Can you provide Paradise Remastered loading footage from Switch 2 for comparison? The improvement varies from game to game and I'd like to be able to quantify it for this game specifically, especially if we're adding new platforms for that reason.

We do have to keep in mind that differentiating by loading speed is a slippery slope. There were already platforms that load faster than Switch; in fact, Switch is the slowest platform by far. Even compared to a base PS4, it gets obliterated. And that doesn't consider that you could put an SSD in a PS4 to make it nearly a PS5 in loading speed. So I'm not sure making that the sole reason for further separating runs is a good idea.

On the topic of platform additions, I’d like to see the addition of PS5 and Xbox Series S/X, or perhaps PS4 and XboxOne can be renamed to just PlayStation and Xbox to account for the various options

Might make sense if we're adding Switch 2. speedrun.com uses a preset list of platforms, so there's no renaming them. There are PlayStation and Xbox options but they're for the original console and get used for sorting by platform on this page. I don't think there's a clean solution here, we'll just have to add every platform that applies.

Lower framerates are better, but the difference between 60 and 75 is minimal. I'm guessing your monitor is 75 Hz and doesn't support VRR, in which case 75 FPS is the only good option. If I guessed wrong, check your monitor's specs and use the guide below. This applies to anyone else reading too.

• If your monitor has variable refresh rate technology (VRR), ideally your framerate would match the minimum refresh rate listed in the monitor's specs, or if that's too low you'd play at the lowest framerate you can take.
• If your monitor doesn't support VRR or you're willing to go below the minimum rate supported by VRR, divide the refresh rate by positive integers, discard any results that aren't integers, and choose the one closest to your target. For example, with a 144 Hz monitor you could choose 72, 48, 36, or 24 FPS.

To coalesce my thoughts from Discord over the past week:

• Given the majority of G25 owners on overtake.gg couldn't get the manual clutch, I don't believe it should be allowed.
• Kleis's tester tried with his G25 and the clutch acted as a brake.
• It's probably possible to get manual clutch with a G25 through specific actions, but I don't know what those actions are. Even if we knew, I'm not sure that's enough to justify allowing it.

The moderators have said they'll ban manual clutch. I'm leaving the DLL up anyway since it'll be useful for anyone wanting a clutch pedal for legitimate reasons, even though the game's clutch is janky. You'll be able to tell if someone's using it from the speed of their shifts.

As for H pattern, it's sounding like it'll be allowed. I'd like to see an annotation variable (at least on levels) to differentiate transmission types, but it doesn't seem like the moderators want that. I'll work on a DLL to make H pattern more usable on controllers when I have the time to do so; it won't affect gameplay, only accessibility/ease-of-use.

Ah I see. Can confirm that's bugged, yes. Sounds like something serverside wasn't updated when the run was updated.

Nope, it's where it should be.

It was initially resubmitted as 3 minutes instead of 3 hours, and I did get a notification about it getting WR on April 6th because of that. Is that what you're referring to?

Small update with some new discoveries. For one, brakeAsClutch (a setting mentioned in the initial post) turned out to be the determining factor in how the transmission_type setting is set. If brakeAsClutch is 1 (brake), which is the default, H pattern will be set with auto clutch. If it's set to 0 (clutch), H pattern will be set with manual clutch. Sequential always uses auto clutch.

Given this, it's interesting to note that once set, it seems that nothing can change brakeAsClutch. Not device detection, not changing controls, not changing transmission type. It appears permanent.

Which makes this post all the more interesting: https://www.overtake.gg/threads/colin-mcrae-dirt-2.6371/page-12#post-409617

And this: https://www.overtake.gg/threads/colin-mcrae-dirt-2.6371/page-11#post-409600

So we have a firsthand account of the G25 with a manual clutch on PC and two secondhand accounts that describe it working on PS3. Sure, it's not the strongest evidence, but it's more than we had before.

Unfortunately, the rest of that forum is filled with people who couldn't get the clutch working, saying it acted as a brake. I'm guessing there's some specific criteria that has to be met to set brakeAsClutch to 0 before anything else happens. Since kleis found someone to test it, I wonder if they can get it working or if it'll just be a brake.

