meganothing wrote: Sat Apr 18, 2026 11:25 am
I would guess many
mod portals don't make the effort to count distinct users as that would mean you would have to save an ident-number that can actually identify the downloader. Small instrusion to privacy and more effort, also makes the download numbers shrink (
mod portals want to show big numbers). Also in the case of factorio not easy to generate an id for servers and users not using steam.
Only wube can answer if those are distinct users though.
I think you may be under-estimating factorio's
mod portal, in the past, it didn't count distinct users as precisely as now, so alien biome which is an old
mod may have a data aggregated from 2 method, but nowadays it rely on the user account to identifiy unique download. Even if you don't have steam and use a factorio.com account, you will not be counted twice if you download the same
mod and its updates ( or on different computer or month apart the same
mod again). It count as distinct "release" download, maybe 5 different release if you download 5 updates, but still 1 "
mod" download ( per user account ).
shuzen wrote: Fri Apr 17, 2026 9:17 pm
I just searched the FFFs for the keyword 'mods', and found a few numbers in FFF 372 and 391 ->
https://www.factorio.com/blog/post/fff-372 and
https://factorio.com/blog/post/fff-391
Of course the 3.5 million were accumulated over a longer time period than just 2022, but given that the
mod users are mostly addicts that also have been playing for years, I think my guess was actually not that bad.
Guesses that are not based on any meaningful data i think are not correct, i feel if the AI gives a number, it has to have been retrieved from somewhere or made up, and if it's from somewhere the AI should be able to tell the source. If the AI itself is making a guess , it should explained on which basis. I feel those links you give are more valid basis for guesses than AI answer, here i think it risk being the case of hallucination, if the data don't exist in the first place or regurgitating the data you linked in the first place.
However i still disagree with your reasonning, when you divide by 3.5 million players, if you fast forward 10 years from now, or maybe 20 , or long enough so that "very few people are still playing and downloading
mod" , you would have stats like 2099 = 100 players downloaded 1000 mods. And you would be dividing by 4 or 5 or 6 millions players, but in any case your guess would rather reflect the state of players who stopped playing versus players that are still active, rather than the % of players that ever downloaded a
mod (imo).
You could cross your observations with multiplayer games, every day how many modded vs non-modded game exist. It leaves plenty of room for unknown, but if you see something that is close to 50% ( i feel) , you may want to reconsider your initial guess of 20-30%. Because if 50 % of multiplayer games are modded, you may imagine the % of player having already downloaded a single
mod is most likely higher, as those playing non-modded may have played a modded game once in the past.