Winning live player yet?

[steamboatin]

Statking keeps me honest.

[TappedOutAgn]

This is a simple statistics question.

If winning nights go to $300 in 4 hrs, sounds like your standard deviation for these games is around 20 BB/100. So 86 hours -> 1600 hands, std will be 80 BBs which is up to $800. If you won $1500 that means you're 3% likely to be a losing player who got lucky and 80% likely to be at least an average 1BB/hr earning player. An average 1BB/hr player has a 20% chance of being this lucky over 1600 hands.

[CrashPat]

you are running good.

Wait until you have a month where you are doing everything as correctly as you have the discipline to do and you still lose. The draws never come in, when they do the redraw comes in for the other guy stuff like that. Have every big pair drawn out on by 72o, and flopped broadways beat by running dueces to make a guys boat. That's usually called running bad, and even if you can competely avoid any kind of tilt you lose a lot of BB before the deck turns around.

I just play the game, keep good records of time played and money made / lost, and get hundreds of hours logged. Then every once in a while I see where my winrate is, remember that I only have several hundred hours logged and the sample size is way too small, and keep playing.

[A_C_Slater]

I will share with you my 167.5 hour B&M sample size just to show you how fast a good run can get bad in a hurry.

5-10 limit hold em dates 10/28, 2004 thru 2-12, 2005 (I haven't played 5-10 live since, and here's why.)


10-28 to 11-13; 29.5 hours +30.5 BB

11-16 to 12-13; 36.5 hours +55 BB

12-16 to 1-13; 39.25 hours +31 BB

1-14 to 1-29; 37 hours +68.5 BB

2-1 to 2-12; 29.25 hours -123 BB



I was begining to think I was invincible and then BAM!! 123 BB downswing, nearly all my "work" gone in 29 hours. I haven't been back since, except to play the 1-2 NL for fun.

I decided I needed to practice more and have been playing 3 tabling the 2/4 on Paradise ever since.

At most I played 5025 hands in this sample size (167.5 x 30) and I only say 30 hands an hour because of the shuffle machines.


Expert players can breakeven for 30,000 hands at this game. And they can easily be losing 100BB+ after 15,000 hands.

I just tell you this to prepare you for the worst.

[BWillie]

Thanks TappedOutAgn that was what kind of response I was mainly looking for. That gives me some hope knowing I only have a 3% chance of being a losing $2/$5 player. I can handle being an average 1 bb poker player if it makes me about 7 dollars an hour while I can do something I enjoy [img]/images/graemlins/smile.gif[/img]

As for padding numbers, I haven't padded numbers at all, and at 86 hours it's easily a believeable number, I just wanted to know if I'm on my way or not to possibly becoming a winning live poker player. My poker pals keep me honest from padding numbers, they watch my chips like a hawk to see if they come up more money than me that night so they can brag about it the next day or make fun of me for doing something stupid.

[SamIAm]

Tapped, this is a good post, but I don't think the stat question is quite as simple as you suggest.

I agree with your SD suggestion. I haven't played the .5/1 game online in forever, but I beat that game for 5.5BB/100 over a few thousand hands. (I haven't seen a low spread-limit game that was any harder than .5/1. Certainly Mississippi $4/$8 spread was looser than Party .5/1. YMMV.) Anyway, PokerTracker says my SD was 19.7BB/100hands, so I'm willing to assume BWillie's is similar.

Here's my problem with your post. (And the part that BWillie latched onto.)[ QUOTE ]
If you won $1500 that means you're 3% likely to be a losing player who got lucky and 80% likely to be at least an average 1BB/hr earning player.

[/ QUOTE ]
You're using the outcome to make predictions about his true winrate, which is NOT the same as using his win-rate to make predictions about the outcome. For instance, I totally agree with your calculation when you said[ QUOTE ]
An average 1BB/hr player has a 20% chance of being this lucky over 1600 hands.

[/ QUOTE ]
In the first quote, you're not using Bayes's Theorem correctly. You're making an implicit assumption that all winrates are equally likely. If that were true, then whatever winrate gives the outcome with the highest probability would be the most likely guess for BWillie's winrate. However, not all hidden winrates are equally likely.

