Comments/Ratings for a Single Item

It sounds like perhaps you're misunderstanding the rules. Each turn a player attempts to make a move as usual. If the opponent wishes, they can force the player to retract that move and make another.

What you describe sounds like No-Chess
http://www.chessvariants.org/other.dir/no.html
For that game, I think the answer to your question should be "yes". In particular, white can "refuse" some impossible move.

I don't think so. The variant I'm referring can be found in Variant Chess
magazine, issue 1.

In that variant, each player refuses a possible next move of his opponent
while making his move.

Okay, you got me to dig out my complete set of Variant Chess issues. 

First: Refusal Chess [article by Paul Novak]

"Refusal Chess (also known as Rejection Chess or Outlaw Chess) was invented by C.H.O'D, Alexander... The only rule change from normal chess is that you may refuse one of your opponent's moves each turn (you cannot leave your king in check though and refuse your opponent's piece takes your king).

Since its conception two very similar siblings have appeared; that the number of refusals is limited; and where two moves are proposed together on each turn... 

Different pawn promotions count as different moves..."

So there is a confusion among similar games, which is causing the problem.

Based on my reading of this, a player may or may not refuse a move. So white is not obligated to refuse a move of black's. But black has the right to refuse white's move R-h5+. So white must make a different move. Still, any rook move on the H column would mate, except of course, R-h7. 

This does indicate a problem - if white did play R-h7+, could white then refuse k x R, leaving black no legal moves at all, and if so, would it be considered stalemate rather than checkmate? It's maybe slightly shaky logic, but not totally outside the realm of possibility.

In this game to me KNB vs K doesn't seem a forced mate anymore, but I have not checked yet, does anyone know better? Anyway if not so, many other like usaual close wins may not be. Point being the game is more drawish, maybe a rules may be made that you cannot refuse a checkmating move, or something on those lines. I doubt that is in the spirit of the game though :(!

The rules are not unambiguously specified, w.r.t. checking. In FIDE rules it is not legal to expose your King to capture. But in Refusal Chess, do you expose your King to capture when you move to a square that is singly attacked by the opponent? I would say "no", because you can always refuse the move that captures it, and so there is no danger to your King. An alternative is that you lose the right to refuse after your King is captured. Then Refusal and orthodox Chess have the same legal moves.

Under these rules even KRK would seem a draw, as there always is only one way in that end-game. (The Rook has to check along the orthogonal that contains both squares ot covered by the attacking King, and a Rook can always reach a given orthogonal in only oe way.) This makes the game very drawish. I therefore think it was a mistake to adopt the rule that you cannot refuse the opponent's only legal move; it would have been better to make that a checkmate or stalemate, depending on whether he is in check. That makes checkmating easier, rather than more difficult. It still does not seem too easy, as you still have the option to refuse a move that would put you in a position where you only have a single legal move. That is, you never really have only a single option, as you always have the choice to make that move or refuse the opponent's previous move. Another ambiguity is what happens when you have no legal move at all. Are you forced to refuse then, or does it count as stalemate?

Yes, you pointed out many problems, but I think this game brings an interesting concept, espeacially for middle game (in the opening there are usual many sensible options). I also believe that the larger the board and army size the cocept gains in complexity as there are more moves to choose from. Anyway something has to be done about the endgame. If all balanced games end up in draws what's the point?

Does anyone know how well a chess engine will play this game against a human?

It's not convenient, but this game can be played against an engine by setting the engine to its strongest level, then letting it play a move. If you refuse the move, then you make the engine go back, and play a different move (temporarily setting to a lower level, until you find the engine's "2nd" best move).

The computer will reject your moves by checking if they are "best" or not.

To me it seems that in a chess opening to the mid-game, there are enough move options that the computer's "2nd best" moves will always be very good - better than most human players. Therefore the board position will start to favor the computer.

In an end-game, the engine may falter - because it will not play well knowing that the best move will likely be refused. But to get to the end-game, the computer may already have achieved a formidable advantage, and the human player will already be in a hopeless position.

I tried this once, and couldn't find a flaw with this type of strategy (from the computer's point of view). But I only got to about 8 moves, and then stopped the experiement. The computer was in a winning position, but this was winning as judged by normal chess - not refusal chess. So I could not say for sure the result was conclusive.

I'd like to try again, but am already busy in other games. Can anyone predict (or know) the result if taken to its conclusion?

I'm not sure if it is the actual game but as someone who tried a bit chess programming I can say it is doable, it is just a matter of optimizing the 2nd move rather than the first. The main problem is that we should, as a community come up with better rules, I myself will soon start (I hope) other things, in my apothecary chess series. So I can't  actually study this bit but the idea seems promising. My belief is that it needs polishing. V. if you'd like you may post a challenge and we should try together (actually I prefer 2 as I find this game promising). All that if there is a game courier version.

