance.
Results:
Based solely on this data, no matter how good one is at poker, if one has the data of the losing session, one will lose many monies. And the fact here is, when I was losing my bankroll there, I was also losing a shit ton of coinflips. Not only that, each category is actually performing lower than what is expected. Since luck is neutral, in the long run those percentages will read 80%, 70%, 50%, 30%, 20%. It doesn't matter how good or bad I am, the expectation for these stats won't change.
So thats a fairly good sign that I'm not completely inept. I think I mostly was freaking out because usually when losing a lot, I am running into big pairs preflop. But as the stats show, I only ran into a 20% underdog situation 6 times.
It also makes sense that a winning session will have similar stats to the ones that were tracked here. I was winning slightly more than my share of coinflips, but also losing in all the other categories. So basically, everything in the luck department was fair and balanced. With luck balanced I can then go to work and play my game, scraping off chunks of EV where I can.
A flaw in this method of tracking luck is basically weighing the impact of different situations. For example, in this basic method, I would track each all in and give it the same weight in the final results. A hand where I called someones 3bb all in with A5 and lost to their KT counted the same as the times where I 3 bet pushed QQ on the bubble and lost to the chip leaders AK. I mean, clearly some all ins are more important than others. But just as in the "coinflips" tally method, I figure that these situations counteract each other in the long run as well.
I'll probably never do this method of tracking again since it took a good 3 hours to load up all the games and find the all-ins. But I think it was needed. I needed some evidence to assure me that my losses were still due to variance, as at times it sure didn't feel like it.
And heres a hand that sums it all up nicely.
Dealt to RikkiDee [8c Qs]
RaiNKhAN: folds
mickael23: folds
Fat Jaffa: folds
afiopeneyes2: folds
RikkiDee: raises 390 to 490 and is all-in
sbeanz33: calls 490
GetItQuietly: folds
trevboy: raises 3035 to 3525 and is all-in
sbeanz33: calls 1625 and is all-in
*** FLOP *** [6d 9d 8d]
*** TURN *** [6d 9d 8d] [7s]
*** RIVER *** [6d 9d 8d 7s] [2c]
*** SHOW DOWN ***
trevboy: shows [Ah Ad] (a pair of Aces)
sbeanz33: shows [As Ac] (a pair of Aces)
RikkiDee: shows [8c Qs] (a pair of Eights)
It's amazing how much a night of playing Wii can cheer you up.
I decided to load up another round of SNGs once I was in a better mood. All signs pointed to me playing a standard, non tilty, +EV game. And the results, while not spectacular, were actually positive (+$204). So theres that.
But I was still very unsure about my game, and the state of the games in general. I needed a way to decipher why I was losing. And since I have a lot of spare time, I decided that I was going to tally up all my all in's from the last 3 days to see how I am doing in the luck department.
I manually went through the last 105 games, searching for every all-in confrontation that went to showdown. I divided them into categories based on probability of winning. I had 5 such categories: 80% favorite (pair over pair, pair vs undercards), 70% Favorite (dominating hands, pair vs 1 overcard), Coinflips (pair vs overcards, 2 unpaired undominated cards), 30% underdog (dominated hands, 1 overcard vs pair), 20% underdog (pair under pair, undercards vs pair)
I ended up getting pretty liberal in my definition for "coinflip". I would basically count any all in where two hands went against each other in which each card was live as a "coinflip". Therefore a confrontation like 8T vs AK ended up falling under the "coinflip" category, despite 8T being a 35%~ dog. While this seems like it would skew the results, what ended up happening is that the times that I was on the 35-45% side of the coinflip, offset with the times that I was on the 55-65% side. Obviously not a perfect system, but still a decent indication of true variance.
Results:
Based solely on this data, no matter how good one is at poker, if one has the data of the losing session, one will lose many monies. And the fact here is, when I was losing my bankroll there, I was also losing a shit ton of coinflips. Not only that, each category is actually performing lower than what is expected. Since luck is neutral, in the long run those percentages will read 80%, 70%, 50%, 30%, 20%. It doesn't matter how good or bad I am, the expectation for these stats won't change.
So thats a fairly good sign that I'm not completely inept. I think I mostly was freaking out because usually when losing a lot, I am running into big pairs preflop. But as the stats show, I only ran into a 20% underdog situation 6 times.
It also makes sense that a winning session will have similar stats to the ones that were tracked here. I was winning slightly more than my share of coinflips, but also losing in all the other categories. So basically, everything in the luck department was fair and balanced. With luck balanced I can then go to work and play my game, scraping off chunks of EV where I can.
A flaw in this method of tracking luck is basically weighing the impact of different situations. For example, in this basic method, I would track each all in and give it the same weight in the final results. A hand where I called someones 3bb all in with A5 and lost to their KT counted the same as the times where I 3 bet pushed QQ on the bubble and lost to the chip leaders AK. I mean, clearly some all ins are more important than others. But just as in the "coinflips" tally method, I figure that these situations counteract each other in the long run as well.
I'll probably never do this method of tracking again since it took a good 3 hours to load up all the games and find the all-ins. But I think it was needed. I needed some evidence to assure me that my losses were still due to variance, as at times it sure didn't feel like it.
And heres a hand that sums it all up nicely.
Dealt to RikkiDee [8c Qs]
RaiNKhAN: folds
mickael23: folds
Fat Jaffa: folds
afiopeneyes2: folds
RikkiDee: raises 390 to 490 and is all-in
sbeanz33: calls 490
GetItQuietly: folds
trevboy: raises 3035 to 3525 and is all-in
sbeanz33: calls 1625 and is all-in
*** FLOP *** [6d 9d 8d]
*** TURN *** [6d 9d 8d] [7s]
*** RIVER *** [6d 9d 8d 7s] [2c]
*** SHOW DOWN ***
trevboy: shows [Ah Ad] (a pair of Aces)
sbeanz33: shows [As Ac] (a pair of Aces)
RikkiDee: shows [8c Qs] (a pair of Eights)
posted by RikkiDee at 1:26 AM
3 Comments:
Good to see that you found yourself again... that was quick.
You might want to have some donking fun out playing poker live with Kat, Astin and me... $40 buy in against a bunch of degenerates near downtown Toronto.
Let me know if you are interested... email is guin36 at gmail dot com.
thats what I want to hear! GG!!!!@#!@#!@#!@#!@#!@#!@#`
we'll see...
Post a Comment
<< Home