Win% statisticsdo not make sense

[FrankJones]

Why does win% count solo games as games played, btu then counts them as losses when calculating the win%?

Example: Someone has 45 wins out of 50 competitive games, which is a 90% win rate. But, this person has also played 50 solo games.
This player's win% will display as 45%, because it counts those 50 solo games as games played toward the calculation, but then uses the win total from the competitive games only.

This is really baffling.

It's especially bizarre if the player won all 50 of those solo games.

I see multiple obvious ways to fix this.

[Kayvon]

The win percentage doesn't work the way you'd expect it to.

Take a look at this FAQ entry:
https://en.doc.boardgamearena.com/Faq?_ ... ic_mean.3F

and this forum discussion:
viewtopic.php?t=17149

[FrankJones]

The win percentage doesn't work the way you'd expect it to.

Take a look at this FAQ entry:
https://en.doc.boardgamearena.com/Faq?_ ... ic_mean.3F

and this forum discussion:
viewtopic.php?t=17149

I remember reading about that. But, I only play 2-player games, and for 2-player games, it works as one would expect - aside from the bizarre fact that I described above, where it incorrectly counts solo games as games played but not as games won.

Example: In "Race for the Galaxy", I have played 74 2-player games, won 37, and it shows 50% win rate, as it should.

For Earth" and "Terraforming Mars" (all 2-player games; that's all I play), it shows exactly the win% one would get by calculating: (pvp games won) / (pvp games played + solo games played).

(As described in the example I wrote in the original post).

So, even for people who only play 2-player, the win% does not work as described because of the way it is counting solo games, which should not be how it is done.

The system for multi-player games makes sense, because the average should be 50%, and if they gave 1 W and 4 losses in a 5-player game, the numbers would look really weird. Except then if they're also including solo games along with the multi-player formula, it will be completely messed up too.

[Kayvon]

Ah, I'm following. Yeah, you're absolutely right. Solo would skew things in that case.

[katmk]

I fully agree that the definition of a win for solo and co-op is strange, and so the percentages are calculated weirdly as is elo. It's also just so inconsistent.

I have a couple of friends I only play co-op games with. One I have played 347 games with but this shows up as 0 wins between us. We are not quite that terrible

One of the games we play the most is The Shipwreck Arcana. This is listed in my XP section as a total of 92 (+1) games, 52 "victories", 28% wins. And still considered an Apprentice.

And yet Regicide is calculated as 57 (+101 solo games), 15 victories, 5% wins. Yet apparently this supposed win record makes me Strong.

Ultimately, it's not that big a deal, we play these games because we enjoy them, but it is irritating because if you're going to show stats and rankings, then you want them to be actually representative.

[Jellby]

28% wins. And still considered an Apprentice.
5% wins. Yet apparently this supposed win record makes me Strong.

The Apprentice and Strong levels have absolutely nothing to do with the win rate. They're linked to elo (or rather, pseudo-elo). How elo works for cooperative/solo games is up to the developer to decide. For some games, it's +1 for every game, win or lose, for some it's +1 for every win, for some it's +10 for every win, for some it depends on the assumed difficulty...

[katmk]

I think that's kind of what I'm saying. The inconsistency is really bewildering, when you have all of your statistics gathered together in one place.