proof that chess is a losing game

Imagine the computer analysis of a game.

/ --- \ ___ / -- \ ___

and so forth.

Now look at this part, where it goes from white majorly winning to black majorly winning:

\___

whose turn did that happen on? It HAD to happen on White's turn - it couldn't have been that Black invented such an amazing move, that it instantly made Black suddenly a huge winner, despite the position being CLEARLYin White's favor before Black came up with that winning move. Bcause if that move EXISTED as a possibility, then it would have meant that the position was ALREADY in black's favor.

So that means the ONLY way to change the evaluation of a position through a move -- IS TO MAKE A BLUNDER!!

That means that chess is a game of BLUNDERS only.

the only way to improve a position is for your opponent to blunder.

And that means one thing and one thing only:

White is the first to move. Since people aren't perfect, it means white must be first to blunder.

That means that White always loses.

thoughts?

Your opponent doesn't necessarily have to blunder so you can win the game - a few inaccuracies can improve your position, too.

Look at this game, for example: http://en.lichess.org/MslZZXP5
No blunders, no mistakes, but white made two inaccuracies and black made six, and white won.

The theory only holds if you identify a line of play for black that always wins, otherwise you can't prove that white's first move is a mistake, and you can't prove that the only way to win is by waiting for a mistake.

In short, you're asking if you've "solved" chess, and the answer is no. Computers can't calculate every single possible move (yet), and until they do we can't know the answer to your question.

Chess.

An entirely information-based zero-sum game with limited choices.

Every move advances the board into a new [Position]. [Factor]s change, as in "King Safety", "Material (im)Balance", and even invisible ones like "Time" (in timed games, the remained time changes, and invested in thinking), "Strength (in)Difference" (during a game of play, the intel of the other players' playing style is learnt), "Mentality".

In a game of chess where both players play perfectly, white has the first move initiative and theoretically will win as long as no mistakes - even suboptimal moves - are made. Therefore, to win games, this is indeed one way: Take the initiative and not lose it. A win is guaranteed.

However, not all moves are perfect. Rather, all moves changes all the existing factors in some way such that, even with a non-perfect plan, the perfect moves may no longer be made. Therefore, an alternate way to win: Counter and outsmart the player with the initiative, forcing them to lose it.

Sure, the only way to improve your position is for the opponent to blunder

But your statement "White is the first to move. Since people aren't perfect, it means white must be first to blunder." does not make any logical sense.

Unihedron wrote: "In a game of chess where both players play perfectly, white has the first move initiative and theoretically will win as long as no mistakes - even suboptimal moves - are made."

This is actually not known. For all we know, chess could be a drawn game if played perfectly. We will know when chess is solved.

I also never understood why for example white can increase computer evaluation by making a move. it´s always the other side making a not optimal move which increases your advantage.

For me, Computer Evaluations were always a hard thing to understand, because let´s say at some point white is up +3. and now from this point, let´s have two computers play against each other. since they can calculate millions of move combinations, it´s likely they won´t make a mistake/blunder/inaccuracy. so, the game stays +3 for the rest of the game?

Or in other words: Let´s say you are +5 (= a rook ahead for instance), if both players/computers only play perfect moves, you´ll end up (simplified) with a king+rook vs king, which is Mate in x!
So, can you say any Advantage is mate in x if only perfect moves are played?

Take a look at the question from a relativistic point of view. Not knowing the "solution" of chess, we don't need to revert to perfection for the sake of comparison. Chess is played by people (or imperfect machines) and can be evaluated from the current situation of the game, given description of the states for each player (what #4 described).
Thus any positional mistake is not a mistake per se but rather it becomes a mistake when noted or successfully counterplayed by the opposite side.
Chess engines present a judgement which is highly limited in its scope by the assessment criteria used by the algorithm. And due to its inherent properties it is more convenient to express position evaluation change in terms of losses. However such factors as the imperfection of the assessment criteria, the psychological impact and player assymetries are not taken into acount. While they could be, by a human judgement, which usually separates between "good" and "bad" moves (which again is a convention). And as far as a person can make any positive statement on the move valuation, in this perspective, chess cannot be considered a loosing game.

