Originally posted by eldragonflyYou are single handedly ruining this entire forum for those who actually want to talk about the subject at hand, instead of posting to engorge themselves at the thought of seeing their own posts in an internet forum. You still have not pointed out why my original post you attacked is at all incorrect or illogical. I also don't see why a game that is all theory and then forced, obvious moves, then ended is at all helpful in the analysis. If a child such as yourself could find the moves, why couldn't another human being?
and this is only because it will show your incompetence as a ex-gamemod as far as determining cheats goes.
Originally posted by !~TONY~!ad hominem=logical fallacy. No !Tony the sad truth is you are dullminded and unobjective, and apparently unable to make even simple decisions. self-absorbed shallow twaddle by any other name.
You are single handedly ruining this entire forum for those who actually want to talk about the subject at hand, instead of posting to engorge themselves at the thought of seeing their own posts in an internet forum. You still have not pointed out why my original post you attacked is at all incorrect or illogical. I also don't see why a game that is all the ...[text shortened]... analysis. If a child such as yourself could find the moves, why couldn't another human being?
Originally posted by !~TONY~!Uh huh. that's what i thought, what pretentious claptrap, hence shallow twaddle. You arbitrarily trying to impose an unrealistic and subjective measure/interpretation upon a chess players moves and/or move calculations is quite silly. You are in neverneverland with that "roll-out" of an idea. You also fail to define what a "theoretical" game is and what is or is not a forced move, and why this is even important at all, in generating a reliable database and/or examining games of a suspected cheat. 😕
Why does my original post not make sense? I'm still waiting for it. You can keep posting that it doesn't make sense, or you can actually prove it.
Here we go, I can actually work with this:
1. I'd still like you to explain why this number is an unreasonable expression of the uselessness of a game as it pertains to the detection of a cheat.
2. A theoretical game is one that follows theory (opening theory, books, databases, etc...known moves in other words) to some result, whether or not it's a win or a draw, etc....It's a game that you could look up in a database, and played without thinking. This would make the game useless in detecting a cheat given that it's impossible to say he couldn't have just looked the game up in a book or a database.
3. A forced move is a move that has to be played such that the game doesn't immediately end in your opponents favor. For instance, after 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 d6 5. Bxc6!? It should be clear that 5...bxc6 is forced. This move will of course match up exactly with an engine's first choice. It should then follow quite easily that you could find numerous games with long sequences of forced moves that will cause match up rates to be quite high, even though it's entirely possible that neither player is using an engine.
It should now be clear that it's conceivable, especially as players get higher and higher up the rating ladder, that they play games that go very deep into theoretical openings (for instance, the KID and Dragon have theory that can extend into the 20th or 30th moves), then a sequence of forced moves (maybe one opponent blunders and the other player finds a tactic that immediately ends the game) follows, and the game ends. It should then be easy to see that this game offers very little as far as engine detection goes. The players almost certainly looked up the theory, then the players played some moves that were forced and easy to find, then the game ended. In this case, the statistic I facetiously named the Ro # would be quite low, pointing out the fact that the game isn't as helpful as one where both players had to make tough decisions about moves of equal value consistently after the opening phase.
Originally posted by !~TONY~!1) i think by theoretical you mean playing by the seat of their pants. Everyone does that, your redefintion here is useless. so those games are not useless, by definition. at the point of repeating what has already been stated here countless times, GM's don't play like chess engines.
Here we go, I can actually work with this:
1. I'd still like you to explain why this number is an unreasonable expression of the uselessness of a game as it pertains to the detection of a cheat.
2. A theoretical game is one that follows theory (opening theory, books, databases, etc...known moves in other words) to some result, whether or not it's a ...[text shortened]... make tough decisions about moves of equal value consistently after the opening phase.
2) i know what a forced move is, but i appears you are mistakenly assuming that every forcing move has but one response. Most games do not have long sequences of forced moves. Another "suspect" and biased interpretation.
3) Higher rated human players go deeper into theoretical openings, another faulty conclusion.
i see what the problem is, you are a database apologist trying to show that engine moves might be mistaken for moves a database user might make. and even worse you trying to supply a measure to your fallacious and subjective reasoning is rather silly.
1. I think you're seriously confused. Playing a theoretical move is exactly opposite to "playing by the seat of your pants". "Theoretical" means that the move is accepted as a valid choice by opening theory. If you can look up the opening move in a database, or an engine opening book contains this move, it's a theoretical move. This in an of itself makes it near useless to analyze since the analyzer will have no idea whether it was made by an engine or human. It's obvious to everyone in this thread but you that these moves should be excluded from analysis.
