best reply to e4
Depends what you mean by "best."
Best chance of winning?
Best chance of not losing?
Best at steering the game into positions you're comfortable with?
Best Chance of winning ... c5
Best Chance of not losing... e5
Best at Steering into positions comfortable with
1... e6, c6, d5, Nf6, d6 (g6), b6
That is just My opinion.
There are 20 possible replies to 1.e4 (16 pawn moves, four knight moves), of which those 10 makes the most sense in number of occurrence :
#10 Owen's Defence
#9 Nimzowitsch Defence
#8 Alekhine's Defence
#7 Modern Defense
#6 Scandinavian Defense
#5 Pirc Defence
#4 Caro-Kann Defence
#3 French Defence
#2 Open Game (1.e4-1...e5)
#1 Sicilian Defence
Best at steering the game into positions I 'm comfortable with .
Not the best but try the Basman Defense. It is so rarely played, you might be able to come up with some new theories of play, even as a non-master.
I am attempting the same thing with the Old Indian.
Trend-setters!
This is sure more of an interresting question than it seems.
I mean by that, letting the engines evaluations aside, it is yes possible to do stats on actually played games, human versus human. One can use a tool such as Chessbase or similar, and perform searches implementing various relevant details.
And here is what one can see, fedding a brute search on chessbase database (covering over 7 million games) with 1.e4:
White scored so far from 50.0% to 89.0% to every possible black reply to 1.e4.
Several problems arises tho, when it comes to refine the search, into an exploitable statistic.
First, "white scores 89.0%" against what black reply to e4?" Answer: "others". I mean by that, that after listing the 20 possible replies, Chessbase, for some reason, gives 5 646 games, where the black reply to 1.e4, has been "other". I don't know yet what that is... Resign? Draw? Win on time? I don't know.
Okay, that being said, next step is "best stat" versus "worse stat" on actually played moves:
1...-Na6 seems the hardest to beat, when white scores "only" 50.0%
1...-c5 comes second, with 50.4% chances for white
Yeah, right. But...
1...-Na6 was played 59 times, when 1...-c5 was played 1 453 033 times.
So, it's hard to tell, if 1...-Na6 has any other quality than surprise and other psychological feature.
So, because you'll find 10 black replies to 1.e4 that has from 12 310 games to 1 453 033 games, and the other ten black replies have been played like a few hundred times at best, with the exception of 1...-a6 who scores 2 711 games.
Okay, so, I'm not going to give you the whole list, it's too much work for no penny and no fame, and ALSO, there are other problems to solve and sort out in all of that: We should then separate the games between, at least: "master vs master", "master vs club player", "club player vs club player", and get all the "full patzer" games removed from the database... Already a work of titan, when then, you'd want to separate the games played before what year from the games played later, etc.
So, it's, to me, an impossible question with still a few hints about what gives honourable chances to black against 1.e4, and what is more risky if you play a strong club player or above.
Good luck.
Check out this article: http://chesswinning.com/10-best-chess-openings/