Mate in 2 after white's blunder Rxa2 (would be Mate in 6 otherwise), from game http://www.chess.com/echess/game?id=58265910
1. Rxa2?? Nh3 2. Ra6 Rg3# 0-1
Puzzles needed for tactics trainer
Sort:
Sep 12, 2012
0
#503
bobes wrote:
Mate in 2 after white's blunder Rxa2 (would be Mate in 6 otherwise), from game http://www.chess.com/echess/game?id=58265910
1. Rxa2?? Nh3 2. Ra6 Rg3# 0-1
While that's the fastest mate 1. Rg3+ Kf2 2. Nh3+ Ke1 and g1=R or g1=Q are both forced mate in 3 making the problem too ambiguous.
TheIronDuke wrote:
bobes wrote:
Mate in 2 after white's blunder Rxa2 (would be Mate in 6 otherwise), from game http://www.chess.com/echess/game?id=58265910
1. Rxa2?? Nh3 2. Ra6 Rg3# 0-1
While that's the fastest mate 1. Rg3+ Kf2 2. Nh3+ Ke1 and g1=R or g1=Q are both forced mate in 3 making the problem too ambiguous.
apologies, now I re-read the original post and noticed this is not allowed. I've seen quite a few puzzles like that in TT but I understand why it's not a good thing.
What would be the reasonable difference between best and second best move to make the puzzle not too ambiguous? I guess if the second best move was Mate in 20, it would be enough but that's an extreme. E.g a slight variation makes the second best move in this puzzle Mate in 4. Is that enough or not?
Mate in 2: 1. Rxa2?? Nh3 2. Ra6 Rg3#
Mate in 4: 1. Rxa2?? Rg3+ 2. Kf2 Nh3+ 3. Kf1 Ne3+ 4. dxe3 Rg1#
A fork from game http://www.chess.com/echess/game?id=58122382. Modified to prevent alternative lines in second move.
1... h6?? 2. Nxe7+ Kh7 3. Nxc6
Sep 12, 2012
0
#506
bobes, unless I'm missing something I think you problem is even better unmodified. I don't see an equivalent second move, as after you capture his second bishop his queen is threatened and he has no time to take your bishop. I guess I should plug it into the computer to make sure.
TheIronDuke wrote:
bobes, unless I'm missing something I think you problem is even better unmodified. I don't see an equivalent second move, as after you capture his second bishop his queen is threatened and he has no time to take your bishop. I guess I should plug it into the computer to make sure.
this is the situation from the original game:
I checked with stockfish, the first move after h6?? is clearly Nxe+7 but after Kh7, Nxc6 and Qd3+ are almost equaly good moves.
If black played Kh8 instead, then white could play Ng6+ instead of Nxc6. The difference between those moves according to stockfish is around 2-2.5 (depending on depth). If that is enough, then it could be a viable puzzle
katonas wrote:
I think its 1.5, but have no idea how to do what you ask. Perhaps you could ask for help from the chess.com community. Please don't use this thread for discussions about tactics trainer, chess.com or Houdini as there are enough pages for me to navigate through already.
I checked the mate in 9 puzzle again. I verified it with Rybka 2.2 , Houdini 1.5 and I got the puzzle from the King v3.23 Engine appr. 2850 Elo... all say mate in 9 and offer no alternate. That takes away any ambiguity, in terms of alternate mates. I think that would be a really good puzzle for the 2800+ T.T. users. One of them complaining was the reason I submitted that puzzle. Thanks again for your time and interest.
It is puzzle #466 on page 24
The source game for this problem is: http://www.chess.com/echess/game?id=26030713
White is to move and win.
Problem ID: 0474799 needs to be taken off..comfirmed
difficulty: hard
Find best move for white.
Reza_R wrote:
difficulty: hard
Find best move for white.
seems like a8=Q+ and a8=B+ are equally good, otherwise a beautiful puzzle
nameno1had, the rules for this thread are written in the first post. They include the lines :
'Examples not allowed :
solution move forced mate, next best move mate in more moves
solution move forced mate, next best move greater than +1.0'
hessmaster wrote:
what about my puzzle on page 26?
When you post a puzzle you must say when the puzzle starts and ends.
katonas wrote:
nameno1had, the rules for this thread are written in the first post. They include the lines :
'Examples not allowed :
solution move forced mate, next best move mate in more moves
solution move forced mate, next best move greater than +1.0'
I'll admit, I didn't look specifically at that, but why is that considered too ambiguous ? Once again I have to ask the question, aren't tactics trainer puzzles supposed to prepare you for real game situations ? Shouldn't that number for the next best move be consider too close for those who are really adept at tactics already? Afterall, I am trying to submit a puzzle to stump a 3000 rated tactics trainer user, not give a false rating to someone to puff their chest out in pride, or confuse a 1400 player with whether it is a potential mate or winning an exchange ...
I really think that this needs more carefully weighed, for guidelines in choosing puzzles.
nameno1had wrote:
katonas wrote:
nameno1had, the rules for this thread are written in the first post. They include the lines :
'Examples not allowed :
solution move forced mate, next best move mate in more moves
solution move forced mate, next best move greater than +1.0'
I'll admit, I didn't look specifically at that, but why is that considered too ambiguous ? Once again I have to ask the question, aren't tactics trainer puzzles supposed to prepare you for real game situations ? Shouldn't that number for the next best move be consider too close for those who are really adept at tactics already? Afterall, I am trying to submit a puzzle to stump a 3000 rated tactics trainer user, not give a false rating to someone to puff their chest out in pride, or confuse a 1400 player with whether it is a potential mate or winning an exchange ...
