ITM rate is only loosely correlated with ROI in large field tournaments.

The problem with finding a larger sample is that people who have, for example, played almost every event over the last 5 years have probably had good results and run over EV. It's inevitably going to suffer from a self selection bias.

Not sure how accurate it is but if theres no data elsewhere then Sharkscope has been tracking live events for 2-3 years including losses so you can get quite a lot of 2011 and 2010 wsop data there. Maybe ppl who played a lot of events can chime in with whether it accurately relfects their schedule for those years.

one drop added to stats? thanks for your time

I updated the spreadsheet on Google Docs to account for this missed cash of $13,031.

New ROI: 30.77%

It correctly reflects my record over the past three year, although I only entered one event in 2010. Also, it incorrectly describes some of the event types (recording a Limit O8 tourney at PL O8 and a Limit HE tourney as NLHE).

Yes, that's why you need a selection criteria that selects players before the results you are measuring are known. E.g. if you want to test results of "top players" from 2011, you could select based on top money winners in 2010, but not top money winners in 2012, since players who won a lot in 2011 probably played more (and thus won more) in 2012 as a result.

"Most Event Played without a cash: 24, Alexander Kostritsyn"

Yet in the pics it states he has played 0 events. What is this, what is this shiit?

Many mistakes have been pointed out, is this even close to being correct?

on the side note, glad to see Arieh down. He always seemed like a huge douche.

If you bothered to read the thread, you'd see that the Krostritsyn error was acknowledged and corrected on the spreadsheet. I couldn't edit the into as it was past the 30 minute window allowed. Many mistakes here, I don't think so. A few, yes. If any more are picked up I will update the spreadsheet on google docs. Please point them out rather than just bitching and complaining. It would be much more productive to the discussion. Thanks.

oh, OK. Very sorry, didnt mean to put you down or anything. Just noticed it right away and kinda figured it would have been corrected if it was noticed before. Silly me, of course it was noticed.. such an easy mistake to spot I knew there is a 30 min limit but I kinda thought mods would have done something to it since this thread has been on twitter and some poker news sites as well.

Made too many assumptions I guess. Just wanted to point out that too many found mistakes kind of kill the whole point of this, since it makes it rather unreliable. But with some fixes it can be a very interesting table. Again, sorry if you took my message as bitching. That was not my purpose.

No one has given this any love, but this is an amazing amazing reference.

I would suggest the investment is understated and ROI overstated. If it were a passive investment such as stocks or bonds, then the buy in would reflect the I in ROI. But this I includes significant amounts of labor hours, skills development, possibly travel. Travel has been mentioned and discarded, and skills development would be difficult to quantify but an important component. I have reflected skills investment in the labor rate proposed below.

But I think we could all agree on the labor piece.Real, tangible hours are invested in the tournament itself.

There were over 300 cashes, and lets assume the labor investment per cash was 20 hours.

There were 1,770 non cashes lets estimate at 8 hours each. Most people would not go to work and bill less than 8 hours.

Total labor investment 20,160 hours.

i would say a fair labor rate would be $60 per hour pre tax given the degree of skills investment required. As pointed out, many would never consider providing coaching for many times that labor rate.

This adds $1,209,600 to the I side, reducing the ROI to slightly less than 17%

Note that the buy ins, assuming you put up your own cash to make it simple, must come from prior post tax tournament income, net of deductible losses,so gains are frequently recycled through taxable events.

Contrast this with a passive investment like a stock, where one can let the gains "ride" for years as there is no forced cash out like a tournament that causes a taxable event. Also the tax rate on long term gains is dramatically lower than fully reported tournament earnings.

Great, now we are going to get tons of meaningless calculations used to draw conclusions from this meaningless set of data, horray.

So you are saying your ROI is lower if you pay yourself $60/hour and subtract this from the figure?

This seems like a senseless calculation. It would be fair to quantify this either in terms of ROI or in terms of hourly rate, but not one subtracted from the other.