Here's an easier example with 2 states instead of a range of win-rate states:

There's a certain disease among a population of 1million. Say 1thousand people are carrying the sickness without symptoms, but we have a pretty acurate test. The test recognizes sick people 99% of the time, but gives false-positives 1% of the time on healthy people. You pick a random guy from the population and his test comes up Positive. Do you think the guy is sick?

Well, being sick would give the Positive result more often than being healthy would, so we're tempted to guess that he's sick. However, Bayes Theorem says that
P(sick)= P(sick & recognized)/P(recognized) = (.99*1000)/(.99*1000+.01(10^6-1000)). That's about 1%. (The guy's not sick.)

I don't know how to do the calculation if we don't know the distribution of sickness (or winrates). I'm happy saying "If you had winrate X, you'd have your outcome with probability Y." However, you're not allowed to say "Given you have that outcome, you have winrate X with probability Y."
-Sam

[Al_Capone_Junior]

Just to give some idea here of a "normal" real life result, let's look at my own results (non-mathematical analysis here) from 2001, where I played almost exclusively 3-6 and 4-8 full kill games in san diego cardrooms...

From Jan-march, nothing special, I won a little.

From March-mid april, you'd think I was a Super-Genius, I was KILLING the games. My top two biggest wins ever in money, and my second and third biggest wins ever in big bets. I had the midas touch when it came to hold'em.

Till about september, I won some, lost some, made a little here and there, but nothing really notable either way.

September was where the dog-doo sessions dominated my results. All the way through mid-october I was sucking major ASS all the way to the poorhouse. I had almost entirely significant losing sessions, extremely BAD losing sessions, and very few, and very small wins.

Finally, near the end of october, I started winning again.

By the end of december, it was time to tally the year's results. I was winning $7.12 per hour in 3-6 and $9.35 in 4-8.

I played almost a thousand hours that year (yes, I lived in cardrooms that year).

This is quite the realistic year of poker. Each year I play, it seems to go about the same. Even the period of the last month, I saw three weeks of mostly losing, major suckouts, and a need for extra tubes of preparation-H, until this last week, when I had several really nice wins, even a small "winning streak." I haven't made much money playing poker the last month, but I sure took some big swings.

Realism in poker is everything. If you're not realistic, you're just like all those magoos who's money you're trying to take. They fool themselves, and so will you.

al

[youtalkfunny]

[ QUOTE ]
I say 1 bb/hour because that is a very REALISTIC win rate. I say that because I did all those things, choose good games, avoid strong opponents, play at times when the morons are out etc etc etc and after playing THOUSANDS of hours, my win rate in california at 3-6 kill was $7.12 hour. In 4-8 kill it was $9.35 or so.

[/ QUOTE ]

Al, I think you need to consider the possibility that there are players out there who are:

--much better than you;
--able to find even softer games than the ones you played in.

When I lived in Vegas (back before the current poker boom, which is packing the low-limit tables with morans), I only played on Fri and Sat nights. After 1,000 hours, I was still making better than 2BB/hr at $1-4-8-8 hold'em. It was $18 and change per hour, which would be 2.25BB/hr.

I haven't kept records since, so I can't tell you how I've done since. Sure, that hourly number is more likely to have dipped, than risen. But the fact remains that 1,000 hours is not an insignificant sample size.

Just because you make 1.2BB/hr, that doesn't make it the most that anyone in the world could possibly make.

[BoxTree]

If you really think you're a winning player...maybe you are.

The question is, "How winning is winning?"

Fact: $4 rake at 2-5 SLimit is huge.

Fact: $4 rake at 40/80 is not so huge.

Fact: 2-5 SLimit has worse opposition than 40/80.

Are the players at 2-5 so much worse than those at 40/80 that the huge rake at 2-5 is offset and you can actually earn more than 1 BB/hr?

Sure.

But it'll take a filthy number of hands for you to figure out exactly how much you're really earning.

Just assume you're beating the game for 1 BB/hr and everything else is just icing on the cake. That way, you won't be pissed when you only make 1.5 BB/hr.