Sub Categories for SF64?
7 years ago
Pennsylvania, USA

Just worth a discussion since I'm going to all the star fox games that could potentially have sub categories and fixing up the leaderboards. It may be that this game won't need them but that's what I want to discuss.

SF64 has an N64 version, a Wii VC version, and an IQue version, each with its own set of lag reduction strats for each. So, would you guys think it's worthwhile making subcategories that specify which console is used? In essence it exactly the way it looks now, just with a 2nd line the shows what Platform was used.

British Columbia, Canada

I think a sub-category for Emulators would be worth having, possibly including iQue in that category.

iQue and Emu have virtually the same lag-reduction strats, which is none. This somewhat breaks the meta-game and, although I respect that some people have better runs on iQue (¤cough Pottoww on Red Line cough¤) , I think it's a bit disheartening to see that you got WR in a game, only to see that you didn't and can't get WR on the console you're playing on because it's so much slower (¤cough anyone not on iQue on Red Line cough¤).

That being said, not dissing iQue. It's a valid console and Pottoww still has the best Red Line run. But I think it would suck if someone got a 32 on console, which would obviously be better than Pottoww's run, just to see that you're 2nd and there's no way to get WR without dropping $100 on an iQue, having to deal with a terrifying Andross and ROB, and basically doing a completely different run.

Here's what I want to see: I want to see emulator on the leaderboard. Straight-up, I'm not big on emulator, but I think it would be a good way to get people interested in running this game. I don't know how many people I've told "emulator is banned" to and they've immediately dropped the thought of running the game. A lot of people don't have the means to run this game ($40 capture card + $40 console + $20 game = $100, not a small amount of money if you have none of this).

Here's what I DON'T want to see: A game dominated with emulator leaderboards. I see a lot of this with another game I run, Mario Party, where the top 20 or so times are all on emulator, and even if you wanna run on console, you can't be even close to competitive because the run is easy and the best time you can get on it is like a 37 (barely top 20). It diminishes the work done by those on console because they see "oh Hayate has a 22:24 Any%, well that's only 16th on the board so he must suck" when the run is fucking godly.

So basically, I'd like emu to be on the boards, but in a sub-category, much alike what was done in Star Fox SNES. Have 2 columns for every category, console and emulator. Console includes N64, Wii(U) VC, and iQue. Emu includes emu. Would require Project64 1.7 or earlier to run, since 2.X is broken for speedrunning in most games, and honestly it's not that hard to find an earlier version online for free. Although I personally lean towards the idea that iQue should be bunched in with Emu, it's a console and it should be treated as such.

tl;dr I'd like to add emulator runs to the leaderboard, but have them separated in columns, much like the Star Fox SNES leaderboard does with warps/warpless/etc.

fling84 likes this
Pennsylvania, USA

So it would be Sub categories consisting of "Console (N64, Wii VC, WiiU VC, Ique) / Emulator (PJ64 <specific version>)".

Console / Emulator (simplified)

I'm more against the means of emulator because saying it's too expensive to get a console to run is never a valid reason to make a leaderboard a mockery, but subcategories make it a bit more friendly. It lets us default to legitimate means of running the game, and allow emulators to be there without hiding them by default. Figure out what emulator would be best to use, and have others respond here so we can see if emulators should be used.

My opinion:

I'd honestly recommend a subcategory for each platform since this game is ran differently w/ lag reduction strats per console used. The only drawside to this is someone could be epic and claim all the console records, which defeats the purpose of having the subcategory. It's interesting to note that some runs state "N64 World Record," so why not make this a thing?

If this can't happen, then I say leave leaderboard as is, but force emulated runs to be hidden by default. I don't think runners who can't submit to a leaderboard are prevented from running the game. Anyone can use emulator to speedrun without submitting to a leaderboard, but a leadeboard is meant to be somewhat restrictive on what runs can be submitted. We don't allow turbo controllers because it's unfair to regular runners, emulation is another reason because it skews times.

tl;dr My opinion is Sub categories for each platform, otherwise keep leaderboard as is but force emulated runs to be hidden by default.

British Columbia, Canada

I'm not really suggesting making the leaderboards a mockery, or if I am I'm doing so unintentionally. That would be putting them all in the same category and it's basically anything goes, which would break the game and a lot of veteran runners would be pissed.

I'm okay with "hidden by default", but then claiming a world record is kind of awkward. Picture this: Someone gets a 32:05 on emulator, claims WR Pottoww gets a 32:10 on iQue. Since emu is hidden, he claims WR because he has it. A fight ensues over "who really has WR".

