online chess games

Online Chess Games

Participate and you will find!
Crazy name, real matters
[ Sign up | Log in | Guest ] (beta)
zhnkiu 73 ( +1 | -1 )
the probability of winning if you see a winning move, how would you measure the likelihood that your opponent doesn't see it? if this can be answered, what distribution would this take? certainly the variance to this distribution could be quite large, but maybe with thought this can be reduced.

for this, we could use:
1. larger difference in rating implies a greater difference in depth.
2. smaller time allocated implies less depth.
3. lower rating in general implies lower depth.
4. the rating itself implies the probability of a correct move

we might also possibly use:
5. knowledge of subtilty of the move.
6. psychological state of the player.

if this question can be defined well enough, perhaps we can formulate a predictive experiment.
rallyvincent 126 ( +1 | -1 )
Not by rating alone Rating alone does not provide enough information. I have a medium rating, but I overlooked a simple mate. This happens and cannot be displayed in a formula.

Some factors I think are important though:

1. Is your opponent running an attack her-/himself? Sometimes players who are attacking do not want to quit the attack and make a defensive move, even if necessary. They overrate their attack and rely upon you being concentrated on the defense. On the other hand, if they defend already for a long time, they are more likely to discover your threat.
The same goes for a tactical move or an idea your opponent tries to pursue for a long time. If you hindered this, until you reached a position where this move would lead to a loss at the opponent, he might overlook this while triumphing because he finally made this move.

2. The more moves are to be made until the threat is visible, the more likely it is to be not met. Even if the opponents moves are forced, there will be a point where he thinks of all your potential threats are countered. Since every move adds a variety of possible answers, it gets harder to see everything.

Greetings,
Rally V.
kewms 18 ( +1 | -1 )
Who cares?

The definition of a winning move is one that wins even with best play by the opponent. In other words, it doesn't matter whether he sees it or not.

Katherine
coyotefan 17 ( +1 | -1 )
A winning move is a winning move There is no way to quantify it, do a predictive experiment on it, or whip up a magic potion for it.

Only one thing is for sure, when it happens I will have a huge smile on my face!
zhnkiu 45 ( +1 | -1 )
a winning move, in how i am trying to define it, is a hidden move with the possibility of a win.

i had read recently (Scientic America?) that a player who has a 200 rating difference above another will have a 3-1 win-to-loss ratio regardless of their actual rating. so i followed with a logical question: under the same conditions, at any winning chance, do you have a (near to) 3-1 cance of your opponent not seeing (and therefore not counterballancing) your trick? this *can* be tested.
zhnkiu 39 ( +1 | -1 )
correction, i think if the game itself has a probabilty win of 75%, then each critical move would have to be taken into account:
"the player saw that one, no win" TIMES "the player saw that one, too, no win." TIMES .... TIMES " the player DIDN"T see that one, finally a win"

reminds me a geometric distribution, but its been a long time to be sure: the probability of getting five tails in a row before getting a head (in a coin flip)
kewms 73 ( +1 | -1 )
Again, I don't depend on "tricks." I try to play good moves.

I don't think it's possible to extrapolate a 3-to-1 win-to-loss ratio into more specific conclusions about particular moves. Someone rated 200 points lower than Kasparov is still a GM, and is therefore very unlikely to miss tactical tricks, but he's still probably going to lose. People rated 200 points lower than me tend to make more tactical mistakes than I do, but they also put their pieces on inferior squares, fail to develop their pieces, and otherwise allow me to get a better position.

Put another way, "luck" favors the prepared. If I have a good position, I will have more tactical opportunities.

Katherine
ganstaman 104 ( +1 | -1 )
my opinion I think some are being a little harsh here. It is certainly an interesting experiment even if it has no practical value (no one's talking about playing 'tricks' -- it's all about playing your normal game and still getting away with some tactical shot -- if you play 'tricks' you will be playing worse than your rating). The problem is as kewms says in her last post -- sometimes you win just because you slowly develop a winning position. It's very difficult to pinpoint the winning move sometimes, and even if you find it, your opponent had no good way to defend against it because they just messed little things up earlier.

