I know I'm a quack for posting this about DR 510 down swing, but

[MMMMMM]

[ QUOTE ]
In fact my worst month ever before this was last January, maybe I should take January off next year.

[/ QUOTE ]

Depends. What sign are you?

[BigBlind]

[ QUOTE ]
Maybe I really don't understand the "long-run" and "long-numbers". However, if you take the time to figure out the chance of a 35% probability coming up only 25% over the course of 1000 trys, you will get a really, really long number!

Dogmeat [img]/images/graemlins/spade.gif[/img]

[/ QUOTE ]

I would be interested to see the number. Anyone care to do the math. I don't know how to do it.

[Baulucky]

-510 is a nasty swing for any to endure. I always wonder how many pros would have taken it had it happened at the beginning of their career. I'm sure you will recover soon enough with the amount of hands per week that you play.

How about taking a couple of days, maybe less, to go thru those 14,000 hands in detail, to fix, once and for all, the several leaks that you must have. It should pay quick enough if it increases your win rate even a tiny bit.

Good luck and thanks for the eye opener, David.

[jtr]

[ QUOTE ]
[ QUOTE ]
Maybe I really don't understand the "long-run" and "long-numbers". However, if you take the time to figure out the chance of a 35% probability coming up only 25% over the course of 1000 trys, you will get a really, really long number!

Dogmeat [img]/images/graemlins/spade.gif[/img]

[/ QUOTE ]

I would be interested to see the number. Anyone care to do the math. I don't know how to do it.

[/ QUOTE ]

The appropriate mathematics is to look at a series of Bernoulli trials where p(success) = 0.35, and ask over 1000 such trials, what's the probability of getting only 250 successes (or lower). This is a straightforward use of the binomial distribution, which approximates to the normal distribution for large N. The result stated by the original poster is indeed unusual, with a probability of about 1 in 100 billion. (To give you a rough idea, this is about the same as the probability of getting 16 or fewer heads when tossing a fair coin 100 times.) So either Party is rigged (ahem), or the original poster should look carefully for any sources of measurement error.

[jtr]

I've been thinking about where the discrepancy might come from.

First, the 35% figure is not the right starting point. If you sit at home with a deck and give yourself AQ, for instance, then deal lots of flops, the chances of hitting at least one of your six outs on the flop is 1 - ((44*43*42)/(50*49*48)) = 32.43%.

But in a real game the expected rate of hitting on the flop could be different. Assume you're a player who will always raise AQ and that the game is tight enough that sometimes your raise will get no callers. Depending on the other players' calling/reraising standards, it's possible then that the distribution of aces and queens in the flops that you actually see is different from the 32.43% figure.

Let's say (just for the sake of the argument) that your opponents will only call when they have an ace (any kicker). Now suppose you get one caller to your raise with AQ. You immediately know that one of your 6 outs is gone, and your chance of hitting on the flop now becomes 1 - ((45*44*43)/(50*49*48)) = 27.60%.

Actually the number could be even lower than this, as maybe more than one of your outs is taken, but let's forget about that for now.

The point is, with a few assumptions about the nature of your opponents in a real game, the probability of success in our Bernoulli trials now moves significantly down from the OP's original 35% estimate. If in fact the right number is something like 27.60%, then the observation of a 24.37% success rate is not weird at all -- the expected frequency of a run this bad or worse is a healthy 1 in 100.

[stinkypete]

you're totally off on the numbers there. if you're counting every single flop, the probability of hitting at least one of your 6 cards should be extremely close to 32.43%.

your logic is correct, in the sense that people are more likely to fold when they have no aces or face cards, but you're totally kidding yourself if you think the "correct" number is 27.6%, or anywhere close to it.

i'd do the math to show you how unlikely that is, but i'm lazy.

[jtr]

Gee thanks, Pete. "Totally off" -- give it to me straight, I can take it.