Oh, also, the string for the G25's name is hardcoded on PC, which it isn't on other platforms. I assume this has something to do with its unique clutch behaviour on PC since no other device names are hardcoded. Just more evidence to support it being a thing.

As a moderator, I would like if this mod gets released. I want to test it out for myself.

Done: https://github.com/burninrubber0/Dirt2Gearbox/releases/latest

DLL goes in the game folder.

Let me know if it works. I only tested on the Steam version but that should be the same as disc and cracked (xliveless) as long as they're updated to v1.1.

im personally against of splitting the boards(again).

I think the idea was to have the variable as an annotation instead of a subcategory so the boards wouldn't be split. At least, that's how I interpreted rkatime's suggestion.

A good idea would be to make a Variable in IL leaderboards for H pattern and Sequential/Auto runs, that way we can keep both sides happy.

I like that idea. May as well have 5 options, auto, seq, seq w/ clutch, h pattern, h pattern w/ clutch. Or remove the clutch options if moderators won't allow it.

surely it wouldnt make a big difference if H pattern&manual clutch was allowed only when the game allows it to? if you have a g25 or something, then go ahead.

Enabling the manual clutch is locked to G25, but keeping it enabled isn't. It'll still be manual on keyboard/controller unless the controls are changed. Also, anyone else who gets the savegame would have manual clutch too. So everything I'm discussing here is viable without mods provided you know someone with a G25. All I want to do is remove that restriction.

The questions

1. Do we care about console players?

2. Do we think using advantages that are only available by having specific peripherals is unfair?

The first question is put forth for a few reasons. Yes, as mentioned here, they can't access H pattern without a wheel, but via save editing they could enable the manual clutch and gain most of the advantage. More to the point though, the top players don't appear to have known about the manual clutch option at all, despite PS3 leaderboards being pretty consistently faster than PC. I don't know how to feel about having an advantage like that when their times are the ones we mainly compare to. On the other hand, the speedrun boards have just one console player and their times aren't exactly competitive, so it shouldn't affect that part.

The second question is more straightforward. Pretty much everything listed here is related to steering wheels and how using them makes you faster than keyboard/controller players. This can't be stopped on the in-game leaderboards, but if we wanted, we could level the playing field on the speedrun boards.

Possible rule changes

Let's assume, for a moment, that nobody cares about the in-game leaderboards or being fair to the people on them, and all we care about is the speedrun boards. Realistically, there are two ways to ensure everyone who speedruns the game has the same capabilities.

1. Institute bans on H pattern usage and/or manual clutch usage. This way everyone is just as slow as one another, regardless of platform and peripherals.

2. Distribute ways of making everyone faster. This would be savegames or programs/DLLs that set the manual clutch option. The latter could make H pattern more usable on controller/keyboard too. The rules around mods would have to be relaxed a bit, or exceptions would have to be made.

Personally, I prefer option 2 because I like to make speedruns faster instead of slower. Yeah, it's more complicated than that, but that's what wins out for me. Making a DLL mod for this is admittedly not easy, but luckily, not to toot my own horn here, but I'm an experienced modder who open sources all his work. So making this a reality wouldn't be a problem, and the moderators could inspect any mods before allowing them. To be clear, even though we'd be using mods to make the speed more accessible, these are still in-game options; we're not enabling cheats here. I'll leave that for the boost option that was left over from Grid.

Anyway. I want to hear people's thoughts on this, especially the moderators' thoughts, before doing anything further. So please, spew them out :)

To way oversimplify the issue: there are "hidden" transmission types/modes that are faster than the usual options, and I'm writing this so the community/moderators can make an informed decision on whether or not to allow them.

The exact questions that need answering are listed after this post. If you skip over everything to read them, they probably won't make sense, so please read the rest of the post first.