If you can figure out a good way to test this, computers may be more useful. They can play many games against each other in a short periods of time without complaining. Maybe even use Chessmaster or something where you can get 'players' of all ratings to play each other.
More: Chess
ccmcacollister 84 ( +1 | -1 )
zhnkiu I'm not sure if the Question as posed is the ultimate answer to winning a game. But a starting place. Certainly the first place to see a great move is if it is staring one in the face. And many get overlooked. And to add to the statistics from there.
I also DO believe in using statistical analyses in Chess, as it can tell you very interesting things. Particularly about a certain opening, or position, or opponent. In the end I would think it is the Winning Moves you see far ahead, and may actually be never made on the board as opponent finally sees them and avoid them, thus having to do something else unpleasant instead, that account for winning of games. { At least at the higher levels.} Yet every move that would win, like every hanging piece, must surely be a consideration in the tactical outlook of any position. imo. Good luck if you do go further in pursuing the matter.
ccmcacollister 138 ( +1 | -1 )
PS/ I used a lot of stats about opponents in winning the Mutual of Omaha Chess Championship a number of times. Mostly about results and play style,myself and opponents. Yet it is fun to notice personal quirks too.
One of my opponents was most surprised when I asked him if he decided to trade the center pawn or not, and asked how I knew he was thinking of that. Usually from watching eyes, but not necessary in his case. He (S.H.) would scratch his nose everytime when thinking of exchanging in the center. :))
That proved to have a lot of good psychological value haha . Besides giving us a laugh. Stats can be Fun too sometimes .
But I also know that I am most likely to blunder on move 18 in otb. Secondly to blunder on move 22. And in danger in between. I usually am at disadvantage by moves 8-11 and gainiing advantage by moves 12-16. After move 22 I start playing stronger, and usually advantage by 26. Moves 39 and 40 are more potential blunder moves, being conditioned there to 4o move time controls, I had to overcome a "Habit" of blunders there even after the time controls were no longer
40/1. And knowing of the tendency, Was able to overcome it. Once the flag was no longer dangling at that point! :)
My motto on stats was First Know Thyself, as those were the most helpful to me.
But all kinds made a difference. They can surely be a LOT of work. At least back then without so much d-base and computer assistance. I'd had to do it all by hand.
zhnkiu 32 ( +1 | -1 )
criticism may be valid. this puzzle may not be well defined. there can be large variances over time, game to game, move to move. also, the measurements can come in small increments, or be catastrophic (dependent on a blunder). the sample space may end up quite small if we want anything significant, so that our model may not apply in general.
coyotefan 41 ( +1 | -1 )
I have thought about the premise of this And I really do not think there is such a thing as a winning move. What there is would be is the proper reaction to a losing move. Maybe even a series of losing moves. The difference in ability of players is to recognize and react properly to a losing move and not letting the opportunity slip by. Again though, I do not think that there is any sort of matrix that can go along with this.
gothicgirl 7 ( +1 | -1 )
zhnkiu I would like to know, why you like to know this..

Julia
mattdw 61 ( +1 | -1 )
I agree with Coyotefan. This is something that I have thought about before and discussed with various people, I believe a 'winning move' is more of taking the opportunity provided by your opponents previous 'losing move'. If, hypothetically speaking, it was always possible to analyse a 'best play' variation to a 'true' quiesant position then the static evaluation of a position would be equal to that of the aforementioned quiesant position no matter how many moves away it was. 'Best play' by the player with the advantage would not improve the evaluation at all, only sub-par moves from his/her opponent would.
zhnkiu 91 ( +1 | -1 )
starting to get confused... gothicgirl
i see a move and wonder if i can get away with it. i haven't thought of application yet. maybe a pseudo-calculational heuristic analysis of a position is enough. maybe construed to witch hunting: is a player consistent with their skill level.

mattdw
please translate what you are saying. i am not trying to figure out what constitutes when a move ensures a win. rather, quantifying when a move should be made or not.

in following a certain line that you hope to better your position, you are able to deduce that if your opponent moves to anything other than u,v,w,z you will have an sizeable advantage. only u and v counterballances to equality but require much deeper thought to see than the weakening w and z. to a lesser player i would expect w,z to be played. to a stronger player w and z to be played. and something else entirely by a GM. in this sense, a winning move was not seen by the lesser player.

zhnkiu 1 ( +1 | -1 )
to a stronger player u,v played.
mattdw 182 ( +1 | -1 )
I'll try to clarify, though it may be a bit OT. zhnkiu, basically what I'm saying is that the opportunity for 'best play' is always there (if such a thing exists of course), this should mean that you cannot ever improve your own position only maintain it, it is the result of our opponents deviating from their 'best play' sequences that we get an advantage. I'll try to give an example to illustrate this:

In an end game with a forced checkmate for white 10 moves away we could say that given:
best play from both sides - mate in 10
best play from black, sub par play from white - mate in more than 10 (or no mate at all)
best play from white, sub-par play from black - mate in less than 10 moves.

