Given that IFDB has just gotten its 20,000th member, I was curious about how the typical membership plays out.
This graph shows number of members with registered play activity over time for three different cohorts: those signing up at IFDB in 2010, 2015 and 2020. A registered play means recording either a rating or review or checking “I’ve played it” for at least one game during a year. Note that a player playing one game and a player playing 100 games will each count as 1 for this metric within a given year. Also note that a large fraction of users do not have public playlists – for these, only ratings/reviews count as registered plays.
Many people enter ratings for several games that they’ve already played when first signing up, so the apparent play activity in their first year is higher than the actual number of plays in that year. This is the same graph, only it excludes any registered play recorded within the first week of IFDB membership.
Interestingly, the evolution of each cohort’s activity is broadly similar, despite their being taken five years apart. By the third year of membership, only around half of a given cohort is still registering plays.
EDIT: For the benefit of those who prefer tabular data to bar graphs, here it is for the above.
It seems to me that the data suggests two types of community-building efforts that would have long-term strategic payoff:
increasing the number of signups each year
decreasing the decay rate of involvement
Some ideas that might help with one or both of those:
visual overhaul of IFDB website
encouraging cross-registration between intfiction and IFDB (i.e. “follow this link to sign up at…” or even “check this box to also register at…”)
opt-out email invitations/reminders to participate in annual play events (for those with a solid long-term track history, e.g. IFComp, Spring Thing, and EctoComp)
I think I would be in favor of removing barriers to participation but less in favor of specifically trying to attract more participation, specifically in the sense that a lot of modern gaming is designed to have engagement loops that make them as involving as possible, keeping people hooked as long as possible.
Most of you ideas seem like barrier-removing ideas. Easy registration sounds nice (and I wonder if there could be a way to allow people to post ‘guest’ reviews without registering, but we have had problems with spam bots in the past).
Didn’t you come up with a game recommendation system once? I remember Dan asked you a couple of times for the code but you never released it publicly. Couldn’t hurt to share that! It seemed pretty useful!
I agree the front page looks kind of outdated (I do like recent reorganizations). Do you have any suggestions on improvements? I always imagined a ‘carousel’ thing like Steam has with cover art (although some people complain about bandwidth so a ‘lofi’ backup version would be nice).
I have a bunch of reviews I’ve posted here that I would port over to IFDB if it was easy to do so. I’ve considered porting them anyway, but it always feels a little daunting. The main things I’m not sure about are how to get the formatting ported (for me, italics and spoilers). I also don’t usually give ratings, so I’d have to figure out how to place everything on a 1-5 scale (or, you know, just give everything a 4).
To be clear, I’m not talking about some kind of dark pattern manipulative gamification of the site, here. I’m talking about periodically reminding the people who’ve signed up on IFDB over the years that IF still exists and that their participation in annual events helps keep it alive.
Yep. I still tinker with Magic 8-Ball every time there is a data update, but I’m not convinced it does anything objectively better than bg’s, which is a much simpler system using the same basic principle. For example, from Social Democracy: Petrograd 1917 bg’s system recommends:
Social Democracy: An Alternate History
Social Democracy: Popular Front
Lilac Song
Kukulcan
The Sumerian Game
Hills of History
Comrade or Czar
How a Bill Becomes Law
Solitary Stars
Bill’s Passage
The Magic 8-Ball recommendations are identical for 7 out of 10 on that list, and the overlap includes what I’m guessing most people would say are the best matches. (The recommendations that differ are: Biennio Rosso: An Alternate History, you are an ancient chinese poet at the neo-orchid pavilion and Ethics AI: Don’t Freeze Edition, skipping The Sumerian Game, Hills of History and Comrade or Czar.)
My intuition is that encouraging “duplicate” tags by different users (to be better able to gauge the applicability of a given tag to a given game) would help to improve automated recommendations, but that’s just a hunch in the absence of a useful amount of data of that type (which is relatively rare at present).
Regardless, there’s no reason to hold back on bg’s effort, which has been ready to deploy for over a year.
You don’t have to give a rating to post an IFDB review! Here’s a ratingless review of mine as an example: Threads of Snow - Details
I do wish it were easier to copy stuff over from the forum without having to manually readd the formatting, though. There are rich text to HTML converters out there but I have yet to find something that works for spoilers.
Given that this forum is probably the main source of Markdown reviews, it does seem like a good idea for IFDB to accept this forum’s particular syntax. I think that’s a good feature request.
I guess my reluctance to add ratingless reviews on IFDB is more cultural than technical? It just seemed like actual numbers were more important in that context, with people looking up things by highest rating, etc. And of course, most of my reviews here are for ifcomp games, which I do rate on a 1-10 scale, but per-comp instead of overall, and usually not public. (By ‘per-comp’ I mean that I give all numbers from 1-10 to some games in each comp, even if I would have given the ‘worst’ or ‘best’ game a 3 or something compared to all other IF in the world on IFDB.)
I would say that reviews are equally vital to IFDB as ratings, and providing either is helpful even if you’re not providing both! But of course there’s no obligation for anyone to crosspost their reviews there either, and I don’t want to be pushy about it.
The spoiler tags are pretty close! IFDB wants <>'s and this forum wants []'s.
Everything else looks like it probably is consistent? Even URLs both use [your-text](your-URL), and there’s blockquotes and lists, which are probably the same.
Was an off-the-shelf library used to do the conversion? How hard would it be to allow [spoiler][/spoiler] instead of just <spoiler></spoiler>?
A lot if the IFDB is centered around reviews, polls and lists which is fine for long-time users, but if you really look at it from a beginners point of view, it is not very friendly or helpful.
If IFDB rather be a repository, then the IF world should have a storefront. A storefront like steam or itchio. Something that won’t confuse beginners.
A storefront would center around:
Choice/parser/hybrid types of games
Genre
Author/game company - some way of branding - like one follows a popular author they like.
Best sellers
New releases.
Put a nice horizontal scroll in. Like ur browsing to select a movie.
Things should be changed if you want a storefront instead of a repository. But i’m not sure everyone wants change.
IFDB is neither a storefront nor a repository. It is, as the name suggests, a database or catalogue. None of the games are stored in IFDB itself, and a few (including former commercial games and games written in obsolete formats) are not available anywhere.
I think the recent reorganisations are great. The team of volunteers have put in a tonne of work. I’ve thought for a long time that the top banner is looking dated. A cleaner banner, somewhat deeper and with a nicely designed logo would go a long way to improving the front page without being a huge amount of work. I think the current banner was designed for desktop computers and doesn’t work well on a mobile phone.
If a carousel were to be implemented, what would be on it? Latest releases?
I genuinely haven’t thought the carousel thing through. Latest releases could be good, or recent poll votes, or selections from ‘random highly regarded games’. Actually, to be completely honest, I don’t really see any problems with IFDB’s front page but I know others do so I’m trying to guess what people want by thinking of other game front pages. I didn’t realize what I was doing, but I think that’s exactly what it is. So maybe I don’t really want a carousel.
