Is it possible to get 100% accuracy in a game?

Sort:
ThrillerFan

The following ought to be 100% accuracy by White:

1.e4 (advance a central pawn)

1...f6

2.d4 (Black did nothing to stop White from taking the center.)

2...g5

3.Qh5# (Undoubtedly the best move)

1-0

FernanJGC

I don't think so

magipi
FernanJGC wrote:

I don't think so

There are hundreds of examples in this thread only, but FernanJGC is not convinced.

professork2014
Ya
professork2014
It’s possible
professork2014
If you get opening win
goobrey

i'm pretty sure it isnt. if someone gets 100% accuracy they might be using deep blue engine!

nguyentrung1234
c124875

Yes. I think the simplest way is play a move then resign or your opponent resign.

MackyMacky1234pro

I got 100% Accuracy:

https://www.chess.com/analysis/game/live/124613373811?tab=review

OutOfCheese

75 pages just to say "yes" seems slightly excessive.

Shruthyukth

I think so! Or not😒

DrSpudnik

Both parties at 100%!

https://www.chess.com/analysis/game/daily/710775027?tab=review&move=15

AtaChess68
Cool!!
AtaChess68

I am new to this threat and am not prepared to go over 75 pages but i suppose the question has been raise if 0% is possible.
If so, is 0% on both sides possible? And what would the game be :-).

OutOfCheese

I don't think it's possible even for 1 player to reach accuracy zero. Accuracy measures how "close" you are to the best engine line, so there's some kind of metric defined to measure the distance. Since a move will always have a finite distance to the best engine move there will always be a positive percentage accuracy. A zero could probably be interpreted as having infinite distance from the best move which doesn't even mean anything on the board.

OutOfCheese

I have looked at that post and I fail to see the number "0" in whites accuracy assessment happy.png

Maybe it's a white font number on a white background but all I see is a white round corner rectangle.

OutOfCheese

As a former mathematician I strongly oppose the notion that the absence of something automatically means "zero". If it was zero they could put the number there, no?

Maybe what they actually have is "NaN" (= "not a number" in programmers terms) which is not the same as zero (which is a number).

OutOfCheese

Good idea, and if I'm wrong I'll accept it. I'm not trying to argue for the sake of arguing.

BPGHchess
Bh6_Anti-Dragon wrote:

I got 99.6% accuracy, but my moves were all book moves or perfect moves. Is it possible to get 100% accuracy?

https://www.chess.com/analysis/game/live/4374335351?tab=report

Yes, it is possible. (Short answer)
Yes, it is humanly possible if your game is very short and you manage to "stay" at a certain wdl or "eval" throughout, meaning you played all the best moves (below 10 full moves maybe) or if you are Hans and Magnus abandons the game. Otherwise, it would be challenging. Here's how lichess computes accuracy https://lichess.org/page/accuracy , not sure if it is the same way as chess.com computes it, but should be similar. (Long answer)