We can see from the above that white cannot get mate in less than 10 no matter how well (s)he plays without the help of black making bad moves, I would say that the same holds for an evaluation in any position at any point in the game. This would suggest that there are no such thing as situations where you can 'force an advantage' because by definition the advantage is already there in that the opportunity exists. I don't know if any of this is particularly relevant though, so sorry if I made it go off topic! :)

Regarding what you mentioned in the first place about probabilities, there is a post somewhere that I made that I think covers what you are saying quite well - it's basically to do with the statistical expectation of candidate moves. This meaning that even if one move is not the 'best line' in the normal sense of the word it is actually possible to win more games or get a better advantage by playing some other move, I think that roughly translates as playing the player or something like that.
mattdw 9 ( +1 | -1 )
Gone. I had a quick search for the post but it looks like all the old threads have been cleaned out.
kewms 127 ( +1 | -1 )
Always assume that your opponent can see at least as much as you can. If you can see a reply that loses (for you), assume that your opponent will also see it. That's much simpler (and safer) than trying to figure out *which* winning moves he will miss and which ones he will see.

Basically, there are three possible situations:

* Your opponent is stronger than you. He sees everything you see, and some things you don't. In any given position, you are more likely to miss important moves.

*Your opponent is the same strength as you. He sees approximately the same things you do. In any given position, you are both equally likely to miss important moves.

*Your opponent is weaker than you. He doesn't see as much, but you have no way of knowing exactly which moves he will miss in any given position.

Now, most players are more comfortable in some kinds of positions than others. Some people love queenless middlegames, some people hate them. Some people are better at handling knights than bishops. You can certainly aim for a position that will make your opponent uncomfortable. But that's a very different tactic from trying to second guess his ability to see individual moves.

Katherine
coyotefan 11 ( +1 | -1 )
zhnkiu My database shows that move 28 is the most popular move for games to end. Hope that helps :)
ionadowman 206 ( +1 | -1 )
ccmcacollister... ...Hi, Craig! looking at yr last posting here set me wondering... Do you play poker at all? 'Scratching the nose' looks like a 'tell'...
Interesting discussion, this thread. Doesn't 'trick', as used early on, mean 'combination' or 'trap' - or just about any tactical coup in chess? So what is being enquired into is trappy play.
It seems unwise to rely on this to win games. If you spot the hole in your brilliant combination, you can just about count on your opponent seeing it too. Assuming he won't, will lose you a lot of games! It is also a good idea not to 'believe it' when your opponent makes a combination. Perhaps the obvious responses will lose for you - but what about unobvious responses?
There is one exception to this: when you are busted already. If you are dead lost, but see a way to present your opponent with a convenient way to go wrong, then play it! It's your best chance!
Here's an example from a club game many years ago:
WKb1, WQd2, WRe1, WRf1, WNd5, WPs,b2,b3,c2; BKg7, BQc5, BRd8, BRh8, BNh5, BPs,a6,b7,d6,e5,f7,g6: Looks pretty dire for white, don't it? That Black pawn roller will pretty soon engulf white, though Black's last move (27...Qc6-c5?) hadn't helped him much. With some vague idea of bringing a rook to the c-file thence to the 7th rank, White played:
28.Re4 Ng3? (Exactly the move White was hoping for! Black has other choices that preserve the winning situation, but, lost anyway, White hasn't hurt himself with this try.)
29.Qg5! Nxe4 (29...Nxf1 actually loses for Black, but with 29...Qxd5, Black can scratch out a draw. Black might, of course, have tried a non-capture, but White can then congratulate himself on being 'back in the game'. This line is more clear cut...)
30.Rxf7+! Kxf7
31.Qe7+ Kg8
32.Qxd8+ Kf7 (Hoping White will be silly enough to take the other rook. A fair enough response by Black. There's no other way to avert the draw...)
33.Qe7+ Draw!
ganstaman 44 ( +1 | -1 )
Either me or some of you... are misunderstanding this situation. Where does zhnkiu mention using this information to play traps? I saw this simply as an interesting study, but not one with any immediately obvious practical uses. No one is advocating playing tricks to try to win, since that would obviously be losing overall (and then your rating would be lower, ruining the experiment).
lucasbeauchamp 43 ( +1 | -1 )
calculating blindness ... strikes me as a waste of energy. I've seen terrible players find winning moves in difficult positions, and terrific players miss easy wins in simple positions. I expect every opponent to find the best move, and plan accordingly. When bad moves appear instead, I aim to exploit the errors without mercy.

