| |
BGonline.org Forums
Error Rates
Posted By: Stick
Date: Tuesday, 28 April 2009, at 8:42 a.m.
I have a fountain of crap I'd like to share/suggest/put forth concerning error rates. This will neither be well organized or short, so bear with and attack any weak links, there are many questions I have yet had the time to track down answers for and even when I do, it usually sprouts other questions.
My first suggestion in staying ahead of the backgammon community is for us to adopt GNU error rates and let Snowie ERs be a thing of the past. This may be a rocky transition as Snowie error rates are more the native language and GNU error rates are like French in Canada, understood by some but on the whole still a foreign concept.
The biggest advantage I see to using GNU error rates is in determining its error rates GNU includes only the non forced moves and close or actual cube decisions. In other words, if you spend a day on the bar or enter only 1 checker out of 2 from the bar these will not be, and should not be, averaged in to your total decisions for the game/match. Snowie 4 uses all the moves (by both players) to determine your error rate, forced or not.
GNU and Snowie's rating titles and classes shown side by side
An explanation by key members of the GNU team on GNU error rates
I found these in a rather short period of time and would appreciate any other links people are able to provide in this realm. There are still other improvements that I believe should be made to crack down on how error ratings are figured but getting the GNU scale adopted is the first step in the right direction.
People quoting their own error rating
We hear it all the time, "I play a 4", my ass. There are numerous faults in almost anyone who gives an average error rating. The most glaring to me is that it's only the 2 ply (or 3 ply) analysis of the match. Nobody has the time to roll out all their matches (nor the desire probably, who would be that sick?) so if they decide to roll something out what is it? Their errors.
Well that's handy ... roll out only the decisions you got wrong and see where your error rate goes, I'm guessing down nearly all of the time. (it is possible to roll out only the plays you got wrong and the size of the errors increase but this is not likely) Even when someone does roll out the plays they got wrong it's doubtful they roll them out to statistical significance so ... well, we'll let Chuck bitch about that.
For argument's sake let's imagine you have some sick puppy who cares about error rates for an inexplicable reason, he has the time and computer power to roll out all his matches and decides to roll out every play that's close to statistical significance when possible, or at least killing off most of the variance and getting it to where the plays may be 'too close to call', good enough?
No! While rolling out plays once you get a statistically significant result and the play made OtB is clearly correct you can stop the rollout at that time, it doesn't matter for the error rate how far back decision B is. However, if you're rolling out your blunders (of which even the best of us have numerous in any given match) and you roll it out a minimum number of games, let's say 108, and after that 108 games it is statistically significant are you finished? Nope, not if you want to be sickly accurate. You need to overroll the plays you got wrong to know how large your error/blunder is. After only 108 trials it may be a .15 blunder and not the correct play by 3.5 JSDs so you'd think on to the next, but that .15 blunder could be a .22 blunder or a .08 blunder with all the variance that's left in that type of rollout.
What else do people do ... they varnick weak sauced up excuses as to why they may have played poorly in this match or the other and ignore it in their perception of their average error rate. Some of my favorites:
- "We hedged." I used this one recently because in the Super Jackpot in Vegas we hedged the entire pool before the match and I may have played a little slack. Tough shit, I still tried to play a decent game of backgammon and while it was my highest recorded error rate of the weekend I'd never 'throw it out' because there was nothing at stake.
- "I was tired...got to bed late last night...didn't get enough sleep...been playing nonstop" or some other form of "I had sand in my vagina." You know what? I was tired too. I played all damn day and certainly coming to this backgammon tournament fucked up my schedule like you wouldn't believe. I'm used to going to bed at 10am, not getting up and being coherent. I'll bitch about it like the next guy but to use it to defend my sloppy play in an actual match more MCG weak sauce.
- The above excuses in the finals/big matches. It's the finals for the love of fuck!? You bring your A game when the money is on the line. Suck it up, Jordan plays when he's violently ill, Tiger swings with 1 leg, you can get over a little fatigue and play well when it's Sunday.
I had more to say when I started but have run out of steam momentarily.
Stick
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.