Determine the length of a game (before playing)

There was a long break when I didn’t play any IF at all (about 1990-2018). This means that I have a lot of catching up to do…

One problem I have is to determine how large of a commitment I need to do playing a game. Is it a multi-week effort or is it a couple of hours? I personally would very much like a field in the “game details” on IFDB that says something like: “Approx. playing time about 4 hours”. But, in lack of that, I need some other method.

Example: Lost Pig
The fact that it is a IF Comp entry indicates that it should not be very big but on the other hand it is a Z8-file that indicates that it is big. Wich is it?

Is there a better way to determine the length of the game besides reading reviews and trying to extract the info from them?

You should definitely check out the reviews on mathbrush’ IFDB profile. He has a nifty categorisation system for games according to playtime. Anything from 15-30 minutes for quick diversions to +10 hours for massive behemoths.

[Edit: here it is: Reviews by MathBrush (]


And if the game was in a recent IFComp, you can look it up - they have been asking for a play-time field for at least the last two. Be aware that’s set by the author, so YMMV.

1 Like

Spring Thing asks for a rather vaguer description of time (micro means 5 minutes or less per playthrough, short means between 5 minutes and an hour, and full means over an hour - which for The Weight of A Soul actually means several hours). Again, it is set by the author.

I did hit the problem that length of playing time is rather “how long is a piece of string?” For Budacanta, my playtesters varied from 20 minutes for the excerpt to one tester who needed 50 minutes to get halfway through the excerpt. (They finished it after I left on the first occasion, so I don’t know how long it took in total, and we used the skip option a lot when testing the music worked on the last test).

1 Like

It’s maybe worth noting that you can’t really infer game size – however you want to measure that: number of rooms, amount of text, playtime – solely from Z-code version, nor from game file size, because a lot of that depends heavily on which development system was used (and which version of that development system). Every system includes at least some of its standard library rules in the compiled Z-code, and feature creep in standard libraries is a good thing in a lot of ways – it adds more functionality for the player without the author having to do much, and covers more unusual cases, but that doesn’t mean it increases the play time of the game. Dropping in an extension in Inform might just add some minor functionality that’s tangential to the game’s main plotlines and doesn’t increase the play time for the player at all, but it does increase the file size and cause the game to creep towards the limits of what a given Z-machine version can handle.

So a .Z8 game written with a recent version of Inform 7 might quickly run up against the limits of the .z8 format, but “the same game” from the player’s perspective, with the same text and similar rules, might be much smaller in Inform 6 or Dialog, or even in older versions of Inform 7. A lot of it depends on what’s going on under the hood, which might in many ways be a black box even to the game’s author.

1 Like

How do you even determine length in a genre that’s about puzzles? Are we using a cap on maximum amount of time spent before a player uses a walk-through to cheat a puzzle? Some people get stuck on a puzzle for days and others guess the answer right away, so…

1 Like

Yes, you’re right…

You’re also wilfully skating around the heart of the question. You know what Henrik Asman meant, you’re just being nitpicky about it.

If ten people play the game and seven of them finish it in about three hours, then that’s a good estimate of playtime.
A crude estimate, unlikely to be correct for some players, and likely to be finetuned if you get a bigger sample group.
But still: a practical estimate.

(of course, if the sample group is two people, the author and their best friend, that’s another story…)


It sounds like you read my post as though I was trying to troll or something but I really wasn’t.

I always found text games to be funny because it actually only takes about 3 seconds or less to complete the game if you can pass correct commands to it using a transcript or something, whereas if you passed the correct input to a shooter or RPG, it would still take pretty close to the same length of time to complete as a player.

Unlike your average video game where you can kind of gauge time based on length of cinemas, character running speed, amount of back tracking and level grinding required, etc… text game length is based entirely on the reading speed and puzzle solving skills of the player (possibly ignoring the ability to get super lost like I do), so I’d imagine length would vary wildly.

Maybe ifdb could use a feature like howlongtobeat or even the one gamefaqs has where users can list the time it took to complete so you can get an average without the author having to guess.

(And no, I don’t know why I randomly picked RE7 as the example game. It was the first that popped into my head.)

1 Like

Not trolling, but going into a theoretical objection that is besides the point. It is possible to give a ballpark estimate of how long a game will take to finish for an average player.