Plan and calculate for the best move always from yourself and your opponent.
ccmcacollister 325 ( +1 | -1 )
Certainly a Best Move ... ... can be made at anytime. Tho just what is Best may always be open for debate, or vary from one player to the next. Expertise is knowing what is solid and objectively best. Imo, Mastery entails some degree of 'Being Right' beyond the objective. Case in point; Tal. In the end, there is no disputing Mate.
***
Any Winning Move must be preceded by losing move(s) by an opponent of course ... unless there is some sequence WT can play from the start position that would force BL to lose(or vice versa)(and such has never been shown), then less than best moves must figure into any Win by the other player. Which may well be a sequence of lesser moves met by better moves, till a small final infraction breaks the camels back, if Capitalized upon. Winning Moves are capitalization upon inferior moves.
Based upon what I have observed over the years. Masters seem almost able to induce errors when needed. Knowing exceptions, and applying pressure until there is a "crack".
ionadowman ... YES! I am the second worst Poker player in the world, insofar as I am aware. Several friends say it pays much better than Chess Tournaments. I certainly agree with you about trappy play. In psot postal I would Never play a trappy line if faced with a choice of an initiative, or the trappy move did not yield a position no worse than another alternative play would, if opponent were to play best thru the trappy line, in a 'serious game'. (Which can be real temptation, and one Ive seen otb players have the hardest time walking away from when trying to convert to postal.
But it is just like other Chess where patience and advantages win out in the end.)
To me, playing a trap that might allow equality when some pull was available would have been almost anathema. Only if absolutely convinced my opponent Could not find his way thru; that it would be beyond his ability, would I even have considered it strongly. Then not to speed the game, but only if it contained some artistic/combinative beauty I Could not resist :) Which is a weakness of character & play I would think. UNLESS of course You're RIGHT ! }8-)
PS/ Yet there is to define 'serious' ... must admit to picking and chosing which games to be Serious about, and using some for TN proving. Which dont always Prove! I guess that is being Trappy but Not Knowing It ?!? :))
I also found it useful early on, to play moves/lines that had troubled Me, against stronger opponents at times... in order to see how They handled it. As part of the beauty of postal Chess was that if you got a good continuation, even at the cost of a game, you might turn around and USE it in 4 or 5 games yourself. Since there was a lag of several years before critical games would get published, and 97.479 or thereabouts of postal games never would be published. Now most everything shows
up all too soon after played, imo.
Perhaps it spoiled me, that 80's-90's really was a Golden Age for Corr. Chess. !
ionadowman 146 ( +1 | -1 )
ganstaman... ... the 'trap' argument - or as I suggest any tactical coup - seems to have been a reasonable inference from zhnkiu's original question. This was: 'if you see a winning move, how would you measure the likelihood that your opponent doesn't see it?' The context of this enquiry is the title of the thread: 'The probability of winning'. It is reasonable for respondents to look for a practical significance to this enquiry.
A couple of points might have been worth clarifying:
1. A winning move for yourself ... or your opponent?
If for yourself, then presumably the probability of your winning (given best play) is 100% (expressed as a percentage). A mathematically trivial case? If for your opponent, though, then all sorts of factors might be involved, of which 'rating' is just one. A practical assumption might be to assume a 0% chance of winning, and resign on the spot! No? hen just take whatever chances you can. All of us have won (or at least drawn) heaps of 'lost games'!
2. How far off is this 'winning move': immediate, or does it emerge 2, 3 or moves off in some likely line of play? The further off, the easier to miss (though it has been observed that strong players do often see the farther whilst missing the nearer). In any case, some ideas are harder to see than others. Simple examples: sacrifices on empty squares are harder to 'see' than sacrifical captures; forcing lines are easier to determine are winning (easier to calculate is what I really mean) than non-forcing lines.