Alekhine's Defense = questionable?

Eventful or not, either of these give a bigger advantage than using alkehine's. 

Stats from Lichess database comparing 1... Nf6 with 1... e5 at various levels of player rating, all games of which are Blitz, Rapid, Classical, or Correspondance:

Masters database:

At the Masters level, it gives Black +5% chance of win, but it also gives White a +7% chance of win, leaving White a 3% "statistical lead". So, stats say that on the master level, 1... e5 is better, if slightly. Or, arguably, they're pretty much the same.

Lichess database (2200 - 2500)


It's basically the same for both players at the 2200 - 2500 level... but Black is "2% better", on the grounds that Black wins 1% more often, and White wins 1% less often, for a total of 2% improvement for Black.

Lichess database (1800 - 2000):

At 1800 - 2000 level, it's arguably better for Black. Black gains +4%, and White loses 5%, leaving a 9% total gain for Black -- not insignificant, arguably.

Lichess database (1400 - 1600):

At 1400 - 1600 level, Black gains only 2% win rate, but White loses 2% as well, with a 4% total gain for Black. Perhaps slightly better for Black.

Lichess database (1400 - 2500):

At the more average 1400 - 2500 range, Black gains 3% win rate, and White loses 4% win rate, so Black is "7% better".

Stockfish assessment: Stockfish likes other moves like 1... e5 better than 1... Nf6.

Alekhine's assessment:

looks like Alekhine had no problem saying how great the move was.

Conclusion

Based on the above, I can see the argument that 1... Nf6 is perfectly fine for Black, unless you're playing at the master level, and that even then it might be ok, depending. But, on the other hand, Stockfish doesn't like it very much, and ranks other moves higher. And there's the argument that White gets to smack the Knight around the board. What if 1... e5 is simply objectively easier to play? I don't know.

@99

"all games of which are Blitz, Rapid, Classical, or Correspondance"
++ That is a pool of different games. If you want the truth, then look at correpondence.
If you want a practical truth, then look at classical. In blitz and rapid everything is viable.
Win rates correlate with rating difference, not with opening.
If all stronger players for some reason decide to play 1 e4 h6, then that would yield the highest black win rate.

@ tygxc

If by truth you mean, how people will play in correspondance and not those other games, then yes.

I agree with you on the rating difference statement -- unfortunately, there's no simple way to filter results by games where higher rated players crush lower rated players, if those are the games you're interested in.

wasn't c5 a fork?