Two 2500 (ELO) players sit down and play a game of chess.

On a whim, one decides to play an opening that he is familiar with though it is not especially strong.

The only edge he has is that his opponent is unfamiliar with it.

How big would this edge be in terms of elo points.

(It is impossible to get any precision on this but it would be interesting to see if a consensus emerged.)

Which one has white?

Didn't you read the OP swindoc?

The one rated 2500 has White.

My off-hand guesstimation would be 50 points if the player is Black or 100 points if the player is White.

This is a good question.

I think it can be simplified a bit by assuming the player with the knowledge is playing white and that the line allows black to equalize easily with accurate play (something like 1.g4 say?)

In this case, objectively white should score 50% since the position is equal and the players are equal strength. The only difference is that the player with more theory is more likely to play accurately during the opening.

I think there would be close to 0 edge here beyond any normal edge from colors. I doubt hardly any GM has spent more than a moment looking at the grob, but all are going to be able to get good positions against it as black. The reason 'strong' openings are strong is not because they are worked out but because they involve good fundamentally sound moves. The same reason that weak openings are not so good. They often involve weaker or fundamentally questionable moves. A GM doesn't need to be booked up to get good positions against the latter.

I remember reading a GM annotation to a Kasparov game where his opponent had tried to play some offbeat grand prix type variation against Kasparov's sicilian. The gist of the comment was that people often do poorly when facing Kasparov's renowned preparation, but they often do even worse when trying to avoid it!

Yeah, I remember that article or something. It was Linares or Wijk or one of the big tournaments and several people used offbeat lines vs Kasparov. He was NOT 2500 though. We need to set up some prop games here.

Rating of the players involved is a key point. In a game between GMs, the player who is unfamiliar with the line is extremely unlikely to stumble into any kind of trap, or find themselves unable to work the line out over the board, so familiarity will probably not be worth much. In a game between two class D players, the familiar player could have a huge advantage, especially in a super sharp line. Since the Fried Liver* was mentioned in another thread earlier today, and I have an unhealthy obsession with that opening, I'll use that as an example. I believe that I could score 80%+ with either white OR black in the Fried Liver, if I was playing against someone of equal overall strength to myself, but who had never seen the line before. Give two GMs the position after Nxf7, though, with one person having made a detailed study and the other not, and I doubt the player familiar with the lines would have much edge, if any.

*1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. Ng5 d5 5. exd5 Nxd5 6. Nxf7

This could be simulated with an engine match, forcing the opening choice to a given line and then have one engine use the book and the other not.

I think a computer playing without an opening book is a substantially larger handicap than a GM playing an unfamiliar opening.

I agree. Unfamiliar openings cause problems because of random tactical oversights (e.g., an unfamiliar tactical motif) but more crucially the question of the proper middle-game plan. Computers aren't good at seeing middle game patterns from move 1 in part because they don't need to be: people just write opening books rather than improve that part of the algorithm.

Also, saying an opening is "not especially strong" is too vague. There are some openings which are "bad" for 2500 players because they may lead by force to an equal position, but Black needs to walk a narrow line to get there. That's different from playing something like 1.g4. In the first case, familiarity may be a significant edge; in the second case, I don't think it matters at all.

surely the time limits matter quite alot. if white is prepared in his offbeat opening and it contains many traps, the equal black player would usually navigate through the opening without trouble and equalise, but he would surely take alot more time than the white player, because he has to double check he's not falling into traps.

so wouldn't the white player get a slight edge in time, so the smaller the time limit, the bigger the edge

OP question: at least, about 100 elo points, imo. At any level.

[QUOTE=BobJoeJim;24700576]Rating of the players involved is a key point. In a game between GMs, the player who is unfamiliar with the line is extremely unlikely to stumble into any kind of trap, or find themselves unable to work the line out over the board, so familiarity will probably not be worth much.


i think i disagree here unless players in question are in the best in the world (aka cap vs marshall marshall gambit).
im pretty sure anyone else but capa would of crumble vs marshall.

and this kind of refrain hapenes often when we hear a top gm been surprise by the opponent chosing a line he didnt suspect and had trouble having a draw.

( i remember fisher match vs sicilian of spasky early and when fisher use d4! for 1st time, both of them lost the game i think when the other wasnt ready enough it seem)

and this is slowplay, so im not sure wich time limit we are talking about but like someone said already, less time = bigger advantage.

so 2500 players arent top gm and i reallly think one will have a much bigger edge imo, and only top gm might be able to put it off imo and even than ( fisher example and countless others)

IMO, it depends entirely on the tactical nature of the position.

Being faced with unfamiliar novelty in a tactics-rich position will cost even a GM a significant amount of time on the clock. If the position remains complex, then a mistake in time trouble is very likely.

There are countless examples of this at the 2700+ level and I don't see how this would be different at the 2500+ level.

I think the question asked is a bit too vague.

Which player deviates? The black player or the white player?
Obviously its a lot easier to find good so-called shortcuts as white.
What exactly is meant by "not especially strong"?
In my library this can mean everything from the Latvian gambit (Which has more or less been refuted.) to Reti or maybe even 1.c4
Do we talk about a tactical opening or a pure strategical one?

I think the question would had been far more interesting cutting it short to just this:
"The only edge he has is that his opponent is unfamiliar with it."