A normal Chess engine should play this game disastrously poorly. It would be very easy to trick it. E.g. it will suppose its pieces are sufficiently protected when they are once protected, and attacked only by a single more valuable piece. E.g. wait until he moves up a Knight, protected by a Pawn. Than attack that Knight with your Queen. The engine will ignore it. It will also think QxN will not be your best move, as it sacs a Queen. So it will let you play that. Then it will try to recapture PxQ, and you refuse that. Bye bye, Knight. Repeat for the other Knight. And then Bishops...

It should indeed be very simple to modify an existing engine to play it. Just make it keep track of the score of the best two moves in each position, rather than only the best, and take the score of the second move as the score for the position. This will need a pair of moves to refute a line, rather than one, so it will increase the branching ratio by a factor sqrt(2), so you will lose some depth. To optimally use a transposition table it will have to hold the pair of refuting moves, rather than just the best. This shpuld be pretty trivial.

HGMuller, how far into a game will that be evident? I didn't detect anything like that happening in the first 8 moves of a simulation that I tried.

(In the example you provided, you say "...attack that knight with your queen. The engine will ignore it." This may be optimal play for the side with the engine)

I certainly believe an engine (as described previously) will falter at some point, but wondering how far into a game is it expected?

For me nothing was evident in the early stages. I could not find a way to out-smart the engine in the opening phase of the game.

Aurelian Florea, I also noticed you offered to play a game. If you'd like, we can do an experiment.

It's not human vs. human. It will be you (human) vs. computer (me).

I will always choose the best move by computer. If you refuse, then I'll play the 2nd best move (also by computer).

Let me know if you'd like to try that. I haven't played any game here on the courier(?) system, but I'm sure it's not too hard.:)

@V.

If you say you would just plug in engine moves, than it wouldn't be that nice would it. Anyway my point was that this game has a slightly different strategy. You should try the game courier it has plenty of nice games and I have plenty of already on challenges. Otherwise you may challenge me directly, my user name is catugo.

I agree that playing your own moves against an engine removes the "competitive" spirit. From my side, I would need to run an engine a few times at each move to learn its 1st and 2nd best moves - but still is not a lot of work on my part.

The idea was an experiment. It would be to test the theory that an engine would play "disastrously poorly". I've seen this discussion of human play vs engine for Refusal chess more than once, and (so far) I have not seen a human willing to go against the engine.

(btw: thanks for your offer - If I played a game against you it would be Enep. That sounds like a fun one too!)

Ok, we can do it, just plundge in maybe even 2 challenges (1white,1black), but I doubt an engine that is not purpuselly desingned for that will manage. But with you assitance situations like when only  one piece attack your queen, will be ignored.

H.G.

Will you kibitz from time to time? I agree with what you said about modifying an engine :)!

Note that engines nowadays almost all have a 'Multi-PV' mode where you can directly ask it to show the best N moves.

In the opening there is not much tactics. This is almost the definition of 'opening'; you just develop your pieces before actually hostilities start. I am perhaps 1500 Elo weaker than Stockfish, but it would not cause me any trouble at all to still be in an almost equal position (or slightly better, if I had white) after merely 8 moves. That is no matter of the quality of play at all.

While at the local supermarket cue, I got on thinking. So even in the opening it could matter. Think about fianchetto-ing your bishop: you move your b or g pawn and then your opponent keeps refusing your bishop move. On the other hand your king pawn opening seems to gain merit, think about the scandinavian, 1.e4 d5 and then you can merely move 2.c4 for example (not sure if it is actually good) as you can always refuse the capture. Point being it is not just choose the second best move, every move has new posibilities. I honestly think this game is more strategic than tactical by comparisson the orthodox chess, or at least is something else both strategically and tactically despite the state space of the game being the same :)!

Ok awesome, to my knowledge, Aurelian is the first person in the universe willing to play Refusal chess against an engine - a historical landmark! I think we only need to play one game for this experiment. If Aurelian plays White and wins, then he has outsmarted the computer.

I've never played a game on the game courier here before, so I hope I did everything right. I set up an invitation here:

(I hope to play one move per day, but might miss a few days. I do have some travel planned this month that might slow things down for short periods.)

Game has started well so far, but no move has been refused yet. I'm sure it will happen soon. I didn't think about it, but I'm not even sure if the game courier allows take-backs. If there is no way to refuse and take back moves we'll have to find another way to do it. We'll see how it goes.:)
 

I just refused a move, so you need to take back (just browse through and redo the last move). Trouble is with that preset there is a ban on the "pass" instruction so I just offered a draw you should not accept :)!

When I go to the game, it says "Drawn Game" and I do not have any options to do anything. (I haven't played a move, or accepted the draw, or anything).

Is there a way to take back my move (since you rejected it)?

All we need is a system to show a board diagram, and I'm wondering if there is an easier way to play this.

