10 move game 100% best move

This position (by Greco) shows why engines have limitations. For those that do not know a bishop and rook pawn cannot over come a king if the bishop is not on the same colour as the queening square.
It is black to move 1. .. Ra1+ 2. Rf1 RxRf1+ 3. KxRf1 Bh3 and that is that. The vital g-pawn is either captured by the bishop or becomes a second h-pawn. The black king goes to h8 and it is a draw.
If you push the engine button, you can see the engine spin on its wheels or tell you white is 1.4 points ahead or some such until KxR is played

I once played a game as Black. The moves were 1.f4 e5 (From

I once played a casual game as Black. The moves were 1.f4 e5 (From's Gambit) 2.g4? Qhh4#.

My computer at home indicated that all my moves were the best. 

That was way back when I didn't know ANYTHING about chess, except how the pieces move. 

i played it for white for about 50 times

hello

Btw, I played two USCF OTB rated games last month. One of those had a 100% score. I wasn't particularly pleased with it tho because I made a mistake in the opening before the computer started analyzing. (The mistake had occurred before, so it wasn't counted as a mistake.)

According to chessbase mega database:

12.Nb3 is the better move.  It gives white a winning percentage of 67.2%, as opposed to 12.Nf3 which comes in at 33.3%

12.Nb3 hasnt been played since 2016?  So maybe something was found that is good for black?

I think the "?" after 12.Nf3 is a bit strong - I don't see a refutation. If there is only a sample size of 3 games relative to the main continuation having 30+ games, statistical comparisons with the scoring percentage are usually close to meaningless without analysis. 16...Qg6+? is a kind of "thank-you move" in that forces White to improve his position.

12.Nf3 invites a repetition, if nothing else. It certainly wasn't a great move. I think White can improve earlier tho.

Nice work smyslovfan!

Makes sense, FishEyedFools.