I don't think people are really that petty, but I could see it happening.

The issue I have with having sub-categories for EVERY platform is that it kinda clusterfucks the leaderboards. You have some categories with 30+ runs (N64 Any%) and some categories with 0 runs (iQue Blue Line). The reason I suggested just splitting it between console and emulator is because that leaves it nice and clean. I 100% see what you're saying, and for the most part, I agree, but it presents to problem of "too many categories".

I like that nobody has seemed to have an issue with at least adding emulator to the boards in some capacity, though, as this was the original intention.

tl;dr don't want to leave emulator hidden as this may cause confusion over who has WR, and not a big fan of splitting boards so that each console has a category, but will accept.

Växjö, Sweden

Emulator and iQue are different things, and emulators are eve faster than iQue. Also we agreed on banning emulator from SF64 a long time ago so I don't see why this is even up for discussion. iQue is official. There are no differences in buying an iQue to get a faster time, than it is to buy NTSC for getting a faster time if you live in the PAL region. This discussion is bugging me so much.

About the argument "we should have different sub categories because you play it differently on different consoles" is not a good argument. Then we should split every single game out there as well. A lot of different games is ran differently depending on console and region. Should be split Star Fox Assault in JP/US/PAL as well since all the three different versions has different runs in theory?

No, if you split it up, I won't be happy. I do not like this idea at all.

British Columbia, Canada

Okay, so we had a very long discussion on this on the Discord, but I'm gonna give a rundown of the points that each person made:

DSX let it known that he was against emulators, but said that he basically didn't care enough and said that other people could decide.

Zallard1 stated that he was against the idea of emulators because they are inaccurate in lag reproduction from consoles, which is true.

Pottoww let it known that he didn't like the idea of having separate leaderboards (as stated above) and said that he would be somewhat satisfied if we could find an emulator which is slightly slower or the same as the current fastest version of the game (iQue). He stated that he wanted the leaderboards to be mixed, but have emulator hidden by default.

Fling84 stated that he was for the idea of having a separate leaderboard for emulator, but not having them mixed because it makes console runners look bad. Zallard1 and I both agreed on this.

Pottoww suggested the idea of having separate leaderboards, but having emulator runs as a miscellaneous category. All people involved in the conversation at this point (myself, Pottoww, Zallard1, and Fling) agreed. The stipulation of this was that we had to find an emulator that ran at similar (within a few seconds over a run) speed to iQue. I am going to compare video evidence of both iQue and emulator using previous PBs by myself and Pottoww to confirm that the run would be of similar speed to iQue, which I will post the evidence here once it is complete.

Ontario, Canada

Thanks for summarizing Lylat. :)

[small]Full disclaimer: I don't currently run SF64. I want to in the very near future, but right now I don't. But I think the emulator discussion is important not just for SF64, but eventually for SF1 as well down the road perhaps.[/small]

Just wanted to also offer my opinion on a suggestion zewing made above -- I think the Console (N64/VC/iQue) vs Emulator separation is enough simply because this is where the real division lies in the community as far as what is acceptable and what is currently banned. Making all platforms their own categories is probably a bit too much as has been said by others already. So far as I know, other then a few lighthearted comments made here and there, there's no real divisions in the community between N64, VC and iQue.

Anyway, moving on....

Repeating a bit of what Lylat said earlier (I think it's important enough to repeat), the whole point of allowing emulator runs to be submitted in any form is to do what we can to help encourage new runners to the game. People who may not already have the hardware to run it on console. e.g. a broke college student living at home, no money to spend, certainly not $100 to get whats needed. But his/her family has a computer and an emulator is free. Telling these people to "just go out and get an n64" is essentially the same as telling them "piss off, you can't run it." :P

At the same time, I don't think any current runners (both those for and against emulator runs) want to really promote emulator speedrunning for Star Fox. Rather (as Lylat said above in one of his posts), we see it as a gateway drug. Getting people interested in the game and eventually getting them to come over to playing it on real hardware. For such a tiny speedrunning community like this, I think our goal should be to remove barriers to entry, not add more.

Keeping emulator runs on a separate, by default hidden (aka. Misc) category does appear to accomplish this idea of not promoting emulator runs nicely. Or as nicely as speedrun.com appears to allow us to do anyway. Putting emulator runs on the same leaderboard as the console runs but hiding them by default is not without problems as Lylat mentioned above.

