# How much does the first move matter?

• 4 years ago · Quote · #1

Consider these two openings (only one move):

A)

B)

How much would it matter for good chess players (let's say 2000+ FIDE) which of these openings their games started with?

Let's say White wins in approx. 55 % of all GM games. In what percentage of games do you think White would win, if forced to start with opening A? And what if forced to start with opening B?

• 4 years ago · Quote · #2

Your question is not formulated clearly. And your premise that White wins 55% of all GM games is incorrect.  Game statistics show that White only uses four "first moves" for 98% of all "tournament" games. These are 1.e4, 1.d4, 1.Nf3, and 1.c4.  All other first-moves come to about 2% of games, and many of these transpose to normal openings.  With masters (ELO 2200+) the winning percentages are about 35% for White, 23-27% for Black, and the rest draws.  As the relative ratings go up for both players, the Draws increase and the wins decrease, with GMs over 2700+ winning only 26-28% for White, and Black only 16-24%. An incredible 50-56% are draws.

Your two examples give openings which are extremely unlikely in master play.  From a database of over two million games 1.d4 g5 (the "Borg Defense") is so bad that I only found two credible games, both of which were won by White.  1.f3 e5 (the "Barnes" opening) had only four games with split results (50-50) and no draws.

For beginners (as well as "good" players) a knowledge of sound opening principles is fundamental.  It is the "ideas" behind the moves that help you learn about openings, not the memorizing of move lists.  I recommend my articles, "Seven Things To Do in the First Ten Moves" and "The Ten Bad Moves," as well as a number of articles here on Chess.com.  For a more detailed opening introduction see The Ideas Behind the Chess Openings by Reuben Fine, or a number of other introductory works, whose names escape me at the moment.  If you want a short list (after checking out my articles and those on Chess.com) let me know.

A good database, and the knowledge to use it, can also help.  The one on Chess.com is OK, but the one on ChessTempo.com I find even more useful.  Register for free and explore to your heart's content.

--Coach Omar

• 4 years ago · Quote · #3

You are right, my question was unclear. It should have said: let's say White scores approx. 55 out of 100, in GM games. Now, how much do you think White would score if forced to play opening A? How much, if forced to play opening B?

I've read tons of pages about openings. Still, I find this question interesting.

Thanks.

• 4 years ago · Quote · #4
cigoL wrote:

You are right, my question was unclear. It should have said: let's say White scores approx. 55 out of 100, in GM games. Now, how much do you think White would score if forced to play opening A? How much, if forced to play opening B?

I've read tons of pages about openings. Still, I find this question interesting.

Thanks.

At master level, I would guess white would score about 75% in opening A, and 45% in opening B.

• 4 years ago · Quote · #5

Thanks, pawn_eater. What do you base this on?

• 4 years ago · Quote · #6

The statistics (although meager) have already answered your question.  With opening A, White "won" 100%, with B they each won 50% (with no draws).  Everything else is idle speculation and, therefore, meaningless.  If you want further statistics your best bet (short of talking some masters into playing the moves) is to let Fritz and Rybka play each other in a match.   ...Be my guest!  Personally I would be happy to play White with "A" or Black with "B."  The opposing sides have nothing to recommend them.

--Coach Omar

• 4 years ago · Quote · #7

The stats tell absolutely nothing, with so few games.

Yes, maybe I should let some engines have a go at it.

• 4 years ago · Quote · #8

And if Houdini can beat Rybka 100% of the time in these openings what does that tell you other than Houdini can beat Rybka 100% of the time in these openings?

Here's some statistics from my database (4,945,000~ games):

1.d4 g5: 13 games (mostly kids and unrateds...plus two thrown games) - White won 9, Black won 3.

1.f3 e5: 62 games - White wins 30.6%, Black wins 62.9%

• 4 years ago · Quote · #9

Of course both engines would have to play White and Black, in the same number of games. If not, this experiment would be worthless.