2. A forced move by definition limits the responses. It doesn't have to have one response, it just has to have obvious choices such that the engines evaluation between these moves and all other losing choices is sufficiently large to overstep the set threshold.
3. Given that you're 1340, I'm just gonna disregard this, since you'll probably never understand what it is to play difficult, competitive chess. I could almost guarantee higher rated players play deeper into openings. I could prove it solely with my games alone. You are 100% missing the point of my original post. I'll clarify. Here is my point. It's easy to construct (I know, because I just did it) a parameter that can show how useful a game is to analyze for engine use. I am simply saying that this number can be calculated for any game, and can easily show whether or not the players had a plethora of moves to choose from, or the opposite. I never said with what frequency you'll have games with a small Ro #. I simply said if this happens, high match up rates are likely to have less statistical significance, given that a lot of moves are either theoretical or forced moves. Games that have a higher # will of course be less theoretical, less forced, probably longer, etc...
Originally posted by WulebgrBerliner managed 100% three times in 1965, all as black! As white he was as low as 75% in two games.
For what it's worth: I finally achieved a 100% match-up with an engine in Game 5222927!
Originally posted by eldragonflyDid you break the upgraded insult engine? Lately your posts have been substandard. Please format your hard drive and reinstall the software.
ad hominem=logical fallacy. and you are pretending that my posts are content free, such a carefree and supremely idiotic assumption, is quite absurd. Surely you can do better than !Tony my boy.
Originally posted by eldragonflyWhat hopeless content free twaddle. I suspect the vacuum you are referring to exists only in your head. How do you keep your ears apart?
Wrong. engine matchup rates are a necessary ingredient of the theoretical framework.
Hey, this is fun! I must get me one of them Mark II Insult Engine doohickeys.
Originally posted by no1marauderI have a problem with this couple of 100 moves.
They are always "relevant" i.e. having a bearing on or connection with the subject at issue. What could be more relevant as regards whether someone is using an engine to help him choose his moves then the number of times he and the engine pick the same moves?
The more data, the better. But a couple of hundred (or less) moves in representative games often suffice as "overwhelming evidence" of engine use.
It could constitute only 4 games, maybe less.
I could guarantee that I could find 4 games here for virtually any player with an 85% match up if I looked but another 4 games from the same player may have a 25% match up.
The problem is the selection of games. Say you decide to select the wins only. That means you eliminate all the games in which your target played badly. Now you decide only to select his games against the strongest possible opposition. By implication these are the games he played best in and will be bound to have a higher match up rate. If you don't like the results of the analysis lets just look at a few more then select the 4 games with the highest match ups. How high would be reasonable in even the targets best games? I am afraid I just do not know.
This is where I have a concern about "manipulation of statistics"
Originally posted by PhlabibitRight, now I know what I was up to and what happened. I tried to post in the "thread that was removed" very late at night. Well, it was late over here in this corner of the universe. By the time I hit the post button the original thread had vanished. I was actually responding to comments about removing forced and "obviously best" moves and also demonstrating the care that is needed in these cases. The 5 in my earlier post referred to forced moves (king in check, one escape square and such like). I should have added that I also removed 24(!) moves that were judged to be "obviously best" by a particular criterion. The 5 forced moves are not in doubt. The opportunity for creative accounting lies in the way that "obviously best" moves are determined. I went with Robin Smith's suggestion that when analysing using an engine a move should be considered forced or clearly best in a positional game if it is the first choice and its score is greater than the second choice score by at least 1 pawn. He also suggests that for tactical games a greater score difference should be used. Since the game in question was allegedly positional (not so sure that is true but that was the opinion of those arguing) I simply ignored all moves where the score differential was 1 or more. The result was that 5 forced moves were ignored and also 24 allegedly best moves were ignored, giving a total of 29 moves ignored and a reduction in match up from 91/2% to 84% or thereabouts. Sorry about the lack of specifics but I only have scrappy notes I made at the time to work from.
Let me know where my math is wrong, but
55 total evaluated moves, ~91% match means 5 moves were Not Engine
remove 5 forced moves (engine matches)
wouldn't you still have 5 Non engine moves out of 50? ( 90% )
If you can give the numbers I'd like to see how this percentage could have gone from near 90 to near 80%
Later folks!