I agree! But this has been brought up before and seems unlikely to change in a hurry.
Best win stats
Sort:
I agree withjonny jupiter
Problem is known but the adjustments required to programming are very problematic-best at present to be aware but philosophical-and few " best wins" are frequently repeatable,they otherwise would equate to your rating; I suggest.
NB Final phrase omitted-"I suggest a look at the difference would be informative !
Nov 3, 2008
0
#6
ManicDragon wrote:
For your specific example, I think if a 1400 beats a 2000, then the higher rated player probably shouldn't have been that high in the first place, thus the adjustment based on their most current estimated rating.
Overall, the most recent rating is the most accurate, so it makes sense to factor in the win/loss that person just had.
Besides, who really cares about the best win? It's online chess with inflated ratings -- a 2000 might only have an ELO of 1600 for all we know.
You know how it goes - someone plays out of their skin against a higher rated player and the 2000 guy isn't fully concentrating while making a move - oops, missed that one. Mr. 1400 deserves this to be recognised - he beat a 2000 player.
I think most people care about their stats, whether they claim to or not. You claim to be very interested in stats elsewhere on this site. What particular elements of the stats are important to one person won't be important to someone else and vice versa. If any stats exist, the more accurately they reflect the information they are trying to represent, the better.
It is hard to say. My first impulse is that it should be before the adjustments.
On the other hand, I have noticed that depending on the ratings that people tend to play, you get a big difference in actual skill level with 2 different people of the same rating (this can depend on the average rating played). Looking at it that way I might tend to disagree.
Another thing is I am not sure about how things work with a rating that has not yet settled down for someone who has not played that many games yet.
There is a lot to think about, and that is probably why it was never changed having been talked about before. I like to use my rating as feed back for how I am playing. Other than that it is hard to say how accurate the ratings are.
Just thinking outloud with a ... variable opinion :)
It struck me as a touch unfair that the best win/worst loss stats take effect after the rating adjustment.
If a 1400 player beats a 2000 player then the 2000 player's rating goes down to about 1940 and the 1400 player's stats register that his best win is 1940. It's not! His best win was against a 2000 player!
The reason I think it isn't entirely fair is that if a 1900 player beats the 2000 player then the 2000 player's rating goes down to about 1970, so even though it wasn't as big a deal for the 1900 player to beat the 2000 guy, the stats make it look better by saying his best win was against a 1970 player.
I think the best win/worse loss stats should reflect what the player's rating was just before the adjustment is applied, not what it is after they lose to you.