Other N64 games which allow emulator runs appear to have standardized on Project64 v1.6 or v1.7. We'll of course have to decide as a community what we want to standardize on and what (if any) standard emulator settings should be used.

I can understand the logic behind the idea of not allowing emulators to be used that run significantly faster then iQue, but I would caution that for N64 emulation, we really don't have much choice. It's pretty much Project64 or nothing (I don't think mupen64 is comparable, but it's been quite some time since I last used that one).

I don't think there's a big problem with emulator runs given all this. We just need to very clearly define the requirements for what will be accepted and what will not be. Moderators validating these runs need to do their part and make sure that people aren't trying to slip a cheaty run through the cracks (goes without saying, but there, I said it anyway).

Anyway, just my thoughts on the matter.

EDIT: Also, I feel I should point out that I'm ordinarily quite anti-emulator. I do not run anything on emulators myself and would never do so. I don't play any games casually on emulators either. However, I do recognize that they are some potential new runners only choice for getting into the game.

Edited by the author 7 years ago
British Columbia, Canada

I did quite a bit of testing comparing Emulator to iQue and I believe that the following settings would be the best for runs done on emulator:

Project 64 v1.7 Emulator Star Fox 64 (U) v1.0 ROM Memory Size 4MB Counter Factor 2 All other settings default

Here are some tests, as well as descriptions, that I did:

  1. Counter Factor 1 setting, comparing cutscene times during the Red Line run versus iQue.

Parameters and results of the test were shown on-screen at the end of the video, but the findings found that, not including the Sector Z cutscene, Emulator with Counter Factor 1 would save 61 frames over iQue over the run. A reasonable estimate for the Sector Z cutscene would be around -2 frames, but it was untested, so these results were mostly inconclusive. The fact that the run went to Macbeth after Zoness on emulator would not affect the results. Test found here:

  1. Counter factor 1 setting, comparing Area 6 times versus iQue.

17 frames were saved over the stage on emulator versus iQue, but there is execution bias in this result. Video showed saving 1 frame every time Gorgon (boss) faded in and out of existence, totaling 9 frames in a perfect double-1-cycle fight. No execution bias in this result. Post-stage cutscene was 29 frames faster on emulator than iQue, no execution bias in this result either. Another comparison was done with Counter Factor 2 (not shown) where the stage was 13 frames faster (again, execution bias), the boss fade in/out was the same, and the post-stage cutscene was 34 frames faster, making the stage 47 frames faster with counter factor 2, versus 55 frames faster with counter factor 1. Test found here:

  1. Counter factor 2 setting, comparing cutscene times during the Red Line run versus iQue.

Same as test #1, except with counter factor 2 setting as opposed to counter factor 1 setting. Cutscenes that DIDN'T lag on iQue were slower on emulator, but ones that did lag on iQue were slower on iQue. This differs from the results from test #1 where all cutscenes were faster on emulator. Overall, including the Sector Z, Emulator w/ CF 2 setting was 41 frames faster. Without Sector Z, it was 45 frames faster, which is 16 frames slower than emulator with CF 1 setting. Test found here:

  1. Counter factor 2 setting, comparing movement speed between N64 and emulator.

Basically travelling in a straight line to see if emulator flies faster. Although there is an RNG bias, there's very little difference between the two systems, which could basically be chocked up to lag frames. Basically all the information is in the video, but 2 tests were done, one with barrel rolls and one without. Neither test has either console touching the control stick, boost, brake, etc., and both boast similar results. The conclusion was that there is no difference in movement speed, and if there is, it's within a very small margin of error and could be approximated as equal. Test found here:

  1. Counter factor 2 setting, comparing Zoness between emulator and iQue.

Playing through Zoness to compare lag. There IS execution bias in this test, but not nearly enough to warrant the result. The result was that emulator was 7.57s slower than iQue in this stage. A few of the lag zones are caused by shooting searchlights, which alone causes enough lag to make the two versions equal, or emulator slower. Although I'm certain that some of this lag could be reduced, I believe the lag in this stage would be at least 41 frames, which offsets the decreased cutscene time on emulator compared to iQue. Test found here:

So tl;dr is that, based on these tests, I can reasonably conclude the following:

  1. Basic movement speed is not faster than console
  2. Emulator with suggested settings would be slower than the fastest official version, which is iQue, at the very least, for Red Line.

Based on this, I believe this meets all required criteria and emulator can be added to the leaderboards as a miscellaneous category, unless there are any outstanding objections.