Games on the list should also be confirmed to be working well on Linux systems, so games like e.g. Jazzpunk (don't know how "hidden" it is) should be excluded, since it has a ton of reported issues on Linux.

Minor issues like broken achievements on Linux are fine in my opinion, but it might be an idea to make a note of such issues in a discussion topic.

How do find out all-time peak players? Nimble quest might be a good suggestion if it fits into your guidlines.

I use the numbers from http://steamcharts.com/app/259780. If you use Steam Enhanced they show up on the store page. Nimble Quest is surprisingly popular according to those stats. Was it in a bundle?

No, but it was on sale for $0.49 cents during the Halloween sale. Guess that pushed it out of the hidden gem criteria. It is really fun.

Since Steam Spy[steamspy.com] came along and revealed huge gaps in estimated sales and max number of players in some cases, I've become even more aware of how arbitrary these rules seem. I never felt comfortable with them in the first place, but I feel like there should be some sort of measure of what to include on our list. I just haven't been able to come up with anything good on my own. Sales would definitely be a good measure, but in most cases, we only have vague estimates to rely on. Any ideas?

In my opinion, increasing the amount of all-time players to 350 or 400 will allow more games into the parameters of this already niche group.

Having a standard is good for this group Flesk, I just think we need to expand the parameters a bit so we can be more inclusive and make the group more active in the activity feed with recommendations.

Yeah, I definitely agree. It has bothered me for a while that a game like Toki Tori 2+ that flopped commercially and is a big favorite of mine, doesn't fit with the current strict rules.

I'm going to compile some numbers from Steam, Steam Charts and Steam Spy for the games we're currently curating, as well as some other games that should probably be on the list, later today to see if we can get a better idea of some good guidelines. I'll post it here when I have something to show.

Either way, I feel like there should definitely be some more wiggle room than we've operated with up until now, and that it doesn't hurt to leave more up to the judgement of the group admins.

Here's the list at last, in the form of a downloadable csv file:

https://www.dropbox.com/s/gavv2nj8nffir7y/hidden_linux_gems.csv?dl=0

If anyone prefers a different format, let me know about it. It's not all pure data, since I've separated out a few games that have been suggested and also added some annotations at the bottom.

I'm not sure how to interpret these data exactly, but what's immediately evident is that being in a Humble Bundle boosts number of owners quite a lot. I'd still argue that a game can be owned by many and still be hidden enough to earn its place on our list though, since I'm guessing many just activates everything from a bundle and don't even realize they might have an un-played gem in their Steam library.

Hmm. I'll have to chew a bit more on this.

Starting with SilasMortimer's recommendation of Syder Arcade, I've raised the all-time peak requirement to max 250 players for now, and will update the first post in this thread to reflect that change. As always, I welcome any feedback.

This change also allowed me to include Toki Tori 2+ and The Tiny Bang Story among our recommended games.

Can I put Jotun on there? It has 267 players which is slightly above the threshold, but it is a great game and has a very low player count. Also the developers have been pretty linux friendly. If not, I understand.

Yeah, I think that's completely fine and I'd appreciate it. :) It definitely deserves more attention (to quote our description).

After the recent changes to the way review scores are calculated, I became aware of a weird quirk in the way games are sorted on Steam. I'm not sure of the details, but as far as I understand games with a positive average but less than 50 reviews are displayed after games with mixed and negative averages, but with more players in the various filters on Steam. It's sort of a catch 22 for games that flop on release, since in order to get more reviews they need more sales, and they're not getting more sales because customers have to flip through 5+ pages to find them.

I'm not going anywhere in particular with this, and as far as I know, it was like this before the recent changes too. It just makes me want to take a closer look at games with few reviews and see if there might be some gems hiding there that we have missed.

How about we go more democratic about it? We don't get many offers at this moment, it's perfectly possible to create a secret thread per game and explain why you think it should be curated by HLG and other admins could yay or nay it. Example (imagine it's a separate thread):

I would not mind that at all. I think that might be a good idea for other games we're not sure about for similar reasons too. I've come across a bunch of games like that, which have had their stats "destroyed" by bundles, without that necessarily leading to more players. Average players is probably a much better metric in those cases.

Yet another approach would be to use an external ranking, e.g. https://steam250.com/hidden_gems - posting just to add to the discussion, though. I can't decide if I like it or not, for now I'm still in favor of voting.