Obviously, a poker player is going to have to "work" to earn money. It's not like they have any expectation that money will just be handed to them while they wait for it. The viability of poker as a profession is in how much you get paid for that work. Under your calculations, the averaged-out player is making about $150/hour. This seems like a pretty damn good result.

I believe Taleb would disagree with you, and there are plenty of his fanboys on this site. The size of return absolutely does matter--asymmetry in payouts.

In your example, what it would likely imply is "don't register for an event that could wipe out 1k worth of tourney winnings unless the price was right and you appreciate the risks". It would be like asking for a black swan or betting it all on the russian ruble. An investor that had ridden along with you to that track record might rightly balk at the concept of buying into the One Drop, no matter how great a player you were. (Much respect to your real-life skills.)

+1

Pure gold

You're right, I stopped at masters degree.

FWIW, I've done analysis on this before and that's why I say 3-4k should be enough to get an idea. We're not trying to pin down ROI to +- 1% at .05 alpha. Best approach is probably to just bootstrap the empirical distribution. I'm not going to generate any estimates because I really don't care that much--maybe somebody else will.

It might be somewhat informative to estimate the coefficient of variation (probably using buyins/game) to get an idea of how much signal/noise we are dealing with compared to cash, for instance.

Also, I'd be interested to hear Levitt respond to a counterargument of a hypothetical tournament where the participants are all very-highly, yet equally skilled. I.e., ranking players a priori makes little sense because they would all be tied. Using this method to "show" that poker is a skill game doesn't work here, even though the players could all be Nash unexploitable bots.

Not sure if has been said, but Daniel Alaei didnt cash six times. He cashed twice. He was on my fantasy team.

Seems like you made quite a few blunders man.


Good job though. Good effort.

http://pokerdb.thehendonmob.com/player.php?a=r&n=25308

What is Russian ruble?

The purpose of an ROI calculation is to compare returns across various investments to rank and rate them. The OP calculated an ROI, not an anticipated "hourly wage" as a result of playing poker, and for good reason. Hourly wage calculations are meaningless as it is required that you invest significant sums of cash side by side with your time, unlike a job.

The investment part of the calculation must include some estimate of what your time is worth otherwise it is not comparable to alternative investments. For example, assuming I could earn $60 an hour, an alternative investment would be to invest $10,000 in a stock portfolio, and separately work for pay.

Small business owners and contractors who build homes on spec often overlook this addition to the investment portion of the equation, as both of these consist of cash and time investments. Here is a thread from a small business owner on exactly this point

http://www.captureplanning.com/articles/79375.cfm

Also ROI should be related to degree of risk assumed, and standard deviation of that risk. This is called the efficient frontier. The greater the degree of risk of capital the greater the return. It would be interesting to know the standard deviation of returns across this group of players, but as the spreadsheet is view only, I am unable to calculate that.

Calculating the true value of investment, including the value of your time, and then comparing this ROI to alternative investments with the same degree of risk, I would say it is a relatively poor investment for players of even this caliber.

In mother Russia, Ruble spends you!

Thanks for this catch. Google Doc spreadsheet has been updated to reflect this.

New ROI: 30.19%

I have posted the spreadsheet up at:

https://docs.google.com/spreadsheet/...mNTBzSUE#gid=0

To be honest, I'm not exactly sure how and what you can do with the spreadsheets on Google Doc, but perhaps you can work with the data from there. It would be great if we got some new prespectives on how to interpret this data (even if it is a small sample size).

Ultimately, we are using this to determine how much money top professional poker players make by playing tournament poker. If you are assuming an hourly wage for professional poker players, and subtracting it, you are completely begging the question.

I guess if your question is, "How much more money do poker players make by playing tournament poker than they would make in some other choice of career?", then it is appropriate to subtract whatever money they would have made in this other career as an opportunity cost. But then your method of assuming $60/hour is far too simplistic. It seems more useful to figure out what they made from tournament poker, then figure out what they would have made from some other choice, and compare them independently.