As this is normal Chess, you could use my turn-based server, which has it already set up as a demo. Because there is no rule enforcement, it is ideal for refusing moves: you just take back the move the opponent just did, so he can play his second choice.

I am ok with that but as my chess skills are very low, I'm not sure if I am the correct person for this experiment, anyway V. if you still agree we may try this way :)!

HGMuller,
I can move the pieces and set-up the board to the current position, but I don't know what to do to save it.

I already have a username and password. The opening window shows a conversation between you and "jon".

So my two questions:
1) how do a save a position?
2) how do I clear the conversation, so it's a fresh board for Aurelian and me?

This may work, but having trouble with some basic things.

So my two questions:

1) how do a save a position?
2) how do I clear the conversation, so it's a fresh board for Aurelian and me?

You 'save a position' by pressing the 'Play Move' button. This only works if you are in a game, though. So you first have to start a game, or recall a previously started game (that is still in progress). Once a game is loaded, and you are on move in it, you can move one or more pieces, and when you are happy with it, press 'Play Move' to submit all the moves you added to the game to the server, as a single turn. In Refusal Chess you should only move a single piece, undoing the opponent's last move if you want to refuse it.

The conversation is only there because I used that page also as a demo for broadcasting comp-comp games, to offer the people watching them to discuss about the game. I agree that it is less applicable when the page is used for depositing their own games, and there is no concept of 'the current game'. But you can just ignore the chat; it is not part of your game, just a general medium for all users of the page to coverse with each other. And as for the time being you would be the only users (I have no plan for broadcastig anything), you could use it as an open (volatile) communication channel.

BTW, when I press 'List Players', I don't see your name. So if you tried to register yourself, you did not succeed.

Ok, I set up "VReinhart" as my username, and setup a game "VReinhart - VReinhart" since I plan to only use this to display a chess diagram. Now two more questions:

What is the data next to "white/black" (i.e. "+23.46/-79.96"). Is there an engine evaluating moves?

Any idea for the best way for Aurelian to communicate moves? Maybe just on this forum, while I update the diagram there?

It would be spamming to post my moves here, maybe I can make an account!

Best way for Aurelian to communicate the moves would be for him to register too, start a game between the two of you, and let him play his moves whenever it is his turn.

The scores should only appear when you load a game that was broadcasted. (These have the engine scores as comments to the moves.) Probably they are not cleared when you load a non-broadcasted game after that; this counts as a bug. Just ignore them, like you should also ignore the PV fields for white and black. This is all stuff I added later to the page, in order to use the turn-based server also for broadcasting comp-comp games.

I will do so in the morning if that's ok :)!

That's fine. Just leave a message once you have your username, and I'll setup a new game with the moves we've made so far. I believe the game so far is:

White      Black
Aurelian   VReinhart
1.d4    ...Nf6
2.e4    ...Nxe4
3.Reject...d6

Since we don't have a working board diagram yet you don't need to make your next move yet. Once I setup the game with your and my usernames, I'll setup the game from here and we can keep playing.

Just for the record, my moves are chosen by the computer but the computer thinks it is playing chess, but we are really playing Refuse Chess.

Well, it should be clear by now that Aurelian is not very eager to participate in this. Because I realized that beating a naive computer is even more trivial then I thought, I now started a game against you on the turn-based server, and played 1.e4. So you can have the computer play against me, if you want. When the game is done, we can post it here.

Actually I forgot, but I'm happy with the way things turned out. Also I play many, many games now on game courier so that's that. Moreover I'm guessing that you are stronger than me HG in regular chess as I am a very poor calculator, although my strategy is ok, not brilliant but ok. But I honestly, and I hope I don't disturb you in any way V., don't understand how you would properly play refusal chess with an engine that was not specifically designed to do that. I'd like to reiterate my belief that this is not difficult to do, but it is still a thing "to do".  I'm more concerned about making the game work as the idea is very good but not polished. I hope your game could shed som light on that, too :)!

Ok, I believe I played a move. This should be an interesting experiment. For the record, I'm playing the "computer" where the computer believes it is playing normal chess.

For clarity, the way I'm using an engine is that it plays the best move. If that gets rejected, then it plays the next best move. It also accepts or rejects moves using this same logic.

(If nothing else, this has gotten HGM to play a game, who I believe doesn't frequently play games on public forums. Hooray!)

I've gotton busy with other stuff so I plan to usually only play one move per day.:)

Good luck to all competitors! I believe HG to be clever enough to come up with the trick I came up and beat the computer, as it is not build for such games :)!

Hmm, there is a small problem that I had not foreseen. If a player makes a capture, and it gets refused, this cannot be done by simply making the reverse move: the captured piece will not come back. I can solve that by altering the game into a drop game, with unlimited pieces in hand, so that you could drop the captured piece back. This is a bit cumbersome, though. Perhaps it is better if we tell the opponent in advance when we are going to refuse a capture. (E.g. through the chat.) This will also speed up the game, as we can immediately do a move that will not be refused, rather than having to wait an extra roudtrip.

