Red Faction 1 rules discussion
7 years ago
Montana, USA

Hey guys, I just want to post a little thing addressing some of the issues we are having with the rules for this game regarding timing, fps, etc.

First, this is only regarding single segment runs.

1: Version Use purefaction, preferably the latest version. The game doesn't run properly without it.

2: in-game FPS We can go two ways here. We can either allow changing of fps during the run, or decide on a set limit or range of fps to do the entire run at. Changing the FPS can shorten the length of cutscenes and postiviely(or negatively) influence enemy AI.

The game runs "best" at 55-60 fps, so, for the sake of consistency across a full game run, I vote for using 55 or 60 fps and NOT switching during a run.

Really though, whatever you guys want to do here is fine. I don't care much at all what we decide on the FPS issue.

3: Real time (RTA) vs In-Game Time

This is the biggest sticking point.

For the sake of clarity, we should define the RTA timing as beginning as soon as the player takes control of Parker after ending the opening cutscene and ending as soon as the bomb has been solved and the player loses control of Parker.

In-Game time is being defined as using a timing program that stops timing during loading screens. This is useful as it gets rid of small inconsistencies relating to load times across varying players' hardware and across any certain players' individual runs. This will result in "lower" times, but that's just a categorization issue--the run isn't really any 'faster," it's just timed differently. The starting and ending of timing for an in-game timed run will obviously be the same as described previously for RTA runs.

REASONS I DISLIKE IN-GAME TIME: This "in-game" timing is done by using the LiveSplit timing program. That's all fine, as one can set up live split to time both real time and in game time. Hypothetically, we could just have everybody using livesplit to record both and use both times for the leaderboards. Right now, in theory, real time would always be "slower" than ingame time and vice versa, but what if something dumb happens and that's not the case. No one wants to deal with that. Or let's say one players Ingame time is the fastest, but the real time for the that particular run is "slower" than another players real time. What do we do there?

I understand that cutting out load times makes for "faster" times and helps even out the SMALL inconsistencies in load times across players and their hardware...but I still dont like it. It requires you to use livesplit. Like...let's say i dont want to use livesplit--how am i going to get my in-game time? Also, let's say one day livesplit isn't supported as a timer, what do we do then?

Using Real timing I think is better in the long run. If someone thinks their computer isn't as "fast" as someone else's, and as such, feels they suffer a time loss from longer loading times then...tough shit right? That's just how it goes. That's how things have been going in speedrunning forever. You start timing when you get control and end when you lose control. Load times might differ slightly, but we wouldn't be relying on an EXTERNAL program to cut those out. Speedrunning, to me at least, isn't necessarily about manipualting time soruces to get the lowest time value--it's about hitting start on a game and beating faster than others. If you're gear isn't good enough to mesh with better times from others with better gear than tough shit. If i want to speedrun the new Doom right now, my times are probably going to be slower than another player with better hardware but I'm not gonna say that the timing method should change.

Red Faction is an inconsistent game sometimes regarding load times, enemy ai, etc. And that' because the engine is jank. But I still don't think we should rely on livesplit's ingame timer function as a way to, in my opinion, falsely lower times.

This is all just my opinion. It's really up to you guys to agree on whatever. I dont even really mess with single segment runs. We could even have separate leaderboards for RTA's and ingame times. Please discuss and give your thoughts so we can nail down a consistent rule set for everyone to use.



I don't think we should have separate leaderboards for IGT and RTA, that just seems excessive. I think we should all run on the same version and cap the FPS at 60 and keep it there throughout the run, I don't think there are any tricks in particular that require FPS changes so staying at 60 sounds good just to keep things consistent. As for using IGT or RTA... I'm not sure what the differences are on PC (I'm on PS2, running PC later) in terms of loading times across different hardware but I think a general rule of thumb should be Vast Differences = IGT and Negligible Differences = RTA. Just my thoughts.


Bringing this up for the timing: The autosplitter for RF stops the timer when the last cinematic starts, and not when you lose control of Parker. It's about 1.5s of an delay before the cinematic starts. I checked some older runs and they stop the timer when the bomb is finished... so I guess that is the way it should be. We just have to figure out exactly how much time there is between the lost control of Parker and the start of the cinematic, and it will be easy to subtract that from whatever the final time is.

