I've recently found a frustrated win (win without fifty-move rule) in this position ...
It can also be won within the 50 move rule with traditionally accurate play by both sides (as above). Whether it can be won by White with traditionally accurate play against any traditionally accurate defence by Black is anyone's guess (but whoever posted this http://galen.metapath.org/egtb50/ could probably tell you).
I forgot to post that this is trash after seeing that I was wrong. Sorry guys. Let's just change the topic of the thread (about the advantages and disadvantages of using nalimov versus syzygy).
I thought Syzygy ran on my computer but actually, it was Nalimov (My Syzygy and Gaviota Tablebases were broken). Sorry.
I've recently found a frustrated win (win without fifty-move rule) in this position ...
It can also be won within the 50 move rule with traditionally accurate play by both sides (as above). Whether it can be won by White with traditionally accurate play against any traditionally accurate defence by Black is anyone's guess (but whoever posted this http://galen.metapath.org/egtb50/ could probably tell you).
I already knew it was wrong shortly after posting it but I don't know how to delete this. So pls tell me how.
I've recently found a frustrated win (win without fifty-move rule) in this position ...
It can also be won within the 50 move rule with traditionally accurate play by both sides (as above). Whether it can be won by White with traditionally accurate play against any traditionally accurate defence by Black is anyone's guess (but whoever posted this http://galen.metapath.org/egtb50/ could probably tell you).
I already knew it was wrong shortly after posting it but I don't know how to delete this. So pls tell me how.
Ask everyone to delete his post? I promoted Ariskokle to first poster. 
If I had this on the board in a real game it would went a but differnt but also I would not be able to checkmate in 50 moves only if black would play it as in one of the sublines. Lot of this endgames goes mate after 50 moves. And the egde pawn is difficult.
Don t know I think they are ok. And also others are just fine.
I have no clue why they are wrong. The OP removed his original post and now it looks as if I claimed the tablebase is wrong. But I didn't!
I have no clue why they are wrong. The OP removed his original post and now it looks as if I claimed the tablebase is wrong. But I didn't!
Glad you clarified that, I was thinking you just had used clickbait!
Disowning a topic should not be allowed. Even if the content is removed, the empty OP post should stay on as an unerasable placeholder.
Hmmm, so how are Nalimov Tablebases actually wrong?
The original post was saying that they don't take account of the 50 move rule, which is correct. The FIDE basic rules of chess don't include the 50 move rule since 2017 so they are now correct unless you're playing competition rules.
If I had this on the board in a real game it would went a but differnt but also I would not be able to checkmate in 50 moves only if black would play it as in one of the sublines. Lot of this endgames goes mate after 50 moves. And the egde pawn is difficult.
You don't have to mate in 50 moves.
If the 50 move rule is in effect you have only to mate without 50 moves elapsing with no capture or pawn move. In your mainline the pawn moves on moves 4, 40, 42, 84 and 85. There is no stretch of 50 moves between pawn moves so no draw would be claimable under the 50 move rule.
The game is similar to a game against the Nalimov database (with a few inaccuracies). Mate can only be achieved against accurate play once the pawn has reached a3. Against Nalimov the main way of forcing the pawn forward is to threaten to mate quicker if he doesn't move it. Syzygy doesn't take much notice of that threat so you may have to physically immobilise the king in order to get the pawn to move.
In some long KNNKP endgame positions it is possible for a program with attached Nalimov EGTBs to draw under the 50 move rule when a program with attached Syzygy EGTBs would not. These do not need to be "frustrated" or "cursed" wins. In some cases it would be possible for a different program with the same attached Nalimov EGTBs to win the same positions.
The issue of this topic was seriously discussed by FIDE in the previous century and caused the 50M-rule to be changed temporarily into a 100M-rule for this endgame. When it became clear, with the rise of the tablebases, that many endgames were susceptible to 50M overshoots, the original 50M rule was restored. Nobody enjoyed the alternative of returning chess players to school to memorize and pass exams on a long list of endgames with FIDE approved DTMs (Distances To Mate).
Today you might say that the 50M rule is no longer necessary since all chess games are finished by sudden death timeouts or by disaster after decreasing time increments (if not by agreement, mate or stalemate). The definitive blow to the ultra-long (duration) game was dealt by the abolishment of adjournments, inevitable when players with better engines started winning adjourned games.
Interestingly, the 50M-rule plays a different role in composition chess alltogether. There is no need to complete an endgame study within a fixed time window, so why require it? And indeed, 50M is abolished in that segment. However, retrograde specialists enjoyed having a 50M-rule - especially an automatic one without claims - as they can create complications near the 50M borderlines. So, 50M is specifically retained for retrograde compositions. What happens when a composition is both endgame study and retrograde problem? Good question!
Summarizing, the game-management rules have been adopted and adapted over time to support a mix of fairness, practicality, necessity and interest in various expressions of chess. I won't be surprised when these rules will continue to develop in the future.