swami wrote:
Pedro deserves some other engine atleast, so my prize to him is still pending, dont know if its a current engine, havent asked him about it, when he comes across this post,maybe he could reply to your query.
As a price, you could run him about 1000 test games for a beta version of Danasah.
swami wrote:I think number of divisions win approach is the best way to determine the winner, isn't it?
I agree. There's a problem with the average approach, that if I only participate in one division and get a very high score, then I'm better off only playing in that division, since playing further will just decrease my average. Still, the highest average should get a prize
well, I gave contestants an option of taking a pause from the division that they have less knowledge about, anyway how about this idea:
I take 5 best scores out of 10 divisions into account for 1 contestant and calculate the average from that?sounds fair enough,don't you think?
swami wrote:
Pedro deserves some other engine atleast, so my prize to him is still pending, dont know if its a current engine, havent asked him about it, when he comes across this post,maybe he could reply to your query.
As a price, you could run him about 1000 test games for a beta version of Danasah.
I accept the prize
Good Idea,Pedro. I will do the tests on Danasah beta and send you the games and results instead of the prize,is that a deal?
swami wrote:
I take 5 best scores out of 10 divisions into account for 1 contestant and calculate the average from that?sounds fair enough,don't you think?
That's pretty much like picking the winner with most division wins. Taking the average is by far the best method in my opinion, and you could make it less erranious by keeping the number of engines the same.
Tony Thomas wrote:I rest my case that taking the average of 5 best scores would be similar to giving the gift to the one with the most division wins.
What's wrong with giving the prize to the most division wins? It's like in chess.
You don't count the material after the end of the game, so even if I'm winning by a rook and a knight, if you mated me, the score sheet will have a 0-1 there, and everything else is irrelevant. In this case, the final score (In my opinion) should be irrelevant, and what should matter is how many divisions you won.
Your beliefs create your reality, so be careful what you wish for.
swami wrote:
Pedro deserves some other engine atleast, so my prize to him is still pending, dont know if its a current engine, havent asked him about it, when he comes across this post,maybe he could reply to your query.
As a price, you could run him about 1000 test games for a beta version of Danasah.
I accept the prize
Good Idea,Pedro. I will do the tests on Danasah beta and send you the games and results instead of the prize,is that a deal?
The idea is of Tony.
I will send 2 versions, DanaSah 3.46 and DanaSah 3.47 to see which is better.
Tony Thomas wrote:I rest my case that taking the average of 5 best scores would be similar to giving the gift to the one with the most division wins.
What's wrong with giving the prize to the most division wins? It's like in chess.
You don't count the material after the end of the game, so even if I'm winning by a rook and a knight, if you mated me, the score sheet will have a 0-1 there, and everything else is irrelevant. In this case, the final score (In my opinion) should be irrelevant, and what should matter is how many divisions you won.
I was trying to point out that the idea of giving the price to the average of 5 best is no different than giving it to the person with most wins. I have a problem with the number of division wins, because I could come in and get the 2nd spot 8 times in a row and wont get a darn thing.
Tony Thomas wrote:I rest my case that taking the average of 5 best scores would be similar to giving the gift to the one with the most division wins.
What's wrong with giving the prize to the most division wins? It's like in chess.
You don't count the material after the end of the game, so even if I'm winning by a rook and a knight, if you mated me, the score sheet will have a 0-1 there, and everything else is irrelevant. In this case, the final score (In my opinion) should be irrelevant, and what should matter is how many divisions you won.
I was trying to point out that the idea of giving the price to the average of 5 best is no different than giving it to the person with most wins. I have a problem with the number of division wins, because I could come in and get the 2nd spot 8 times in a row and wont get a darn thing.
Well, how about we create a poll on this? what other ideas do you have?
is this enough for a poll question and the options:
Which one do you think is the best way to determine the winner:
-Overall average scores based on all divisions that the contestant took part in.
swami wrote:
Pedro deserves some other engine atleast, so my prize to him is still pending, dont know if its a current engine, havent asked him about it, when he comes across this post,maybe he could reply to your query.
As a price, you could run him about 1000 test games for a beta version of Danasah.
I accept the prize
Good Idea,Pedro. I will do the tests on Danasah beta and send you the games and results instead of the prize,is that a deal?
The idea is of Tony.
I will send 2 versions, DanaSah 3.46 and DanaSah 3.47 to see which is better.
From now on I will try more in the forecast.
Pedro
Well,thanks, you can send these betas to me at nswami15 at yahoo.com
I will send you the games and results within a week or in 10 days.I presume its not too late for the tournament that you are looking forward to.