Board Game Arena

JonChambers wrote: ↑06 July 2021, 05:24 Oooh! Null isn't counted for averages!? Okay, this might be useful after all! My game is going to throw a lot of null statistics then.

2-player win after triggering game end: null
2-player win after not triggering game end: null

Yeah, that's going to be at the top of every 3-6 player game.

Null not counting for averages changes EVERYTHING!!! The possibilities are now endless!

If that's a bug, I hope it's never fixed.

Yes, null do not count up for stats but I'm not sure if the division for the total is using games without null or with null.
But again, this is really not the main point of developping game and as Een said, if you really want to compute complex stats, the publisher will be able to ask BGA to raw data (ie stats + number of players and other base infos) once the game get played (which is the real point of developping games).

Tisaac wrote: ↑06 July 2021, 07:36 Yes, null do not count up for stats but I'm not sure if the division for the total is using games without null or with null.
But again, this is really not the main point of developping game and as Een said, if you really want to compute complex stats, the publisher will be able to ask BGA to raw data (ie stats + number of players and other base infos) once the game get played (which is the real point of developping games).

I am the publisher, and I couldn't care less about seeing the stats myself. The game is done. It's awesome. Fully developed. No more work on it is required.

But the central point of the game is knowing how early to trigger game end. It would be useful to the player to realise that they win 40% of the time when they trigger it, but 55% of the time when they let their opponent trigger it. This would help the player understand that they're triggering game end too early, and should let the game play out a bit longer in future.

This is a stat that's actually worth buying premium membership to see.

That central point is only applicable to the 2 player version, so I need a way of stopping other player counts from influencing the stats.

It seems null is the solution, assuming Een was correct and the division is only by the number of non-nulls.

If it doesn't work, I'll just remove all stats from the game.