The last puzzle I tried to castle but couldn`t.
New Elo Estimator
Sort:
Interesting, the FEN suggests that neither player has castling rights at this point, here is the FEN of that position:
r1bq1r2/1p3pbk/p2ppnpp/2pP4/2P1PPn1/2N1B3/PP1QBNPP/R3K2R w - - 0 1
Although it looks like it should be from the position, I will look for the game to see if I can find it and verify
Jan 22, 2014
0
#4
Good! Some things to add...
1) Let users know where they are at...how many puzzles will there be?
2) Some delay...it is jarring to submit an answer and be immediately face with something new. Maybe a progress bar?
3) An explanation of the methodolgy used.
4) at the end, show the best answers
Thanks for the feedback, those are all very good suggestions that I will certainly work on. Good idea about the prorgess bar. Just quickly for those curious, each puzzle has 4 moves that you can be credited for, and it is also timed. However, if you come up with the best move in every puzzle and take 3 hours, you will still have a very good score. The timed element does not make an astronomical difference, however it should separate (for example) a 1780 player from a 1650 player (that's the idea at least).
Jan 22, 2014
0
#6
how are you determining the approximate rating? Is this your own test, or did you find it somewhere else?
I created this test myself. I don't know of any tests that are accurate at estimating Elo.
@tigerprowl I thought long and hard about that, the test is heavily rewarding to those who do it fast, but it doesn't harshly punish those who take their time
tigerprowl wrote:
"and it is also timed."
I don't think you should have it timed.
In addition, I blitzed it and got 1512 rating. I have no idea if that is accurate to the moves. If you look at the Houdini analysis for the Tatachess tournament it gives a number (.5 or -.5) and lists sequence of moves.
So, when I guess the next move during the game, I can know how it relates to the actual move played. If both the GM and I make the same next move I can know. To me this is more important than an ELO which doesn't evaluate several moves. A true ELO is a rating of you playing a game from beginning to end, not just jumping into a game midstream and playing 1 move and leaving.
Of course, I agree completely. But getting an official Elo also takes months or years! The whole idea of this test is to estimate the Elo, I do believe it is possible to build a semi-accurate (I admit it is not yet).
Jan 22, 2014
0
#9
youngrema wrote:
I created this test myself. I don't know of any tests that are accurate at estimating Elo.
@tigerprowl I thought long and hard about that, the test is heavily rewarding to those who do it fast, but it doesn't harshly punish those who take their time
not meaning to criticize your chess, but how are you qualified to creat a chess quiz that estimates ELO rating?
I posted the solutions, but chess.com screwed it all up (only showed first).
I guess I'll try posting them one at a time...
First move check with computer. Replies aren't always best. In those cases I make simple moves to show the basic idea.
Hey, if you want to solve the test with an engine, I can give the FENs, it may save you time, let me know if this is your desire.
Okay guys, I've been working on this for the past 48 hours and it's finally ready. I expect it to be a bit buggy, please bare with me, but please also do critisize. I expect the Elo to be slightly overestimated right now, but please let me know how accurate the results are that's the most important part. I think it would be cool if I could actually make this accurate and if people dig it, I'll buy a real domain name! Please give it a try and give me feedback, don't hold back if you think it sucks :)
http://rscorecalculator.com/chess/
edit: it may not look great on tablets or phones yet, sorry about that