| |
BGonline.org Forums
Random thoughts on adjustment of play, PR level and expected win rate
Posted By: Henrik Bukkjaer In Response To: Random thoughts on adjustment of play, PR level and expected win rate (Christian Munk-Christensen)
Date: Friday, 25 November 2011, at 10:18 p.m.
CMC: A couple of thoughts about your data:
The figures you compare are dated many years back. Your's and Michael's win rates are calculated on data going back 12+ years.
Now, while you might have played at the same level (at least error-rate-wise) as Michael for the last 3-4 years, I guess there must have been a big gap on your levels of play going back 5-8 years!! Back then Michael was even more in his own league, and you had just started playing competitive backgammon. 7-8 years ago, there must have been at least 700 places between the two of you on the Danish ranking list.
So I feel your comparing apples to oranges, when you say Michael won so many more matches than me, even though we (now) play at the same level ER-wise.
Furthermore, it's not certain that you and Michael played the same strength when you jousted heads up against each other, and when you played in the Elite division. Michael might have raised his game in the leagues and maybe even relaxed it a bit vs. you, and/or you might have done the opposite. 5-6 years ago, it might have been more important for you to beat Michael in a non-tournament match, than it was for him to wipe the floor with you!! I don't mean anything stupid with this comment, just looking for possible explanations of the "phenomena" you mention :-)
PR is not god, like Timothy also points to. Some players perform when it matters, while other crack under pressure. So if you have a tendency to play very well when things are going smooth, but you start to make errors when the cube go up, or the game gets near to the end, then your low PR will not perform the magic it promises to be capable of!
Other things matter as well. I just made a post on the DBgF forum that you might have read, in which I suggest that certain player types (defined by their error patterns) will fair better against a strong playing field, compared to their relative performance vs. a soft field.
Now, some error types have special effects on ER as well. Eg. if you tend to drop early and/or cash out rather than play on for a gammon, you won't find yourself playing the same number of endgames, and that will probably affect your average error!
There are a lot of scenarios that will work ER and winrate in opposite directions. Here's one I encountered in one of my recent matches. In one game, a player makes a checkerplay error on 3 of his first 4 moves. 2 of them real solid errors, and the last one an outright blunder. Then I double him, and it's a cash - no doubt about it! However, he takes (yet another error) and go on to loose a gammon after many moves. However, he didn't make any real errors after his take! So, what happened? He made an error by taking, it did cost him a considerable amount of match equity, but that error did in fact lower his ER (which would had been sky-high had he passed correctly).
:-)
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.