Thank you for the numbers, but that's still not enough to say much statistically about the strength of these openings. Maybe regarding 1. f3 e5, it might at least give us a slight hint. However, to be statistically reliable we would needs 1,000 or more games played with each opening. Preferably more than 10,000.

I thought I could have some engines have a go at it, automatically playing tons of games. So far I haven't found a way to do this.

• 4 years ago · Quote · #10

Stop wasting time on these two.  It's not productive.  There's a reason there are so few master games with these opening moves.  Think about it!

--Coach Omar

• 4 years ago · Quote · #11

1.d4 g5 strikes me as a good test of engine endgame ability - what is an extra pawn worth at stratospheric level ?

I would expect pro players to score above 80% from this position (at least those over 30, who can play decent endgames

• 4 years ago · Quote · #12

I suspect both positions are still probably drawn on best play, although I'm not entirely sure about the first one, since white does gain a pawn.

• 4 years ago · Quote · #13

I think the last move is more important than the first move.  Also the next to the last move is paramount in a chess game!  The first shall be last, and the last shall be first!  White must attack, and black must defend!  The King's can change colors; the white king can become the black king, and the black king can become the white king depending on the positional initiative.  It really makes no difference if you: "Black or White!"  King of Pop, Michael Joseph Jackson, August 29th 1958 - July 25th 2009.  "You gotta play the game!" Tony Dungy ESPN Football Analysis!

• 4 years ago · Quote · #14

Omar, I have no intention of playing any of these two openings. The reason why I choose these openings for raising this question is that my engine considers opening A to be the best possible opening for White, and opening B the best possible opening for Black.

• 4 years ago · Quote · #15

Laurent, if a score of 80 % with White with opening A, then what do you think would be the score for White after opening B?

• 4 years ago · Quote · #16

madhacker, if both of the openings lead to a draw with best play, then first moves essentially do not matter. Somehow I doubt that.

• 4 years ago · Quote · #17
cigoL wrote:

Omar, I have no intention of playing any of these two openings. The reason why I choose these openings for raising this question is that my engine considers opening A to be the best possible opening for White, and opening B the best possible opening for Black.

I think your engine is broken

• 4 years ago · Quote · #18
Arctor wrote:
cigoL wrote:

Omar, I have no intention of playing any of these two openings. The reason why I choose these openings for raising this question is that my engine considers opening A to be the best possible opening for White, and opening B the best possible opening for Black.

I think your engine is broken

Why is that? Which first moves (White and Black) would you propose as being better for White and Black, respectively?

I think my engine evaluation makes sense. In opening A, White already got a hold on the center, and after 2. Bxg5, White will have a piece in play and be a pawn ahead. Black has nothing.

In opening B, White has exposed his King, haven't laid claim to the center, and need to move another pawn before he can get a Bishop out. Further, the pawn on f3 is standing in the way for the Knight on g1.

• 4 years ago · Quote · #19
cigoL wrote:
Arctor wrote:
cigoL wrote:

Omar, I have no intention of playing any of these two openings. The reason why I choose these openings for raising this question is that my engine considers opening A to be the best possible opening for White, and opening B the best possible opening for Black.

I think your engine is broken

Why is that? Which first moves (White and Black) would you propose as being better for White and Black, respectively?

I think my engine evaluation makes sense. In opening A, White already got a hold on the center, and after 2. Bxg5, White will have a piece in play and be a pawn ahead. Black has nothing.

In opening B, White has exposed his King, haven't laid claim to the center, and need to move another pawn before he can get a Bishop out. Further, the pawn on f3 is standing in the way for the Knight on g1.

Apologies, I read it as opening A being best for Black and B being best for White

• 4 years ago · Quote · #20
cigoL wrote:

madhacker, if both of the openings lead to a draw with best play, then first moves essentially do not matter. Somehow I doubt that.

Only if you're a hypothetical perfect computer which evaluates a position as either 1, 0.5 or 0.

For us humans, even though two positions are both drawn on best play (known to above computer but not to us), one can still be far preferable to the other in practice. This is what the symbols += and =+ mean.