Carlsen resigns on move 2

[JustinHorton]

May have gone by now, but my recollection is that he didn't even want a move, just a kick in the butt to alert him to the fact that something was up.

Yes, this is exactly the kind of thing people are saying, and they are not backing it up with any particular detail or any particular discussion of the potential pitfalls. In a rational discussion, this ought to be a red flag that people may be speculating.

There is, by the way, an example upthread of such a situation being recently discussed.

[JustinHorton]

One thing that might be useful would be somebody giving us a definition of a game-changing move. I don't mean in general terms - we all understand that - but I mean something that you could give to an associate or a computer program so that they could recognise it and inform you (we'll leave the details of the "informing you" for now).

[Paul Cooksey]

.

[JustinHorton]

Let me try again, since among other objections a computer program is not able to stand up.

One thing that might be useful would be somebody giving us a definition of a game-changing move. I don't mean in general terms - we all understand that - but I mean something that you could give to an associate or a computer program so that they could recognise it and inform you (we'll leave the details of the "informing you" for now).

If we can't even get this far, then the argument is pure bluster.

[Ian Thompson]

One thing that might be useful would be somebody giving us a definition of a game-changing move. I don't mean in general terms - we all understand that - but I mean something that you could give to an associate or a computer program so that they could recognise it and inform you (we'll leave the details of the "informing you" for now).

That seems obvious to me, most easily explained by example. On move n, already played, White has an advantage of 0.5 pawns which is a clear advantage, but perhaps not enough to win the game. Black replies, and White's best move on move n+1 will give him an advantage of 2.5 pawns, which is a game winning advantage. You alert the player that he has a game winning move. Engines already tell spectators the advantage in the current position and the prospective advantage if the best moves are subsequently played, so the information is readily available to pass on to the player.

[David Sedgwick]

Some [arbiters] make a specific point of not following game commentaries whilst the games are in progress.

That is prompted more by a desire to avoid accidents that by any concern that they may inadvertently become a party to deliberate cheating.

[Paul Cooksey]

.

[JustinHorton]

.

[JustinHorton]

That seems obvious to me, most easily explained by example. On move n, already played, White has an advantage of 0.5 pawns which is a clear advantage, but perhaps not enough to win the game. Black replies, and White's best move on move n+1 will give him an advantage of 2.5 pawns, which is a game winning advantage. You alert the player that he has a game winning move. Engines already tell spectators the advantage in the current position and the prospective advantage if the best moves are subsequently played, so the information is readily available to pass on to the player.

Thanks very much! Now what I'd like to do is thrash through this a bit, not for picky purposes (I have no objection if you want to change your values or criteria at any point) but to see if it's going to be problematic. OK?

[JustinHorton]

(Just for clarification, and it's not fundamental, are we talking about +2.5 regardless of the present assessment, or +2 compared to the present assessment?)

[Paul Cooksey]

.

[Ian Thompson]

(Just for clarification, and it's not fundamental, are we talking about +2.5 regardless of the present assessment, or +2 compared to the present assessment?)

As we're talking about alerting the player to game changing moves it would be +2 (or whatever other number the cheat's decided they want it to be before the game starts) compared to the current assessment.

The intelligent cheat wouldn't need to be alerted if they're clearly winning though. They're told if the assessment goes from +0.5 to +2.5. They're not told if the assessment goes from, say, +5 to +7.

[JustinHorton]

OK. So our putative cheat, let's call them GM Grifter, is sitting at the table and they feel a little buzz which tells them an error of the previously-agreed magnitude has occurred, if they can only find the right response.

We'll neglect the possibility that they have already moved before they get the nod (this will presumably occcur sometimes but not too often). Now what I wonder, in the first instance, is - how often will such a situation occur, and they will as a result see a move they would not otherwise have done, and they will win a game they would not otherwise have done? Has anybody actually researched this question, or are we just relying on guesses?

[JustinHorton]

Second question. Let us suppose that Grifter either fails to find the gamechanging move, or finds it but, alas, lets the game slip later on. Do they get a second chance - should one occur - or are they restricted to one go per game? If the former, do we think that's a free hit - it may be - or does every instance increase the possibility of being caught?

[JustinHorton]

Third question. What effect does the availability of this cheating service have on Grifter's overall play? Is it entirely a bonus? But what about when the dog doesn't bark? When Grifter's looking at the position and there's no buzz, perhaps they're going to assume there's nothing special there. But maybe there is - except it whaps up his evaluation only by 1.9 or 1.6 or 1.2. Is he more likely perhaps to miss this kind of chance as a result?