Board Game Arena

same number of rolls each game. 25. And each game has the percentage to the hundredth. I would say that's pretty accurate.

tbhp wrote: ↑09 October 2026, 04:53 But those stats are rounded, you would need to actually count everything for the exact numbers.

Also, if different games have different numbers of rolls, it's not a good methodology to just average all their stats.

And as has been discussed before, when a game is abandoned, are you able to see the stats for that game?

tbhp wrote: ↑09 October 2026, 03:50 But everyone goes out of time sometimes, even if it's the once in a century occurrence where they get lost in a tropical jungle while in vacation and don't have access to an internet connection.

If they somehow got bitten by a Burmese python while feeding flesh to piranha in the Amazon, after a boxing match with some kangaroo, and then were captured by drug runners or a US agency while trying to escape, and were stung by scorpions while captured, then maybe I'd give them a pass.

mizzou95 wrote: ↑09 October 2026, 05:12 same number of rolls each game. 25. And each game has the percentage to the hundredth. I would say that's pretty accurate.

Why would percentages of 25 need hundredths places? Or even tenths places? It should always be a whole number.

If it's not a whole number, something is wrong with the stats logging.

An integer divided by 25, expressed as a percent, is always a whole number (and always a multiple of 4).

mizzou95 wrote: ↑09 October 2026, 03:35 yes, I realize it's 500 games. That is why I went back and found out my rate over ALL the games.

May I ask how you did this? Seems a lot of effort to do this by hand for almost 4000 games.

However if you did so I guess you have collected the data per game. Could you put this in a table with the links to each of the games and the percentage of your doubles in each game?

Silene wrote: ↑09 October 2026, 13:48

mizzou95 wrote: ↑09 October 2026, 03:35 yes, I realize it's 500 games. That is why I went back and found out my rate over ALL the games.

May I ask how you did this? Seems a lot of effort to do this by hand for almost 4000 games.

However if you did so I guess you have collected the data per game. Could you put this in a table with the links to each of the games and the percentage of your doubles in each game?

I have done similar things, manually, retroactively creating a data spreadsheet by clicking my games played history and opening the end-game stats page one game at a time and entering data.

But that was for 50-100 games. 3900 games is a whole other level.

If every game has exactly 25 dice rolls, can anyone give me an explanation for why the "percentage of doubled rolled" in a single game would ever be a weird number like 23.847 ?

Let's not forget, one of my possible explanations I gave in this thread is that the end-game stats are not always logged correctly.

https://boardgamearena.com/table?table=928509743

The percentage of doubles rolled are not possible for an integer divided by 25.

And the magnitude of the error would be sufficient to explain the overall deviation claimed by the OP.

I check a few games. On the games I checked, the percentage of doubles is indeed always a multiple of 4... except when the game didn't finish.

Mizzou, when you calculated the average, did you include the unfinished games and give them equal weight to the finished games?

Aha. The game I linked was indeed not finished.

Also, conceded games are different from abandoned games. Conceded games still show end-game stats. Abandoned games often do not.

mizzou95 wrote: ↑09 October 2026, 05:12 same number of rolls each game. 25. And each game has the percentage to the hundredth. I would say that's pretty accurate.

Games that end prematurely aren't going to have 25 rolls. You didn't answer if you can see the stats of those games. And if you can, did you use those stats or did you go manually count the doubles in those games?

Can confirm end-game stats are displayed for abandoned games:

https://boardgamearena.com/table?table=925105170

Okay, so, the only remaining issue is, how much are the stats skewed by not taking a weighted average to account for conceded or abandoned games?

There would need to be a lot of conceded or abandoned games to explain the differential between 16.67% and 17.25%.