Transmission types

When you go into the control options, you'll find this:

There are two control options available to users: automatic and manual sequential. Most of the currently submitted runs use the latter. However, under the right circumstances, there can be additional options:

This third control option, H pattern (or "gated"), shifts faster than sequential. This wouldn't be a problem if it were available at all times on all platforms, but while this seems to be the case on PC, on console it's hidden unless a supported wheel is connected. This means PC players (and console players that own a supported wheel) have an acceleration advantage over console players that only have a controller.

Braking technique

Slamming the vehicle into reverse and using throttle and brake together leads to a braking advantage over sequential. I'm not sure how difficult this would be to use in practice since it messes with the steering a bit, but as a standalone maneuver it's fairly easy to do.

Launch advantage

I performed limited testing with the DeJong MXR and found that being able to start the race in 2nd gear provided a slight acceleration improvement over starting in gear 1, which sequential is forced to do. This is somewhat vehicle-specific so I'm not sure how widespread the advantage is.

Auto/manual clutch

Whereas Dirt 3 uses a dedicated fourth transmission type, H pattern w/ clutch, to allow the user to intentionally enable manual clutch control, Dirt 2 uses a separate option that's hidden from the user but is still present in the profile. Normally, this is set to auto clutch, which manifests as a small delay between gears and is often accompanied by hitting the rev limiter if using H pattern. When set to manual, use of the clutch is ironically unnecessary since gear changes are instant. Acceleration increases dramatically as a result. This works with sequential too, with equivalent speed to H pattern.

While I lack a suite of wheels to test and haven't yet checked the relevant code, based on the PC game's action maps I believe manual clutch control is enabled on at least one of the supported wheels, the Logitech G25, but is disabled otherwise. (The G27, Driving Force, and MS SideWinder FFB potentially enable it too.) I also believe the clutch control is rebindable in this scenario. The necessary tools for extracting PS3/X360 files are unavailable and I don't want to spend a bunch of time making them just for this, so I can't check their action maps but I know they differ from PC.

The game also has another hidden setting that sets the clutch pedal to be used as a brake instead of a clutch. Judging by this Steam post, I believe this is the default behaviour and would only be changed if one of the above wheels is used. Fortunately, since there's no need to actually press the clutch it doesn't affect runs.

Given these findings and under the assumption that at least the G25 enables the manual clutch, an acceleration advantage is granted to those who have such a wheel.

Recently, a new strat was used by myself and Demo to set times on Burnout License:

1. At the start of the run, using a bike, start an event at a junction where a car's Burning Route is.
2. Use the load option in the save/load menu to return to the junkyard.
3. Switch to a car, then restart and complete the event.

Because the junction was saved, it will start as the Burning Route despite being the bike route and not having the right car. The Burning Route car will unlock and become usable. In this case, we used the GT Concept Burning Route and switched to the Cavalry to unlock the GT Flame.

Now on to the big question: What category/ruleset should this fall under?

There are some things I'd like everyone to consider here. First and foremost, to be absolutely clear, this is a glitch being exploited to set faster times and would violate the current ruleset of all license runs. This is new to us since no glitches were previously known to affect license runs, meaning the rules were set somewhat arbitrarily. Now, if people want, the rules could be changed to allow glitches in these categories provided they can be done entirely offline. After all, the ultimate goal of any speedrun is to reach the goal as fast as possible. However, in earlier discussions on this issue, there were people who wanted glitchless runs to remain the primary category and for glitched runs to be a separate category, or even multiple categories. So this is up for debate, and I'd value any ideas anyone has.

There's also some confusion regarding whether this particular glitch counts as Paradise Cars Only since it uses bikes for the glitch. It seems obvious until you read the rules, which state, "this is the same as the 'Without DLC' subcategory on the original game's leaderboards." Problem is, the original game's rules don't limit it to Paradise Cars; they read, "DLC vehicles, timesavers, and Big Surf Island access are not allowed." Okay, bikes were a free DLC pack included in v1.4, so that means they aren't allowed, right? Well, not quite, because the original game's Full Game rules also disallow DLC vehicles (except island vehicles) even though it requires the use of bikes. My best guess here is it was supposed to mean no paid DLC vehicles, which would mean bikes are allowed in Full Game, and that would extend to the license runs. It's just that there was never any reason to use bikes in a car license run previously, so nobody did, and that led to the Remastered rules reading as though limiting runs to Paradise Cars is the same as the Without DLC rules. Regardless, it's obvious the rules need to be clarified and I'd value feedback on which direction to go with them—should we truly limit runs to PCO or just exclude paid DLC?

