Online Poker : 75% of loosers

[Onaflag]

Is this the new "Test" thread? Just wondering.

Onaflag..........

[twolf]

[ QUOTE ]
[ QUOTE ]
I read it was 90-95%

I bet there's a lot of players who think they're winners, but have put in hundreds before they got any good, and just 'forgot' about all that expense.

[/ QUOTE ]

Hundreds? How about thousands?

[/ QUOTE ]

For an online player I don't think this is true. Mostly because of the lower limits than B&M and the amount of hands/experience/information you can get. All you have to do is study/practice and have decent comprehension and bankroll management skills. Now if you only try to learn playing B&M it will likely take much much more money to become a winning player

[speirs]

On a side note, you can now officially add me to that 90% I mentioned earlier.

Assuming a 5% rake online (??) That means that every dollar wagered pays out .95. So, to beat this, one must be 1.05263 times better than "average".

Where "average" is 1.0, "winner" is (> or =) 1.053.

Let's consider a "crushing" full-ring limit player is a +3BB/100 player. After 100 hands, a break even (or 1.05) player would have to win 7.5BB just to be at even (you pay 7.5BB every 100 hands). A "crusher" would have to earn 10.5BB over this period.

3/7.5 = .4, which means a crusher is 40% better than a break even player.

So, where "X" is average skill (slight loser), and equals 1X. Break-even is 1.05X, crusher is 1.47X

Among 1000 players, if the 1000th player was a 1.47 player and 500th was a 1.0 player, then about 565 would be a 1.05 player, (if it's a direct relationship ??), which means about 435/1000 players are showing a profit*, or 43.5%, while 56.5% are losing.

*This assumes nobody plays more than 1 table at a time. Since we know that the more likely you are to win, the more likely you are to multi-table, we can easily assume that the winners are much more rare than this. Guessing that the "average winner" plays 3 tables at once, I imagine this makes the winners 3 times as rare, which would bring the number down to 14.5%, which would render 85.5% of online players losers.

This is all quite fuzzy math of course, but hey...whatever. I think the number is probably around 90% in all honesty. No way it's lower than 85%.

[Cincy Peach]

Don't you all think that the number one reason for losing is refusing to change? I have noticed that changing my poker style is like changing your personality - and some people are just not up to it. I am guessing that two-thirds of the players online do not search for and plug holes.

[meow_meow]

[ QUOTE ]
Assuming a 5% rake online (??) That means that every dollar wagered pays out .95. So, to beat this, one must be 1.05263 times better than "average".

Where "average" is 1.0, "winner" is (> or =) 1.053.

Let's consider a "crushing" full-ring limit player is a +3BB/100 player. After 100 hands, a break even (or 1.05) player would have to win 7.5BB just to be at even (you pay 7.5BB every 100 hands). A "crusher" would have to earn 10.5BB over this period.

3/7.5 = .4, which means a crusher is 40% better than a break even player.

So, where "X" is average skill (slight loser), and equals 1X. Break-even is 1.05X, crusher is 1.47X

Among 1000 players, if the 1000th player was a 1.47 player and 500th was a 1.0 player, then about 565 would be a 1.05 player, (if it's a direct relationship ??), which means about 435/1000 players are showing a profit*, or 43.5%, while 56.5% are losing.

*This assumes nobody plays more than 1 table at a time. Since we know that the more likely you are to win, the more likely you are to multi-table, we can easily assume that the winners are much more rare than this. Guessing that the "average winner" plays 3 tables at once, I imagine this makes the winners 3 times as rare, which would bring the number down to 14.5%, which would render 85.5% of online players losers.

This is all quite fuzzy math of course, but hey...whatever. I think the number is probably around 90% in all honesty. No way it's lower than 85%.

[/ QUOTE ]

Your math is all wrong. I don't know where the "paying 7.5BB/100 hands" number is coming from, but this number is highly limit dependant. For example, at 5/10 6max, the average player pays a little over 2BB/100 (tight players pay less because they win less pots).

[meow_meow]

[ QUOTE ]
[ QUOTE ]
Interesting, what kind of datamining are you talking about? I would think you would need several thousand or tens of thousands of hands on an opponent before you can classify him as a winner/loser with a high degree of confidence, and I'm fairly sure most PT dbs out there do not have a lot of opponents with this large number of hands. My PT db is pretty small, so I'm just speculating here.

