The game in question is Final Fantasy Fables: Chocobo Tales. There isn't much of a community for this game, so I figure I'd pick the brains of the wider speedrunning community.

Up until now, I've been timing runs on the leaderboard by in-game time for any%. The in-game timer is accurate to the second, and the nice thing about this is that I can accept runs that don't have a full video. The DS needs a specialist capture modification to record runs, and this is a niche game, so I like being able to accept runs above 3 hours IGT that just show a short video of their final time. I understand that someone could do a segmented run in this case to reduce their IGT, but it's a trade-off I've been willing to make.

Today, I found out that the credits (which are timed in IGT) vary in length depending on the version of the game: JP takes 9:47, NA takes 11:42 and EU takes 13:14. Unlike most games, the credits in Chocobo Tales are a minigame of their own (you tap the o's in the text to get points), so even though the fastest thing to do in any% is not get enough points to get any prizes, it's technically a gameplay segment.

The only other differences that I know of between versions are text differences between languages, and the inability to get promo cards in the NA version (free cards you can get before the story mode starts so that you don't have to collect as many cards in-game to fill out your deck).

In light of this knowledge, I have a few options:

1. Keep timing any% runs by IGT, and recommend that runners run on the JP version for the fastest time. This preserves the purity of the run at the expense of putting Western runners off unless they import another version.

2. Change any% timing to RTA up to when the game fades to black just before the credits, and remove shoesockgaming's run from the leaderboard since they don't have a video of their complete run. This goes a long way towards equalising different versions, but removes a gameplay segment from timing.

3. Take a fixed amount of time off a runner's final time based on how much time they lost to JP version credits. It feels more "fake" than option 2 to me, but I would be able to keep shoesockgaming's run because I can tell that it was done on the NA version.

I'm not clear at all on which option is the best right now, so I'd appreciate your thoughts!

Option 1 is what I would go with since it just seems like if someone wants to save time, they should use the fastest version. If there aren’t any differences aside from text and a few things like the codes, it seems like that’s the best option. Using JP for the fastest time is hardly a new concept to runs, so I see no issue with that.

If the game is hard or expensive to import then that can kinda change things. In that case, you could create a variable to divide the two languages (or three if you have a lot of EU runs) to keep them a bit more isolated. But that’s really up to your discretion about whether you want to take that approach to the LBs.

Option 1 feels the most "pure" to me, but Option 3 I feel would be easier for most people to work with (they still don't need to import other versions or record runs). Option 2 to me feels like an awkward middle ground where people are forced record runs even though the version of the game doesn't matter (if that makes sense?)

Obviously t's up to you overall; that's just my opinion :)

As a runner of the game, let me give my 2 cents

Personally I've always been a big fan of RTA and if it makes this much of a difference I'd be very okay with just going fade-out into the credits. It solves the issue without fundamentally changing the run (as most runners would need switching to JP which Id say isn't ideal if it's just about the credits). The credits are technically a gameplay section but the strat is "do nothing" for 10+ minutes which doesn't add much to the run and isn't particularly interesting either. It wouldn't change anything about the run except just removing what is basically more of a cutscene then anything else. If the concern is video evidence you could still allow the final time piece as it is now for times above a certain time. If the difference in time is static it wouldn't be too hard to verify either. I'm not a fan of this personally, but then I've never been a fan of not requiring full video evidence on a run in general

1. Is the only notable version difference the credits? If so, I really don't like this option. Having to buy and play a different version of the game to get a competitive time, just so the credits (that arguably shouldn't be included in the timing at all) go by a bit faster seems very strange to me. That said, if there are other version differences that make JP faster then you're already going to need a JP copy anyway.

2. Using real-time and ending timing before the credits makes the most sense to me, especially since as you noted the "gameplay" section that gets excluded is the credits (which is often considered the end condition), and is effectively a 10 minute long autoscroller.

3. Bad, bad, bad. Even with having a very specific reason for how much time get added or subtracted, having time penalties is super weird and confusing, and just make your timing method even worse. I see no benefits to this over just using real-time.

My thoughts on it: Its up to the existing runners / community, but I were to choose one of your options: I'd go for option 1. Timmi explained the reasons quite well in his post.

Also, I strongly advise to not accept runs without full video, even if the final IGT is supplied but the video evidence is partial, it could very well be spliced/faked. Besides that, you'd also be setting arbitrary standards as to how much video evidence is actually necessary in the first place.

Thanks, everyone. I've considered all of your posts, and ultimately I agree with Crrool in that as far as Any% is concerned, the credits are purely an autoscroller. In order to provide a more accurate reflection of the quality of each run on the leaderboard, timing has been switched to RTA, and ends when the screen fades to black before the credits. Thanks again!