Here are some mods which are allowed in runs. They either fix the game or bring it to parity with console versions. The mod loader included in BPR Modder is required for all of them.

• Core Bugfixes (via BPR Modder) - needed for some players to run the game.
• Unfinished stuff - Various fixes, particularly disabling the delay when leaving the junkyard after changing vehicles, which brings it to parity with Xbox One. Other options have no effect on gameplay. Disabling cursor clipping is recommended for windowed mode users.
• Webcam disabler - Forces the game to not use webcams, shortening event finishes. Brings it to parity with console.
• Save/load adder - Re-adds the save/load feature, enabling bike in car event bugs. Brings it to parity with console.
• Saved window position - QoL mod for windowed mode users. No effect on gameplay.
• Fix lost focus bug - QoL mod which (sometimes) stops alt+tab from losing focus and forcing a game restart. No effect on gameplay.
• Exception reporter - Reports details when the game crashes. No effect on gameplay.

Core Bugfixes also corrects the downforce of the Toy 88 and fixes junkyard category selection rollover. The latter is also an option in unfinished stuff (disable this if using with Core Bugfixes). These would impact runs if they were used, thus the disclosure here, but no since current runs would benefit from them they're acceptable.

Other mods may be allowed if they don't affect the run. Contact a moderator if you're uncertain whether one would be allowed.

Nah they're the same game. Literally identical executables, the first 5 races in Glacier are identical to Tesco, world and vehicles are the same albeit retextured a bit. Not worth submitting a new game in the slightest, just make a new category here. Glacier% or something.

Greetings, dearest Tesco fanboys! Like you, I can't get enough of this game, so can we get a leaderboard for the full version as well? It doesn't have Tesco branding but don't worry, I'm sure we can find a way to add some :P

If you haven't seen it yet, behold:

I'm a little late on this, but although it isn't necessarily a bad idea, I don't think there's any potential ruleset for the category that could be properly enforced—not one where it's still worth doing, at least. The world's just built around shortcuts too heavily. That, plus everything @Ventilo_ mentioned, means I don't see this happening even as a category extension.

I've implemented the 3rd option into the rules after unanimous consensus among runners. This thread will be left open for further feedback until the first run under this category is verified.

I'd like some feedback from runners on this. At the moment, no runs have been done for this category, but there are people planning to run it soon so now's the time to set the rules in stone.

I've thought of three options (if you have more ideas, feel free to share):

1. Restrict the mainland portion of runs to Paradise Cars only and the island portion to BSI Cars only. (In other words, make Full Game runs identical to running Main Game PCO + Big Surf Island BSICO.)
2. Restrict the mainland portion of runs to Paradise Cars, but allow both Paradise and BSI Cars in the island portion. (This is how the original game's Without DLC category works.)
3. Allow Paradise Cars and BSI Cars to be used freely throughout both the mainland and island portions of the run.

What do you think we should do?

That sounds complicated, does it mean BSI cars can only be used for BSI collectibles, or does it mean PC cars can't be used for BSI collectibles and you have to use BSI cars?

It's the same rules as Main Game + BSI. Although, it looks like the BSI rules are meant to match the original game's Without DLC rules, which would mean Paradise Cars are allowed there. Perhaps it's worth having a dedicated discussion about what we what Full Game Paradise/BSI Cars Only to be, because there's never been a run for that, hence all this confusion.

Edit: I've opened a thread.

the current all billboards and all super jumps categories are for the mainland specifically, which is why I think it should be the same for all collectibles (and by extension all smash gates)

So you're saying the full game all collectibles category shouldn't exist at all? Or am I misunderstanding? (I've gone ahead and added All Collectibles/All Smash Gates to Main Game regardless)

Do you think the run should end after all the mainland smashes have been counted by the game (so it shows 400/400 smashes in the menu and unlocks the carbon GT Concept) or would just getting all of them be enough, even if they don't get counted?

