What is criteria for naming an opening

Sort:
Streptomicin

I was thinking of this before, but just recent events made me ask this question in public.

I played a game Fritz 12 vs. Chessmaster grandmaster edition (11 I think). Game started like this. 1.e4 e6 2.d3 c5 3.Nf3 Nc6

After game was over and I did analysis Fritz named game as sicilian closed, and Chessmaster as French defence. How do we name opening in game. By order of moves or by the position that is reached and that represent one opening. I know when I analyze my games in Chessmaster if position that I reached is one, but it is created with different order of moves it says sicilian transposed.

Like this game I'm having now.

I wanted to play Blackmar-Diemer Gambit, but we ended up with Caro-Kann.

So what would be right way to name opening in that game. Blackmar, Caro-Kann, or Caro-kann transposed, even that I did not see that transposed is mentioned in any games I have seen.

rooperi

Well, the 2nd is clearly a Caro Kann. I dont know ebough about Frnch/Sicilian to comment on the 1st one.

It is position, not move order, in my opinion.

TheRisingStorm

The criteria for naming openings is the position that is reached, not the moves that are played.  I know this because I tested it once on an opening database by playing the moves 1. e3 c6 2. e4 c5 to see what would happen, and the database recognized it as the sicilian.

Streptomicin

But 2 professional chess programs named same game different. So I guess there is no rule about this.

Tenna

I think it's the position that determines the opening.

The reason some programs do it by move order is because that's much easier to program: start from the beginning of the game, as long as in you're in book, following that opening line and call it what it is, and once you're out of book, keep it as whatever it is forever.

Whereas if you do it by position, you have find a clever search algorithm that doesn't take forever if you have a decent sized opening book.

Scarblac
Streptomicin wrote:

But 2 professional chess programs named same game different. So I guess there is no rule about this.


Of course. Chess opening names aren't some official thing, everybody can do what they want.

marvellosity

I rather think that one of your programs was using an incorrect algorithm, rather than it just being subjective.

I'm fairly sure it's by position reached. I can't remember exactly what right now but there's a Sicilian line that turns into a French about 8 or 9 moves down the line that's a fairly common occurrence.

rookrascal

Its a good question!

However, in logical theory an e4 opening is actually Kings Pawn openings until there is a reply and second white move, at which point the opening progresses and changes. In that sense it is the order of play that is crutial, snce the second white move is determined by the black reply to 1.e4. The name of openings therefore have to account for play and reply. I like to play kings gambit, but a 1...e6 reply forces me down an alternative route and I usually reply with 2.d4.

It's a bit like cooking. Same ingredients does not always provide the same results, its the process as well. chow.

Scarblac

A good algorithm (if you have an entire game to work with) is to work backwards from the end position, and use the name of the first named position you see.

But even then result depends on how deep your list names goes -- 1.Nc3 e6 2.d4 d5 3.e4 may be a French, but if your list of names only has 1.e4 e6 2.d4 d5 listed as the French and nothing deeper, then the transposition will be missed. And quite a few of those transpositions happen around move 10.

And for those IQP lines that can arise from the Panov Caro-Kann and the Karpov Nimzo and the English... it's just completely unclear  there what should have which name anyway.

Dimitrije_Mandic

Caro-Kann Defence, Advance Variation?

P.S. Mislim da je tip pozicije dosta bitan za nomenklaturu.