[PROTOTYPE] Great Rescalening: a "new" movement base

[Stolid]

Discussion about movement has slowed down somewhat, but many people are still not entirely happy.
The main critiques are towards ADAD spam that is too fast, map travel that is too fast, dodging that is too slow and some general mushy behavior when it comes to movement especially around edges in map geometry.

What this mutator is
This is a proposal to become the new base movement in the standard game as a point to start tweaking from. This mutator changes the movement in such a way as to match UT99 properties. This doesn’t just include speed, height and length, but also duration of jumps and dodges as well as dodge delays.

What this mutator is not
This isn’t a way to go back in time or to say UT99 is superior. It’s not a 100% emulation of UT99, all the new ways to move like walldodge, wallrun and kneeslide are still in.
This also won’t magically make old ported/remade UT99 maps play perfectly. If a map is scaled by a different factor than this movement, it will be out of scale by that amount.

Why
There are 2 main reasons for making this new movement mutator.
First, it’s to improve the movement by addressing the aforementioned issues people still have with the current movement.
Second, it’s to serve as a way of excluding movement as one of the variables in the core gameplay. By using movement from an older UT game that has shown to “work” historically it will aid in pinpointing where other problems are. Be it the weapons, the maps, the animations or the netcode.

The Challenge
This is easier said than done. While time consuming it’s not especially hard to copy the same properties over, but doing that and not break many (any?) maps requires some more thought.
Here are some movement properties and their scale factor to illustrate the problem.

Running speed	2.14 times UT99
Acceleration	2.42 times UT99
Jump speed	1.86 times UT99
Jump height	1.85 times UT99
Dodge speed	2.14 times UT99
Dodge length	2.48 times UT99
Gravity	1.87 times UT99
Character height	2.36 times UT99
Character width	2.47 times UT99

Things are all over the place relative to UT99.

The main thing that can break maps is the dodge length, if you make it shorter you can’t reach certain places anymore. If you make it too long you either shoot right over things or are able to reach places you aren’t supposed to. It would make sense to keep the dodge length the same and scale everything else by the same factor used for it. However as you can see it happens to be the property that is scaled up by the biggest factor compared to UT99. Using this factor for everything would mean acceleration increases even more and so going against the first goal of slowing down ADAD movement spam.

By only picking and choosing what things to scale by this factor would mean to go against the second goal, so this isn’t a viable option either.

There doesn’t seem to be a great way to do this well, but from my previous movement adventures I happened to remember there was one other thing that is different in the current movement compared to previous UT games and luckily it opens the door to make this happen after all.

Dodging in UT maintains your (running) momentum. Dodge speed is and has always been 1.5 running speed. Pythagoras would tell you that means your dodge speed is ~1.2 times your base (standing still) dodge speed if you dodge perpendicular to your running direction.
This isn’t the case in UT4, there is an artificial cap on dodge speed set to 1.11.
We can open this back up to 1.2, scale the dodge length back so that it matches that old 1.11 length and we’ve got a dodge that doesn’t break maps. Since you can almost always run before dodging you should be able to get to the same places you used to.

Instead of a scaling factor of 2.48 we now have one of ~2.29.

Scaling everything by this factor achieves both goals at the same time.

A third thing this achieves, though it wasn’t a goal, is that it very easily allows for porting UT99 maps. Scaling it up by the same single factor this movement has been scaled by means the map plays the same way it used to, not accounting for the new movement possibilities in this game of course.

What has changed
I ended up having to change 55+ values so I won’t go over them all.
What you can expect in general is faster running speed, slower acceleration, higher jumping, faster dodging speed, lower duration of dodges and more delay between dodges.
While dodges shouldn’t break maps there is a possibility some jumps (lifts jumps in particular) could allow players to go up to places they shouldn’t be able to.
The list of scale factors above gives some indication on the amount they changed by.

Basically what has changed is how current UT4 differs from UT99.

Some additional things I’ve changed are giving the character capsule a flat base, changing/improving water behavior (getting out of it in particular) and speeding up Sprint.

Download
https://ut.rushbase.net/Stolid/99HCM...wsNoEditor.pak
https://multiplayerforums.com/files/...core-movement/

Put the file in Documents\UnrealTournament\Saved\Paks\DownloadedPaks folder. Create it if it doesn't exist.
Create a game from the menu. Go to the Custom tab.
Select the UT99 Movement Properties mutator in the list and move it to the right.
Click Play to try this movement out offline

Or add
?mutator=/Game/Stolid/Movementv01/HardcoreMovement.HardcoreMovement_C
to the (server) command line.

Updated
Use the new command line above.

Update September 1st 2016
I ended up with changing the height of the mesh to 105% from 115%, but kept the widths at 115%. This makes all characters noticeably more bulky, IMO closer to the look of previous games.
Eye height lowered back down to where it was.
Animation speed reduced by 25%. This makes opponents movement appear a lot less chaotic and easier to track.
I also decided to move away from the slower deceleration that has been in the game for quite a while now. It is now back to where it was in previous games.