I am not committed to the right answer being at any particular point between 32.34% and, let's say, 27.6% as a lower bound. But I should have phrased things more carefully -- I can see that my comment "actually the number could be even lower than this", for instance, is misleading.

The reasons behind my post were 1) to show that the original poster's 35% number was too high, 2) to indicate that the right answer is no higher than 32.34%, and 3) to note that the discrepancy, while it might not look like much, would certainly affect how weird an observation of a 24.37% hit rate is over 1000 hands or so.

On point 2), if you want to debate just how much lower than 32.34% the number should be for a given game texture, fine. I grant you that to get down to the 27% ballpark we'd have to assume a weird game full of quirky rocks.

And I'm not sure that it's a straightforward question of summoning up the energy to "do the math". The calculations are easy; the hard part is figuring out a model of the typical opponent's preflop play that we can all agree on, and coupling it to the original poster's raising / limping behaviour with various ace-and-a-face-card combinations.

[dogmeat]

Guys, thanks for doing the math. I understand that the statistics are certainly flawed by the fact that the only time I can count an ace/face and is when there is a call to my hand, but even so, the numbers seem so improbable now that I have 1000 chances to look at........

Anyway, its probably just a funky streak I'm going through. Maybe this is what happens when bad streaks come upon us and I have just been lucky enough (in spite of missing the flops) to have won at my usual rate, and I like that assumption instead of thinking this has always been the case (and the percentage hit) and my earn would be considerably higher if the flops hit at the expected average.

Dogmeat [img]/images/graemlins/spade.gif[/img]

[BigBlind]

Thanks for the calculations jtr

[James282]

[ QUOTE ]
Hey, I hate to see a respected poster/player like David Ross take such a large hit (-510BB in nine days), but I for one have always believed that swings even out. I would bet a steak dinner for David (and friends/family) in June at the WSOP that he will have a very significant increase in winning percentage by Valentines Day that easily wipes this amount out. I know that makes no sense from a mathematical standpoint, but I'm more on the Doyle Brunson side of this - streaks happen, and they even out.

That said, the upper part is not the part where I could be considered a quack (well, yeah, I guess there too), but....

I posted last month about this and nobody really cared. But what if Party and affiliates did the following:

They juiced the cards preflop for aces and faces. Would this not increase the number of players each hand, and would this not make the number of times a mediocre hand won higher? Especially straights and small connectors?

I know the following stats are not entirely accurate because they (1) only apply to my starting cards (2) Don't include any hands where I did not see the flop.

When I started with an ace and a face card, I expect to at least pair a card 35% of the time. However, for 15,000 hands my average has been 24.37%

That means I have flopped a pair/2pair/trips to the ace and/or face in my hand 25% less than the expected amount - for 15,000 total hands played (not actual hands with an ace/face).

I held the ace/face hand 1087 times (is this amount so high as to prove the hands are hitting more than they should preflop) and hit the flop 265 times for a percentage hit of 24.37% - is this percentage for 1000+ samples so low as to be impossibly small for a fair game? It seems to me, yes.

On the other hand, I'm still winning at my usual rate. But I wonder, if these numbers are too small for chance, would not the lessening of aces and faces on the flop hurt the aggressive players that always pop the pot preflop with these hands? And are not most of the best "pros" aggressive, and don't they like to raise/cap pots preflop with bigslick etc?

Yeah, I know I'm a Quack. Maybe the numbers I am getting are a statistical abberation and over the course of 300,000 hands will even out. I swore off poker tracker a few months ago, and don't have anything else to look at, but

Quack, quack

Dogmeat [img]/images/graemlins/spade.gif[/img]

[/ QUOTE ]

I thought I read somewhere that you play for a living. Is this true? Because if it is, and I'm not trying to be a jerk here, you may be in the wrong profession. The simple fact that you counted how many times you flopped a pair in the last 1000 hands worries me. The fact that you think this indicates anything about PP's random number generator worries me even more. I don't really know what else to say.

-James