Elektrik
05-13-2005, 07:49 PM
This whole thing's essentially pointless as it won't change anything, but I thought someone might enjoy it so I'm posting it.
I finally sold out and bought poker prophecy; I figure there are situations where it could definitely come in handy. My first reaction: WOW, they're missing a lot of data. I'd estimate they're hitting 20, maybe 30% of the SNG's played, if even that, and I'm looking at the $200 level. Who knows about the lower ones.
I don't have time to do this, so someone else would have to undertake it, but here's what I propose as an experiment to quantatively test poker prophecy's inaccuracies:
Abstract
To determine both the precision and accuracy of poker prophecy as a data measuring tool.
Methods & Data Collection
Find as many players as you can from all levels; a good sample will include players that play 0 SNG's per week, a few a week, and many a week. Pick a time when none of them will be playing, say 10AM monday morning, and record Poker Prophecy stats for all of them. Have them keep close records of every SNG they play after that time for the following week, two weeks, and/or month. The longer the period and therefore more SNG's played the better.
Results & Discussion
Testing Precision and Accuracy:
Precision is determined by how much variation occurs within the data set. If prophecy records 100% of SNG's played, it is completely precise. If prophecy only records 30%, it is not as precise but results are still relevant.
Accuracy is determined by the method of data collection, ie did prophecy pick up extraneous data? While we cannot determine the exact accuracy as precision will be off anyways, clear examples of inaccuracy would be extra SNG's played, as well as SNG's played at different levels that were never played.
If anyone's up for it, have fun!
I finally sold out and bought poker prophecy; I figure there are situations where it could definitely come in handy. My first reaction: WOW, they're missing a lot of data. I'd estimate they're hitting 20, maybe 30% of the SNG's played, if even that, and I'm looking at the $200 level. Who knows about the lower ones.
I don't have time to do this, so someone else would have to undertake it, but here's what I propose as an experiment to quantatively test poker prophecy's inaccuracies:
Abstract
To determine both the precision and accuracy of poker prophecy as a data measuring tool.
Methods & Data Collection
Find as many players as you can from all levels; a good sample will include players that play 0 SNG's per week, a few a week, and many a week. Pick a time when none of them will be playing, say 10AM monday morning, and record Poker Prophecy stats for all of them. Have them keep close records of every SNG they play after that time for the following week, two weeks, and/or month. The longer the period and therefore more SNG's played the better.
Results & Discussion
Testing Precision and Accuracy:
Precision is determined by how much variation occurs within the data set. If prophecy records 100% of SNG's played, it is completely precise. If prophecy only records 30%, it is not as precise but results are still relevant.
Accuracy is determined by the method of data collection, ie did prophecy pick up extraneous data? While we cannot determine the exact accuracy as precision will be off anyways, clear examples of inaccuracy would be extra SNG's played, as well as SNG's played at different levels that were never played.
If anyone's up for it, have fun!