I really think that this needs more carefully weighed, for guidelines in choosing puzzles.
The typical tactics trainer is not rated 3000.
Anyways, I think this should be in a seperate thread....
iixxPROxxii wrote:
nameno1had wrote:
katonas wrote:
nameno1had, the rules for this thread are written in the first post. They include the lines :
'Examples not allowed :
solution move forced mate, next best move mate in more moves
solution move forced mate, next best move greater than +1.0'
I'll admit, I didn't look specifically at that, but why is that considered too ambiguous ? Once again I have to ask the question, aren't tactics trainer puzzles supposed to prepare you for real game situations ? Shouldn't that number for the next best move be consider too close for those who are really adept at tactics already? Afterall, I am trying to submit a puzzle to stump a 3000 rated tactics trainer user, not give a false rating to someone to puff their chest out in pride, or confuse a 1400 player with whether it is a potential mate or winning an exchange ...
I really think that this needs more carefully weighed, for guidelines in choosing puzzles.
The typical tactics trainer is not rated 3000.
Anyways, I think this should be in a seperate thread....
Where do you propose is the best place to discuss if there should be a uniform ambiguity factor for puzzles for 400-3000 on tactics trainer ?
Part of being 3000 as compared to 400 is that you can much more easily discern the right answers, inspite of there being a more difficult ambiguity factor potentially between puzzles. I realize that the T.T. puzzle moderator doesn't necessarily decide these things, but if a person wanted to get chess.com to consider any variance as the rating goes up, I think having the moderator's backing, or more specifically the prompting it would be best.
@ Katonas
Where would be the best place to discuss this or bring it to Chess.com's attention...
nameno1had wrote:
iixxPROxxii wrote:
nameno1had wrote:
katonas wrote:
nameno1had, the rules for this thread are written in the first post. They include the lines :
'Examples not allowed :
solution move forced mate, next best move mate in more moves
solution move forced mate, next best move greater than +1.0'
I'll admit, I didn't look specifically at that, but why is that considered too ambiguous ? Once again I have to ask the question, aren't tactics trainer puzzles supposed to prepare you for real game situations ? Shouldn't that number for the next best move be consider too close for those who are really adept at tactics already? Afterall, I am trying to submit a puzzle to stump a 3000 rated tactics trainer user, not give a false rating to someone to puff their chest out in pride, or confuse a 1400 player with whether it is a potential mate or winning an exchange ...
I really think that this needs more carefully weighed, for guidelines in choosing puzzles.
The typical tactics trainer is not rated 3000.
Anyways, I think this should be in a seperate thread....
Where do you propose is the best place to discuss if there should be a uniform ambiguity factor for puzzles for 400-3000 on tactics trainer ?
Part of being 3000 as compared to 400 is that you can much more easily discern the right answers, inspite of there being a more difficult ambiguity factor potentially between puzzles. I realize that the T.T. puzzle moderator doesn't necessarily decide these things, but if a person wanted to get chess.com to consider any variance as the rating goes up, I think having the moderator's backing, or more specifically the prompting it would be best.
@ Katonas
Where would be the best place to discuss this or bring it to Chess.com's attention...
Why can't you just create another forum and message Katonas the link so it doesn't take up this forum? Also, most people who are 3000+ in TT either have done all the problems their level and memorized the solutions, or just do the ones they are certain of and skip ones they aren't sure of.
nameno1had wrote:
katonas wrote:
nameno1had, the rules for this thread are written in the first post. They include the lines :
'Examples not allowed :
solution move forced mate, next best move mate in more moves
solution move forced mate, next best move greater than +1.0'
I'll admit, I didn't look specifically at that, but why is that considered too ambiguous ? Once again I have to ask the question, aren't tactics trainer puzzles supposed to prepare you for real game situations ? Shouldn't that number for the next best move be consider too close for those who are really adept at tactics already? Afterall, I am trying to submit a puzzle to stump a 3000 rated tactics trainer user, not give a false rating to someone to puff their chest out in pride, or confuse a 1400 player with whether it is a potential mate or winning an exchange ...
I really think that this needs more carefully weighed, for guidelines in choosing puzzles.
looking at your discussion, it seems to me that you did not get the main problem in tactics trainer: there are no alternate solutions which means that you either make the only "solution move" of fail the problem.
this means if you really allow ambiguous problems, people will be punished even though they play a perfectly winnng move. of course you should be always looking for the very best move in a real game but there are a lot situations where you just cannot afford to, especially when you are running out of time as tactics trainer imitates. if you have, say 30 seconds on the clock, and see a simple mate in 5, why invest your remaining time to look for a complicated mate in 4 which you do not even know if it exists? or even if you leave out the time, maybe you get in a situation where you can choose to go for a clearly winning pawn endgame which you feel confident to win, or a complicated mating sequence in which you are afraid of having missed something? real game results clearly prefer the former option but tactics trainer would punish it.
unfortunately there are currently a lot of ambiguous problems in tactics trainer and the report button does not work, but the goal has to be to remove the ambiguities (or make tactics trainer allow multiple solutions which is unlikely to happen), and the first step is definitely to not add more ambiguous problems.
NOW END OF THE DISCUSSION, or this thread will get even more spammed ...
I think its 1.5, but have no idea how to do what you ask. Perhaps you could ask for help from the chess.com community. Please don't use this thread for discussions about tactics trainer, chess.com or Houdini as there are enough pages for me to navigate through already.