The word in french is Échecs. If we remove the plural "s", when the french word is translated to english it means failure.
If we add the "s" it should be failures, but it is not. It is not because it is a name.
Arm wrestling is physical and one person must loose, but it is not a loosing game. The same goes for chess, it is a mind wrestling game. When we participate in a tournament, it becomes a sport, which is competitive.
I do believe that once chess is solved, white will always have the upper hand and black will be forced to work the position to create a draw.
If we remove the spaces between the chess pieces and test the following position, white will always wins. Note, both players should be of comparable strengths. Try this position out with your friends and if the black player wins, they are obviously stronger than white. It is only normal in this position for white to win.
http://fr.lichess.org/analysis/8/8/rnbqkbnr/pppppppp/PPPPPPPP/RNBQKBNR/8/8_w_-_-

solakamen: That's almost the gist of it, however;

To say chess *has a solution* (mate in N) is very courageous! The rules are defined where, when the board is laid, the defender has as much advantage (minus the initiative) to defend. I only said that white has a winning play that can be guaranteed, under the following conditions:

1

Chess is a zero-sum game (if one gains, another loses). The aggregate change from all interactions results in a equal value, opposing shift between all [Factor]s, where the pie cannot be enlarged. (Of course, only a few of these factors matter to achieve a victory.)

2

There are limited choices. Therefore, there will be a best move. Like tic tac toe, it is possible to predict a [Position] is "won for a side" before it has even finished; That is, you win no matter what your opponent responds with. This is a result of such a property where real-time games (you can choose to commit whenever) with unlimited choices (you can choose to commit whatever) does not have.

3

When both players play [Perfectly] from any "perfectly balanced position", there may or may not be a winner. The rules of chess allows for "draw", which the "first-move advantage" may not enough to overcome. (It may be easier to understand this if you compare chess to tic tac toe.) However, "the opening position" (eight pawns, a queen and king, three pairs of menaces, you know the drill) is not perfectly balanced - both sides has a weakness (F2, F7) and do not have as much territory controlled, and so white can use and expand their advantage.

... To visualize how such an advantage can be expanded, consider a tic tac toe played in, not 3 x 3, but a 300 x 300 playing space. Because neither side has secured territory, the player with the first hand, who shall be designated "X", is able to secure a win in just two moves: Were the "O" to play horizontally adjacent to the "X" played on the first turn, the "X" can play vertically adjacent to their previous move, forking the two squares above and under the chain where the "O" has only one move incapable of defending.

... Of course, with this tactic a 300 x 300 playing space isn't even needed, but as big of a 5 x 5 board, the "X" already has a guaranteed win in three moves. (If you follow the logic in the previous paragraph and is not an idiot, you'll figure it out.) However, for the 3 x 3 board, the lack of territories is still "perfectly balanced" - while at it, it is solved for "O" to have a drawing tactic no matter what "X" starts with, or in other words the "first-move advantage" is no use against this balanced game of 3 x 3 squares.

... What about tic tac toe set on a 4 x 3 board, or a 4 x 4 x 4 space (en.wikipedia.org/wiki/Qubic)? Well, unfortunately, both has been solved for "X" (the solution of 4 x 3 is particularly easy): The space is not "perfectly balanced", so while the starting player only has the "first-move advantage" - It is enough to win enough [Factor]s to secure a "winning" advantage.

4

Under perfect circumstances, the advantage (... by the way, we're back to chess) in "the opening position" can theoretically be expanded to an (either tactical or positional) advantage. This is not merely the "+0.1" value or the likes presented by an engine's brute force attempt, but a threat that the first player imposes by having the first move advantage and using it to the fullest by playing perfectly. The resources are in the first player's favor.

Rather than saying "white is winning", it is my belief that "white has the resources to win". However, even the "reliable depths" that machines of the modern day are able to brute force through doesn't *near* the degree of perfectness I long for, for even the top engines (sic; intentional plural) of the TCEC can't maintain a no-loss stream (not even just counting the games it played given the advantage of white)!

Machines brute force in the ways it's supposed to, ill of predicting the "best move" the opponent can play but only to acknowledge a very few factors (in order to convert the [Position] into a "problem" that can be solved and optimized by the model taught by the engine). Even if we reach the date where chess is solved, no human or machine computing a [Position] will be able to find it, so all it does is to fill in the number in "mate in N"!

However, since imperfect moves are so abundant, one misstep and you'd lose the initiative before you even know it - When the opponent plays a suboptimal move that you aim to punish, but in turn neglecting how the "perfect move" has changed because of it and missing it. The winning initiative will then be lost. Truly there is no way to rely on making "perfect moves", but sparring with a computer will probably be the best we can get in this field.

Having a player who is strong enough to introduce that sense of threat capable of challenging and countering your initiative as a comeback from a losing position is a gift -
Creative play can throw off the perfect players who aims to optimize their decisions.
Outrageous play can throw off the ill-prepared players who aims to deathmatch them with the theory they brought.
Sacrificial play, when still in the "counter the perfect player" sense, can throw off the players who aren't capable of identifying it as a trap.

That's what makes chess an amazing game.