My opinion about the fps: Use 55, since that's the highest fps that allows for proper cut scenes in the run. 56 and higher adds time to the cut scenes, and the difference is quite great (1.5 seconds extra on the capek death cut scene between 55 and 56 fps). If you want to avoid lag, run at a higher fps, but not allowed to switch the fps during a run or during cut scenes plus the time loss on the cut scenes adds in to your final time. Any fps changing during a run should disqualify it.

Sxnclair, Coffee Hero gains 9 seconds on me due to loads. That's a vast difference for this game. The loads themselves might not cost much individually, but since there are so many of them, they add up. I am kind of in the middle here, since I'm the one at a pretty heavy disadvantage, I should probably root for igt but I honestly don't like it either. The thing is, I don't like being dependent on Livesplit to know my in game time, since I do agree with what Duane is saying, and this is the way I've been trying to improve for since I started. Plus this is the way older runners timed their runs too. Two rankings could work though I guess even if it may seem excessive, or if it was possible, to display the in game time next to the rta time or whatever. If someone would have a better igt than someone else who has a better rta, the rta is the one that's being ranked higher, but displaying the igt would make it even for both players in a sense lol.

Edited by the author 7 years ago

9 seconds in my opinion is too big a difference to use RTA, I assume those seconds are because of better hardware which is nobody's fault but it's still unfair. I don't see why we can't just rely on Livesplit IGT, that's what most runners use and I've seen categories before that pretty much demand Livesplit to handle the timing. If there is ever a point where Livesplit isn't the best option then we can always adapt.


Likely differences in hardware, my machine is not exactly "latest news" anymore. 9 seconds is indeed a lot. I know there are several games that use live split to pause the timer during the loading screen and have that as a set rule.

I guess we can see this both ways, as in: Your loading screens are slow due to your hardware sucks and you don't have the money to update it, tough luck bro. Or: You don't want to use free software (Livesplit) to time your run to know your in game time, tough luck bro.

Looking at it like that, common sense tells me that Livesplit in game time timer is the way to go. It just feels kind of shitty to "cheat" down the wr like that, but then again, it takes zero skill to watch a loading screen (which can be argued for the cut scenes as well, to bring down the wr even more, but cut scenes are constant unless you are stupid enough to run at a higher fps, unlike the loads).

Or just have two rankings... so both IGT and RTA gets credited.

Edited by the author 7 years ago

At the end of the day Livesplit is available to 99% of people and more cash for better hardware is not (if only am I right) I'd rather the rules be fairer so that the we get a more solid Leaderboard that people can get behind, even if it means being at the mercy a little bit of an external program.


Hello, I think for the any% single segment leaderboards, both the RTA time and and the in game time should be submitted. I have seen on several games where on the leaderboards you submit the time with the loads and without the loads. I do not know how you could add this option when submitting a run but I am sure it could be figured out.

Montana, USA

I'm leaning towards tracking both times for runs if everyone is going to use livesplit

"Sxnclair: At the end of the day Livesplit is available to 99% of people and more cash for better hardware is not (if only am I right) I'd rather the rules be fairer so that the we get a more solid Leaderboard that people can get behind, even if it means being at the mercy a little bit of an external program."

It's true that livesplit is free and available vs better hardware being NOT as available. The thing is, though, that "better" hardware doesn't necessarily equal better loading times for Red Faction. The engine and the way it's in-engine timers control events and loads and triggers, etc, is jank and doesn't necessarily always get faster on "better" hardware.


Yeah, having both will probably be the best. There really is no point in replacing rta with igt and I would not like that.

I would like to think that the loading themselves gets faster on a good machine :)

Figure out the loads would bring down the rta another 10ish seconds in my case seemingly. It's getting dangerously close to 54:xx rta being possible, so some more time savers and a "perfect" run...:P


I agree with what Kotti said, you do not have to rely on LiveSplit 100% for the IGT, you can use a video editor or time them (which would take a while but yeah)


Good arguments Kotti. Looks like its leaning more and more towards IGT afterall.

Some more: I looked it up and the autosplitter has an 1.8's delay from when you lose control of Parker until the timer stops. I suggest, when using the RF autosplitter, that we should subtract 1.5's from the final time. It's the same for both RTA and IGT. Not subtracting the whole 1.8 just ensures that we are not subtracting too much time, and also considering human reflexes to stop the timer on an actual run. Plus 1.5 is a solid number and makes the most sense :P


We can have the Leaderboard display both RTA and IGT although the focus and the actual time of the run should be IGT, it's just more accurate and fairer for everyone. Regarding the autosplitter, I think we should just remove the full 1.8 seconds delay as this will give us the purest most accurate time of the run we made, if people don't want to use the autosplitter then they'll just have to deal with the inaccuracy's that comes with splitting manually.