E.g. after the already played 1. e4 c5 2. Qh5, I expect the computer to play 2... d3 (to protect the c4 Pawn). Then I would say "3. Qxf7+ {refuse 3... Kxf7}", so black can immediately play Kd7 (the only legal alternative, which I now cannot refuse, as I already used up my refusal pre-emptively). Then I can continue with "4. Qxd8+ {refuse 4... Kxd8}", etc.

Ok that's fine. If the commputer makes a capture I'll inform you rather than playing the move (and wait if you reject it or not).

Stockfish (set to level 10) actually plays e6 (another way to protect the pawn, as you predicted it will do). It did accept your move because Qh5 (by normal chess) is not the best move.

Well, as I had already written (in the chat) that I would refuse 2... e6, you can try again now.

I'm not sure what is going on at the game. The pgn doesn't seem to match the board position. I (the engine) tried to play 3.Nf6 but I'm not sure if it accepted it or not. It doesn't show that on the diagram.

I believe this is the gams so far:
1.e4...c5
2.Qh5...e6
3.(refuse)...Nf6

It is not clear what you mean with 3. (refuse)... Nf6 . Was the knight move a white move or a black move?

For move 2, White played Qh5, and then Black played e6.
For move 3, White refused Black's move, and then Black played Nf6.

At each move, each player can do one thing - same as normal chess.

Wouldn't this be the normal way to notate a game of Refusal chess? Maybe the parenthesis aren't needed. Does this look better?

1.e4...c5
2.Qh5...e6
3.Refuse...Nf6

Are you sure you flushed your browser cash? It seems an empty move was submitted to the server, instead of Nf6. This could have come about by a bug in the script for that page, which I fixed two days ago: it was periodically refetching the game from the server to see if your opponent moved, and then automatically show you that move. But it should not have done that when it is your turn to move. For one, it would be pointless. But it caused the move that you just played on the board to clipped off the game again, when it refetched the game from the server before you could press 'Play Move'. This could have happened to you, if you are still using the script from before I fixed this (because of browser caching).

Anyway, I repaired the game on the server, adding 4... Nf6, and played my own move. Note that I am going to refuse a capture now (see chat).

@V. ,

Ok, thanks, now I understand :)!

Perhaps it would be more intuitive to write 'undo' instead of 'refuse'. The turn-based-server page, which was not specifically made for Refusal Chess, considers the 'undo' a normal move: it wrote ee7 (SAN for e6-e7; it adds a disambiguator to indicate which Pawn moves to e7, because none of the black Pawns can move there legally). It will disambiguate the moves in the normal SAN way, though, assuming that you know which color is performing the move. Which in this case is not true.

About clearing browser cash - I never do that, I don't even know how. Is it "Ctrl-f5" (maybe varies with OS?)? I'll hit that next time if there's problems.

When using the nubati server, we can "undo" moves because that's what were doing. (I haven't done it yet because engine-player has not requested it yet)

But when producing pgn notation (to describe games), I think "Refuse" or "Reject" is better because that is the terminology used historically, and I believe it matches the intention of the author(s). But either way is ok for me.:)

Sorry to be pedantic, but it's cache, not cash. To address your actual question, you flush the cache via Alt-F5.

Thanks Nicolino, I appreciate the info.

I've been watching some of the Top-10 chess engine tournament, and am also swamped with my normal work, so have been playing slow. But I have not forgotten about this game.:)

What top 10 chess engine tournament are you talking about?

chess.com just held a 4-day tournament of the top chess engines. Three days were qualifying rounds. The match between the top two engines was just finished today (20 more games to determine the best engine).

I can reveal the result, but it might be better for visitors to just go there. There is video commentary, chat forums, and a pgn file of all the games. It will probably continue to be discussed and analyzed for several more days.:)

Link is here:

https://www.chess.com/article/view/chess-com-announces-computer-chess-championship

I'm whatching the TCEC championship which seems to be a mote detailed competition :)!

That's interesting too.  I actually prefer games with longer time control.

But I did like the chess.com live commentary - like watching a baseball game with experts discussing the games. I think they did a pretty good job overall considering it was the first year.

HGMuller:

Sorry for no moves for a few days. I tried to get the engine's moves today, but had problems. Not sure if it was chess.com or my connection. I'll be travelling for a few days, so might not play for a few days. But hope to continue game soon. Sorry for the delay.

Now that I am into refusal variants, it was of course easy to also make an Interactive Diagram for the simplest of those.

As for the interface: the AI just takes back the move you play when it wants to refuse it, printing a message above the board in red. The user can refuse the AI's move through a button.