jvanelteren / blog_archived

Blog about my personal projects
Apache License 2.0
5 stars 1 forks source link

Diving into BoardGameGeek | Jesse van Elteren #15

Open utterances-bot opened 2 years ago

utterances-bot commented 2 years ago

Diving into BoardGameGeek | Jesse van Elteren

Key insights on board game ratings

https://jvanelteren.github.io/blog/2022/01/19/boardgames.html

hilcoklinkert commented 2 years ago

Wow. You seem to have time to game, analyse AND make a picture with the kids. What is your secret!!!??

Nice article too.

HappyCerberus commented 2 years ago

The best way to rank game for non-experts is to rank by number of ratings. This is because the number of ratings is actually a good proxy metric. Games that are bad don't get many ratings, similarly games that are to heavy or to simple don't get many ratings, etc...

msaari commented 2 years ago

@HappyCerberus, I agree sometimes the simple approach is the best.

I run Lautapeliopas.fi, which is a big Finnish board game site, and I used to rank the best games based on BGG rankings from Lautapeliopas users. This wasn't bad, but when the API got harder to use, I decided to stop doing that, and instead started simply counting votes from the yearly "best games you've played this year" voting I've organised for a long time.

This method is very simple, but produces a list of best games that I find very good for non-expert readers. It highlights the evergreen titles, which is what the non-experts probably should gravitate towards anyway.

I've always thought Monikers is the best party game there is. This article proves that, especially when you consider that Time's Up! Title Recall! and Time's Up! Edición Amarilla are versions of the same core game.

robblatt commented 2 years ago

I wonder if a game being released more recently means that it has more reviews from more recent reviewers. Are new users flooding the site and rating games that came out in the last five to seven years because they’re more readily available? I wonder what the distributions would look like based on age of account.

kaizhu717 commented 2 years ago

Thanks for sharing! The review data does not have a column for timestamp. Is it possible to add that it to the data?