Montana, USA

"The only argument for using pure RTA is laziness and that's not a good argument."

I respectfully disagree. RTA isn't "lazy" or "inaccurate." It's just a DIFFERENT way to measure time.

And timing to the millisecond is important, in my opinion. The fact that different machines might have tiny variations in load times doesn't mean that milliseconds aren't important, or that RTA is inaccurate.

As far as the games "auto splitter" and all that nonsense, isn't possible just to hit STOP on the timer when you complete the bomb puzzle? Like...manually? Rather than rely on livesplit to stop it for you?

There's plenty of games that are measured RTA that have load times. And those load times aren't consistent across hardware. Wolfenstein: TNO, BioShock, etc. Just because it's possible to use an external program to remove load screens, which is just stopping the timer when the game is sending pause-frame data, doesn't mean it should be used or needs to be used.

Here's a far fetched example--let's say that using livesplits IGT pauses the timer when the player pauses the game or hits escape and is taken to the menu--I dont know if it actually does, but let's say it does for the sake of this example. So, if a runner flubs a keypress and enters a menu or pauses the game, that doesn't count towards the timer, so there's no consequence for making that mistake, whereas an RTA would punish the runner for that, correctly. Just a dumb hypothetical situation, probably would never be an issue.

I'm totally cool with whatever everyone wants to do--I just want a little deeper discussion than "I like in-game time, it cuts out loads and makes the time look better."

Montana, USA

oh and regarding this from surrealistic infinity:

" I looked it up and the autosplitter has an 1.8's delay from when you lose control of Parker until the timer stops. I suggest, when using the RF autosplitter, that we should subtract 1.5's from the final time. It's the same for both RTA and IGT. Not subtracting the whole 1.8 just ensures that we are not subtracting too much time, and also considering human reflexes to stop the timer on an actual run. Plus 1.5 is a solid number and makes the most sense"

Just have a hot key to stop the timer when you complete the bomb. There's no reason to just be subtracting time off the autosplitter time or whatever. IMO. Or if manually stopping the timer is too much of an issue for you (back in my day we used non-auto splitters, non-igt timers and had manually start and stop everything. And we had to walk barefoot to school, backwards, uphill both ways, ya damn young'ins!) we need to find a way to fix the auto splitter.

Montana, USA

You're right, I was wrong about bioshock.

Did my humor get missed as far as purposely ridiculous reminiscing about the "old days"? Sometimes my sense of humor is too dry and doesn't come across well in posts.

I like RTA because I like the idea, especially within single segment runs, that it's all about how fast you can sit down and beat a game. Like competing against the clock on the wall. I don't even do single segment runs often. I got into segmented running specifically to do things like cut out loading screens, edit segments together and spend an inordinate amount of time obsessing over tiny details.

If everyone wants to do in-game time and rely on livesplit and all that jazz, that's fine with me. I'd rather it be RTA but that's not because I disagree that there are benefits to using ingame time and external tools to assist with the timing--I totally realize those benefits and how said benefits even out the playing field for everyone. That's fine.



I didn't think about using a hotkey to stop the auto splitter. That will work as well, at least for me. Some guy named SuicideMachine (doesn't sound worrying at all haha) made it. I have no idea who that might be but I was surprised that there was an auto splitter available for RF, and why not use it when there is one available. My guess is that nothing actually happens for 1.8's when you do finish the bomb even if you lose control of Parker, so the auto splitter had to use the last cinematic to stop the timer. I guess I'm a lazy young'in but the auto splitter is very convenient, times change ya old dinosaur ;)

As of right now, I have set my splits to be RTA, and the big timer as IGT, so both are tracked at the same time on screen.

Kotti, maybe thats because no one has submitted both times? What happens if both are added? Braydon's is without loads though so that one is correct at least. Braydon's RTA was 57:12.9 for that run, although that one is not his PB anymore, maybe he haven't just submitted the new time yet. My IGT stands at 55:00.9. I have no clue about Turntohard's IGT though, he never told me when I asked him about his IGT before.

Edited by the author 7 years ago