[/ QUOTE ]

Do a search on the forums for datamining threads, there are many. Some people are pulling down 500,000-1 million hands+/month.

[/ QUOTE ]

Here is the fallacy in your argument.
Nobody is capturing every hand that every player plays.

When I look at my own 5/10 database, I have an average of 55.3 hands per player. Only 57.7% of the 10816 players in my db are losers.

I'm guessing that the average number of hands/player in even the largest mined db isn't anywhere close to 900, but lets use that as an example.

Lets say that we have a huge population of players, 90% of whom have a true winrate of -2BB/100, and the remaining 10% have a true winrate of +2BB/100 (lets also say everyone has a reasonable SD of 16BB/100).
Now, lets observe each of these players over a 900 hand stretch. What percentage of these players will be losers over 900 hands? Just 61.8% , even though 90% of them are long term losers.

What if we watched each player for just 100 hands? 54% would be losers.

Even if we were able to watch each player for 2500 hands, only 68.9% would be losers over that stretch.

[BarronVangorToth]

Another factor to consider when people provide their opinion of how many are winners...

Imagine if I went to a game and raised and reraised any time I could regardless of what my cards were.

And then imagine if I did this for 100,000 hands.

Do you think most of the people I will have played against would be winners?

Does that make everyone winners?

Is the "X-Factor" (i.e. the person giving the information) relevant?

When one person says 10% and another 20% -- are we assuming they are equal?

That they can read the data the same?

That there aren't other factors?

Barron Vangor Toth
BarronVangorToth.com

[Sniper]

[ QUOTE ]
Here is the fallacy in your argument.

[/ QUOTE ]

The statement you quoted was a fact, not an argument, so I'm going to assume for purposes of this comment that you were refering to my earlier statements regarding the 40%win/60%lose rate for online ring game players.

[ QUOTE ]
Nobody is capturing every hand that every player plays.

[/ QUOTE ]

For starters, the poker rooms themselves have all this data [img]/images/graemlins/smile.gif[/img]

Secondly, sites like poker edge and poker prophecy are datamining almost all hands played.

Third, there are several others datamining more than enough data to make their observations relevant.

Fourth, you can also draw conclusions from the aggregation of many players relatively smaller sample sizes.

The data in all cases points to the same roughly 40/60 split for ring game players!

[ QUOTE ]
When I look at my own 5/10 database, I have an average of 55.3 hands per player. Only 57.7% of the 10816 players in my db are losers.

[/ QUOTE ]

Thats close enough to 40/60.. see [img]/images/graemlins/wink.gif[/img]

[ QUOTE ]
I'm guessing that the average number of hands/player in even the largest mined db isn't anywhere close to 900, but lets use that as an example.

[/ QUOTE ]

I've already covered that... there are others that have compiled substantially more data than you.

However, lets take a closer look at the average ring game player...

The average ring game player... is a "recreational" player, losing roughly -2BB/100 (mostly attributable to rake), and plays somewhere around 1,000 ring game hands/month.

Its probably worth noting at this point, that this is a relatively insignificant amount of money. In fact, the vast majority of ring game players neither win nor lose an amount that they would consider significant.

[ QUOTE ]
Lets say that we have a huge population of players, 90% of whom have a true winrate of -2BB/100, and the remaining 10% have a true winrate of +2BB/100 (lets also say everyone has a reasonable SD of 16BB/100).

[/ QUOTE ]

You have just created random numbers... try doing the statistical analysis on factual data.

[ QUOTE ]
Now, lets observe each of these players over a 900 hand stretch. What percentage of these players will be losers over 900 hands? Just 61.8% , even though 90% of them are long term losers.

[/ QUOTE ]

Play around with the numbers a bit and you'll see that it really doesn't matter if any particular player is a winner or loser, that in the aggregate the 40/60 holds up.

For statistical purposes, try assuming that the BB/100 "spread" between a top 10% player and a bottom 10% player is relatively small around a -2BB/100 average, and see how the numbers look then [img]/images/graemlins/wink.gif[/img]

Wonder how many of the losing players quoted are people who only played for a day and then quit after failing to get the sign-up bonus.