Board Game Arena

First, there is a statement from sourisdudesert that only around 8% of Arena games are turn-based, so the real-time games will make up the bulk.

There is no denying that for most games, you can fit in more 2p games than 2p+ games in the same amount of time. Of course preference will also play a role in how many games of each player count get played overall, but you can't just ignore that playtime is important, too. There is no proof that there is selection bias in the guru sel, but this is a totally logical argument, not an argument from authority. This is also not hypothesis or an estimate but pure logic.

Wow, 8% is surprising.

Anyway, I know bias thing is not purely imaginary. The problem is, we cannot distinguish it from preference precisely.
Yet again, multiplayer is not THAT good as a competitive format.
Also, play time logic may not apply for certain games (e.g. Carcassonne, Sushi Go!, etc.).
Or 2 player might be the only option practically possible (games with small player pool).
Requesting an additional weight on multiplayer gurus without considering all these factors is an argument from authority.
I personally hate 'there is no better alternative' narrative, but this topic fits perfectly.

ufm wrote: ↑07 October 2021, 23:03 Anyway, I know bias thing is not purely imaginary.

That sounded quite different in your other posts.

ufm wrote: ↑07 October 2021, 23:03 The problem is, we cannot distinguish it from preference precisely.

And why is this exactly a problem? The argument by Vandaley is still valid. Just because we cannot quantify the selection bias, it doesn't mean that we can simply ignore it per se.

Ideally, gurus are not chosen by number of played games but by amount of playtime, it would combat the possible guru selection bias precisely. E.g. BGA knows that 2p games of Agricola go ~40 minutes on average and 4p games last ~85 minutes on average. But maybe it is not worth the effort and I can totally understand if BGA admins go with a more simple and practical solution. But what annoys me if people think that this is an easy decision and selection bias can be shrugged of.

dschingis27 wrote: ↑07 October 2021, 23:37

ufm wrote: ↑07 October 2021, 23:03 Anyway, I know bias thing is not purely imaginary.

That sounded quite different in your other posts.

ufm wrote: ↑07 October 2021, 16:08 Real time games, probably, or not for games with adjusted time limit or simultaneous actions.

dschingis27 wrote: ↑07 October 2021, 23:37

ufm wrote: ↑07 October 2021, 23:03 The problem is, we cannot distinguish it from preference precisely.

And why is this exactly a problem? The argument by Vandaley is still valid. Just because we cannot quantify the selection bias, it doesn't mean that we can simply ignore it per se.

ufm wrote: ↑07 October 2021, 23:03 I personally hate 'there is no better alternative' narrative, but this topic fits perfectly.

I just started playing Happy City in the new Season and noticed it's completely ruined as only has 2 Players??? What on earth is going on??? What a dumb thing to do - they've ruined the game completely

To be clear, you asked me how I knew there was selection bias. My training and profession were my reasoning for knowing it exists. It was not an argument from authority, it was my response to how I know. I know because I am an expert on sampling and the criteria fits.

Point taken.