PDA

View Full Version : Estimating Tournament Win Rates


11-12-2001, 01:23 PM
Greetings,


I am a live action player with approx 600 hours under my belt. I have reasonable win rates in all games that I play. Often quoted is the approximation of "one big bet per hour"; and my experience confirms that it is a win rate to be expected by competent players.


Recently, I started playing in tournaments and I have not seen much discussion on estimating win rates in tournament play.


Do tournament players use the winrate (i.e $/hour) as a measure of success as live action players do? If so, what is reasonable for various tournament structures/buy-ins?


I have attempted to analyze this concept using a % advantage. The case in mention is a single table tournament like the ones found on Paradise Poker. $10 buy in with $1 entry fee. $50 first place, $30 second place and $20 third place.


For example, in blackjack players can achieve anywhere from 0.5% to 2% advantage using a card counting system. Basic strategy yields approx a -0.5% disadvantage. These percentages are quoted for all kinds of casino games. For example, if you have 0.5% EV, for every $100 dollars you wager you expect to win $0.50 in the long run.


Let's try to figure it out for live action play, say $5/$10 holdem. Assuming 30 hands per hour. You will pay blinds approx. 3 times per hour which equals approx. $21 dollars in blinds. A tight aggressive player plays approx. 15% of hands so you might play 5 hands in an hour. Assume you wager on average $35 dollars on each of these hands (NOTE: THIS FIGURE MAY BE WAY OFF!!). Some you will fold on the flop, some will go to the river, some will involve raises etc. That is $175 + $21 wagered in an hour ... about $200.


A competent player wins 1BB per hour which is $10 . Therefore

$10/$200 this translates into a 5% EV. So, for every $100 dollars you wager on the poker table you win $5. I have disregarded the rake in this discussion.


Now let's apply this to tournament play ... in our example you have wagered $11 dollars ($10 plus $1 entry fee). These tournaments usually last 1 hr.

Here is the big assumption: YOU SHOULD HAVE THE SAME WIN RATE IN

TOURNAMENT PLAY AS LIVE ACTION PLAY!!! (OR CLOSE TO IT) ... LET'S

ASSUME YOU ARE PLAYING AGAINST THE SAME PLAYERS AS IN THE LIVE ACTION TABLE.


Therefore 5% of $10 is $0.50. Therefore I expect to win $0.50 an hour (remember this means I get back $0.50 plus my $10 )at the single table tournament in question. A complete waste of time! Playing $5/$10 holdem appears to be 20 times more profitable than a single table $10 tournament.


What is the point of all this?? It is my attempt at getting a grasp on predicting winrates for various tournanement structures ....


I will leave my post at this point since it is getting rather long.

Any feedback, comments etc. are much appreciated. Also, if anyone knows of literature that discusses estimating tourney win rates ,please let me know.


Thanks,

JM

11-13-2001, 09:00 AM
I think there is a fundamental error here. When you calculate your win rate for live action, you use the total money put into action each hour, about $200 in your example. However, when you buy in for $200 and play for 8 hours, you get to use those same chips over and over again, right?


Same thing for a tourney. Just because the buy-in is $10 doesn't mean that's your total action. Your advantage is based upon the chips you put into action, and the dollar value of those chips. Since you'll reuse the same chips over a few times, your advantage can't just be multiplied by $10 to calculate your average win. You need to use some number higher than $10.


An old rule-of-thumb is that a very good tourney player will win somewhere around double the average. Since the average player wins back his total entry fee (minus the vig), the very good player will win back about double. So, if the tourney is $10 + $1, the very good player might win an average of $20, for a profit of $9.


Of course, I've never seen any actual proof of this old rule-of-thumb, it's just thrown around a lot. The variance is so high compared to the win rates in tourney poker that nobody really knows (and may never know) just how well the very good player earns in the long run.


Later, Greg Raymer (FossilMan)

11-13-2001, 10:51 AM
Greg,


When you calculate win rate for live action it is simply in the units $/hour.


My example uses the concept of % advantage. For example if you have a % advantage of 5% this means for every 100 dollars you wager you will get a 5% return . I.e a return of $105 for a profit of $5.


In the example approx. $200 is wagered in an hour and 1BB is $10 which corresponds to 5%.


I see what you are saying but if you use the chips over again it counts as another wager and i figured approx. $200 in wagers in an hour. So what you are saying doesn't make sense for live play.


Where your argument might apply is in tournament play. Depending on what way you think about it.


If the buy-in is $10 , really you have only wagered $10 of your own money in the time taken to complete the tournament approx. 1 hour. In the grand scheme of things, you have only risked $10 ... i.e it is your total action.


You said,


"Your advantage is based upon the chips you put into action, and the dollar value of those chips. Since you'll reuse the same chips over a few times, your advantage can't just be multiplied by $10 to calculate your average win. You need to use some number higher than $10."


I agree with this especially the dollar value of the chips part.


For example at Paradise you start with 800 tournament chips. so at the outset each chip is worth $10/800 or just over 1 cent each.


If you calculated the action this way it would still be very minimal since the dollar value of the chip is so little.


I have not heard the rule of thumb you mentioned before ... that's kind of what I was looking for . That's hard to believe though. Averaging $9 per tournament in a $10 buy in single table event. That's pretty high.


I suppose the only way to know is to keep good records and play a tonne of tournaments.


Later,

Joe

11-16-2001, 07:57 PM
Due to the enormous volatility in tournament results, structures, and payouts, I can't imagine any validity to a dollar "win rate" for a tournament player. However, I do believe that certain players know with a reasonable degree of certainty how often they will win a tourney and how often they will finish in the money. Further, I think the idea of an hourly win rate misses the main point in playing tournaments, and that is to make a knockout score, not grind out a few bucks "per hour".