All smashes need to be counted for the run to be valid. I know there are issues with them counting, that's the reason this category wasn't added sooner, but if you wanna run it you'll just have to find a way to deal with it.

in my opinion BSI collectibles and mainland collectibles are 2 different things so it makes sense to separate them.

Do you mean the rules for Full Game PC/BSI cars only should enforce using Paradise Cars for the mainland part and BSI cars for the BSI part? I suppose that'd be the best solution to keep it consistent with other categories. I'll add it now.

This sounds more of a Category Expansion thing rather than a category being added to this page, but it is just my point of view.

Moot point, the category was added years ago. I wouldn't be against it being an extension, but that page is a mess and it doesn't seem like that'll change any time soon, so probably not a good place for it right now.

Vehicle handling, particularly drifting, is affected by framerate. The higher the framerate, the more the car will slide. In normal cars, it also slows down more during a slide. This isn't the case in drift cars, but I haven't done any real testing on those. TL;DR lower framerates are better.

Below are the results of some testing from the past couple days, performed using a CCXR with full eBay bonuses on Nürburgring GP Circuit. Each test pass simply consists of going around the first corner as fast as possible. The game is recorded, the recording is gone through frame by frame, and the minimum speed (in km/h) the car dropped to is logged. Minimums were chosen because using the average of each test pass would unduly minimize the differences caused by different framerates.

Full disclaimer, it isn't possible to test this with 100% accuracy, but I did my best and spent several hours doing a few hundred passes to try to average out any inconsistencies.

I'm no statistician, but this should be enough to get some idea of what's going on. The most important thing here is the average without outliers (average of the minimum speeds of each test pass), which shows a clear downward trend in speed as the framerate increases. This is also reflected by the minimum without outliers (lowest minimum speed of any test pass), which follows the same trend.

The maximum (highest minimum speed of any test pass) is notable because it actually stays relatively consistent. Unfortunately, I can't show this easily in data, but the reason is that these fastest test passes don't involve drifting. There still seems to be a slight downtrend with higher framerates, but I can't conclusively say that lower framerates are faster—more research is warranted.

Even assuming high framerates can be as fast as low framerates, the data proves their consistency isn't equivalent. Despite me being objectively worse at the game at lower framerates (due to a combination of heightened input latency and motion sickness), the minimum speed at 15 and 30 FPS was significantly higher than at 120 or 240. I attempted to show this in the rightmost column, standard deviation, with mixed success. The low-speed outliers only existing at higher framerates also supports this. (If you're wondering, 0 km/h is a spin—and I was unable to spin the car at lower framerates no matter how hard I tried. Maybe initiating a drift is more difficult?)

Finally, to touch on the sub-30 FPS numbers, these are below console performance and generally considered unplayable. However, I suffered through it to see whether scaling continues beyond 30. Long story short, they appear to have diminishing returns. Despite this, there were some results that were out of line, like the 121 km/h result at 20 FPS, and so out of caution I have to recommend that framerates below 30 FPS be officially banned from use in speedruns to maintain parity with consoles (and also to be fair to people who can't take the abuse of low framerates).

Anyway, that's all for this loveliness. Let me know what you guys think. Thanks @Night_95 for insisting the problem exists.

Edit 2024-03-10: Apparently all cars have a "skid type" and drift cars don't lose speed because they use the drift skid type. They're still less consistent at higher framerates, but the extra sliding might be beneficial if you're going for high scores.

Also, just to be clear, the decrease in consistency scales linearly with framerate. That means going from, say, 30 to 60 FPS is less impactful to your times than going from 120 to 240 FPS, even though it's visually more impactful. That's why I'd still recommend playing at 60 if you can't handle 30. Funnily enough, that also means ultra high framerates like 1000 FPS could make the game literally unplayable.