I think this is a really good idea. So, for example, when you check the little “I’ve played it.”-box, you get a list of options, also with little boxes to check.

–>How long did it take you?
-15-30 minutes
-30-60 minutes
-1 - 3 hours
-more than 3 hours

We could add a little checkbox for that:



Unfortunately, the proposed method only works if the sample of those who provide information is somewhere near the average of those who haven’t provided information (I assume that as this is a proposed IFDB feature, we can ignore any player who wouldn’t take the initiative to go to IFDB, which would take out a fair number of the longer outlier times).

Given the majority of the latter don’t have IFDB accounts, this is difficult to assess - especially since someone who has no account and sees a blatantly wrong (or wrong-for-them) answer is more likely to dismiss other time suggestions on the site than they are to create an account to update the record, or even to consider that the previous reviewers might be more/less experienced at IF than they are. I believe this is why VNDB (which is about visual novels, that often don’t have puzzles to complicate this further) only has divisions at the 2, 10, 30 and 50-hour level - anything more “precise” is apt to become inaccurate for too large a proportion of people using the database, except possibly division(s) for the shortest IF (the sub-30-minutes ones, which are rare in the VN space but considerably more common in other forms of IF).

This is a knotty problem…

There are some suggestions on IFDB’s suggestion tracker for a “time to completion” field and the ability to search that way:

No one has commented on the tracker to say they are working on these issues. So as far as I know, they are up for grabs.


If you were going to have pre-set categories, these seem like good time ranges (although I wonder if it might make sense to have another category beyond 3 hours–not sure what a good cutoff would be. Maybe 3-10 hours and more than 10 hours?). You would also need a category for the shortest games, since some are under 15 minutes.

1 Like

Maybe there could be a checkbox for “I used solutions/walkthrough for a lot of the game” (like, half the game or more) when reporting play time, or something like that. Not sure how it would be figured into the average (or whether there should be more than one average).

Although…do a lot of people actually play this way? Using solutions/walkthrough for most of a game? If not, maybe those answers could be excluded from the average.

Or maybe if it’s a really hard game, and more than half the people check that “walkthrough” checkbox, the average with-walkthrough completion time could be reported as the average playing time, with the note that most players use a walkthrough.

I’ve heard people mention that they do. I think it happens a lot more around competition time because the reviewers can’t afford to get stuck on a puzzle for a single game. The options then become to cheat so they can see the rest of the game, or drop the game for lack of time.

I’m not sure if this experience translates, but a lot of my RPG-playing friends play with a strategy guide or walk-through by their side the entire time because of FOMO and they don’t want to miss a single thing. But walk-throughs and hint guides for IF tend to only tell you how to solve puzzles and find items but not any easter egg stuff, so it might not apply.

I usually download games and play them in an interpreter, but my understanding is that many games can be played on the web. Perhaps the servers could measure and aggregate anonymous playing time statistics and feed those back to a database. Imagine:

“Pineapple Upside Down Cake”
Playing time (hh:mm)
Average: 3:14 (n=37, std. dev. 1:22)
Median: 1:41
Range: 0:01 - 11:22

Admittedly, measuring play time on the web has challenges (e.g., player leaves the game running in the browser during the day but only actively plays in the evenings).

The interpreter could send analytics to IFDB, but this gets into a bunch of privacy issues which I really do not recommend.

Admittedly, measuring play time on the web has challenges

Of several kinds. How do you know when the player has finished the game? (As opposed to replaying it for review reasons or to try alternate endings.)

I dislike using in-game clocks because they’re never accurate. My JRPGs always say that I spent 80 hours playing the game but at least 15 of that is idling while talking with my girlfriend, chatting online, eating dinner, etc. It’d skew results.

I could see there being a section under the quick voting links on IFDB - along with “I’ve played it…” drop down a list:

I played for approximately:

  • 15 minutes or less
  • 30 minutes
  • An hour
  • Two hours
  • Between two and four hours
  • Longer than four hours
    Do you feel you reached the game’s conclusion?
  • Yes
  • No

Average playtime/interest is probably also as valuable as “time to completion”. You could show the stats on the main page:

  • 10 IFDB Users report playing this game. Average playtime reported is (30 minutes). 6 Players reported reaching a conclusion.