Research Study -- Participants Wanted

jcm4ccc · #1 10-18-2005, 02:17 PM

The latest theoretical discussions made me think about this. This forum should be a good source for collecting data about how much a chip is really worth in an MTT. So I'm asking people to participate in a research study.

I will focus solely on the Party Poker $40K guaranteed, in order to reduce variance associated with different sites and different tournaments. The study will run until December 31, 2005. Participants will enter the following information in an Excel worksheet that I created:
<ul type="square">[*]Date of Tournament[*]# of Entrants in Tournament[*]In what place did you finish?[*]Gross Dollars Awarded[*]# of Chips after 15 minutes (end of 1st level)[*]# of Players after 15 minutes (end of 1st level)[*]# of Chips at 1st break[*]# of Players at 1st break[*]# of chips at 2nd break[*]# of Players at 2nd break[*]etc etc etc[/list]
I tried to make it as simple as possible, so that people would find it easy to participate. With this kind of data, we could figure out a person's M or Q at certain points in the tournament, and see how that relates to a person's place in the tournament. For example, what does it mean when you have twice as many chips as the average player at the first break? Are your expected earnings doubled, or something less than that? Does increasing your chip stack increase your chances of winning in a linear way, or in some other way? etc etc etc. Of course the study will be biased, since the participants will probably be better than the average player. But I still think the data will be meaningful, though it may not generalize to the average player or to different tournaments.

If you participate in the study, I will send you a copy of the raw datafile before I do any work on it myself. That way, you can run your own analyses if you like, and share them with the forum.

If you are willing to participate, please do the following:

1. Post a reply to this message (this will encourage other people to participate).

2. Send an email to uncpsyc28@nc.rr.com with the following subject line: 2+2

I will send you a zipped file of the Excel worksheet to fill in your data from now until December 31, 2005 (the end of the study). Hopefully enough people will participate that we will get some meaningful data. I would say we need a minimum of 50 people to make it worth the effort.

SumZero · #2 10-18-2005, 02:20 PM

I would love to see the data and the results for this. I would love to participate but I only play on UB and only in smallish buyin ($30 and less, usually $5 and $10) tournaments so don't know if that helps.

But if it turns out there aren't enough from one source I'm willing to throw my hat in the ring.

One other thought would be to try it with 3 table sit and goes. As there we may be able to get significant data quickly in a format that is a very small MTT.

#3 10-18-2005, 02:23 PM

Thanks a lot for doing this.

I hope you get a huge sample size.

AtticusFinch · #4 10-18-2005, 02:31 PM

Good stuff. I hope you collect a ton of data. I also hope you're willing to share [img]/images/graemlins/wink.gif[/img]

For my part, I play the 40k occasionally, but I've been lazy about keeping records lately. I'll send you my data henceforth, though.

zoobird · #5 10-18-2005, 02:50 PM

Maybe I'm missing something since I don't play on Party, but can't you view that data for ALL tournament participants at any time during the tournament (or any tournament, on any poker site)? If so, I would think you could do some sort of 'screen scraping' to get much more data, much faster at whatever intervals you need.

AtticusFinch · #6 10-18-2005, 02:57 PM

[ QUOTE ]
Maybe I'm missing something since I don't play on Party, but can't you view that data for ALL tournament participants at any time during the tournament (or any tournament, on any poker site)? If so, I would think you could do some sort of 'screen scraping' to get much more data, much faster at whatever intervals you need.

[/ QUOTE ]

The trouble is that the results wouldn't be independent, then (one player's result in a single tourney is obivously dependent on the other players' results).

To make a good statistical analysis, you need to have independent data.

jcm4ccc · #7 10-18-2005, 02:57 PM

[ QUOTE ]
Good stuff. I hope you collect a ton of data. I also hope you're willing to share [img]/images/graemlins/wink.gif[/img]

[/ QUOTE ] Definitely willing to share. Will send the raw datafile to all participants after Jan. 1.

zoobird · #8 10-18-2005, 03:54 PM

I think you're oversimplifying and getting this wrong. If you look at the questions OP put in his original post, a full set of data for a tournament is going to provide much more meaningful data than just gathering from a few people.

#9 10-18-2005, 04:15 PM

[ QUOTE ]
I think you're oversimplifying and getting this wrong. If you look at the questions OP put in his original post, a full set of data for a tournament is going to provide much more meaningful data than just gathering from a few people.

[/ QUOTE ]

I have yet to see tournament data that gives chip counts for every player at each break, as well as the other data the OP is asking for such as number of players remaining at different intervals. Full tournament histories from PP can tell you what place everyone finished. It does not list everyone's chip counts at each break. This is where your idea is flawed as this data can only come from individual players.

KneeCo · #10 10-18-2005, 04:31 PM

But couldn't one just get it from the lobby?

By clicking on the tables individually and take a screen shot.

One could do this at every break and then take the final standings and match names. It's a lot of leg work, but it offers a big sample size and one that isn't biased by being made up entirely of brown trouts (unless of course the intention of the study is to be limited to 2+2ers).