Precision algorithm

Sort:
Sofidellavit
magipi ha scritto:

Your math is only good if you want an exact number for both players. (For example, for both). If any number is good, then it's just 2 / 1000.

But the topic is about the exact number

magipi
Sofidellavit wrote:
magipi ha scritto:

Your math is only good if you want an exact number for both players. (For example, for both). If any number is good, then it's just 2 / 1000.

But the topic is about the exact number

As I read it, it wasn't. It was about that both accuracies are the same, whatever that number is.

Sofidellavit
magipi ha scritto:
Sofidellavit wrote:
 

But the topic is about the exact number

As I read it, it wasn't. It was about that both accuracies are the same, whatever that number is.

Yes, you are right, I've taken in my head the precise number, but actually it doesn't matter which number we have. So, there are 500 favorable numbers for 250000: 2/1000

Sofidellavit
magipi ha scritto:
 

As I read it, it wasn't. It was about that both accuracies are the same, whatever that number is.

Very interesting.. How do you think is it possible to make the same accuracy on purpose somehow?

magipi

Nobody knows exactly how the accuracy calculation works. It's chess.com's secret. Probably very complicated.

Sofidellavit
magipi ha scritto:

Nobody knows exactly how the accuracy calculation works. It's chess.com's secret. Probably very complicated.

If so, then there is really very rare event happend, only if the chess.com haven't done this trick happy

Thank you very much for the time you have dedicated

blueemu

You are calculating it as if there was only ONE way to match the two numbers.

ie: as if ONLY 60.3 vs 60.3 counted.

Try multiplying your answer by 1001. Or if you are only counting the range 40-90, by 501.