Update March 2nd 2016
Hacky fix to make liftjumps go the same height as they do on stock. This mainly effects superfast lifts like the one on DM-Chill to armor vest.
Reduced the eye height to be in between where it used to be and where current stock is. Since I still can't reduce the character height I don't want to go lower as opponents will look like giants compared to you, but this does make first person speed feel closer to where it has to be again.

Update February 10th 2016
Sprint put back to default for now
Impulse imparted by weapons now launches players up to the same height as in the default game, this makes rocket jumps etc work as before.
Dodge delay between dodges slightly reduced.

Update December 12th 2015
You can now jump up onto 200 unit boxes again.
Walldodge reset time has been significantly lowered, it’s now in between the default UT4 value and the previous value in the mutator.
Sprint speed slightly lowered and delay + acceleration from run speed to sprint speed has been changed back to the default UT4 value.
Impact jumps now give the same height as they do in default UT4
Walldodge speed has a max horizontal velocity slightly below the default limit.

[Tobe`]

So much better than the current movement. The movement feels much more grounded and controllable and the walldodges dont feel as overpowered anymore. This would be a great base to start tweaking from. Thumbs up!

[timbur]

Great initiative!!! I would say that UT99 is not only the best place to start in terms of movement/scale, but it's also the best place to start in terms of the overall design of the game, including menus, scoreboards, HUD, the overall code structure/patterns/architecture, etc., etc. We already know almost everything works beautifully and seamlessly in UT99, so it makes sense to start there and use that as a base, and then improve and build on top of it in a way that meets modern demands and expectations, instead of attempting to make every design decision from the ground up, which can get extremely tiresome after a while! I feel like if this was done from the beginning, we'd have a much more solid base to work from. Plus, the 90's are kinda making a comeback lately.

[MoxNix]

Speaking from a mapper's point of view, when it comes to porting UT99 maps a scaling factor of 2.29 doesn't align to the grid at all. 2.1875 and 2.34375 align quite well.

With the current official movement, 2.34375 is the scale most players say plays most like the original.

The main critiques are towards ADAD spam that is too fast, map travel that is too fast, dodging that is too slow and some general mushy behavior when it comes to movement especially around edges in map geometry.

Agree with that. In order of how much of a problem they are I'd rank them.

1. ADAD spam
2. Dodging is too slow
3. map travel is too fast
4. Mantling is lame

Edit: Having tested it now my initial impressions are

1. Dodge is snappier though the cooldown between dodges did seem a little too long.
2. Sprint is insane, it's like warp-speed now, really fast.

Snappier dodge makes movement feel crisper and more precise. I didn't notice any difference with ADAD spam but that's more of a problem when the guy you're shooting at does it and I was just testing solo. Never really tested jumping/mantling either, simple bunny hopping type jumps felt fine but I didn't try anything like the flak crate in Tempest (the one you could jump onto with hardcore on but not with it off).

Edit2: Tested jumping on another map and found it breaks a bunch of jumps. Jump height is lower, some jumps are no longer possible, the entire map would have to be rescaled vertically to fix it. Also lift jumps and lift+boots jumps weren't quite as high either (the jumps were still makable but more difficult to pull off). Regular boots jumps (without lift assist) seemed almost the same.

Testing was done on my own DM-Codex remake ported at 2.34375 scale and my original map DM-Octagon (With official movement Octagon has several just makable "easy" mantling jumps as well as a bunch of lift jumps, boots jumps, lift+boots jumps and other more complex combo trick jumps).

[Stolid]

As far as mapping goes it's not entirely ideal, but then the current situation isn't either.
With either of those scaling factors you won't nearly be able to jump as high as you would in UT99 on that map, though artifical boosts like mantling can help to pull you up onto objects higher than your jump height, but that effect is one of the things that's disliked by many as it feels disconnected and sometimes unwanted. The dodge is the opposite, it's relatively longer even on the bigger of the two scales. There is no way to make it play the same, at best you can make it sort of OK at the moment.

However, if you keep one version around that is aligned to the grid and use the map scaler on it to bring it up/down to 2.29 when you are ready to publish it, you are able to match it up perfectly when using this movement.
https://forums.unrealtournament.com/...ueprint-Editor

DM-Lea is already done this way.

[Infiadux]

Stolid, will you please marry me? <3 <3 <3 <3

This is great. The dodging feels perfect, it's actually fun to just run around a map now. Strafing feels a little slow, but still good, and shouldn't be overpowered anymore. Jumping feels slightly off, but I think it's just because I'm not used to it and at least having it like this differentiates it from dodging. Thank god for that capsule not snagging on things anymore. Slide and wallrun are faster (due to faster dodging I presume), which is a welcome change.

The only thing that slightly bothered me when testing this was the longer dodge reset interval. In the beginning of UT4 development I actually wanted it to be longer, but over time I've grown used to the shorter interval of UT4 and started to enjoy it a lot. However, I guess this is a reasonable change as it keeps map traversal speed similar to what it was before while allowing dodge velocity to be faster. Might be just a matter of getting used to it again (it's bothersome for the reason that I kept double tapping too early :/)

I hope Epic implements this or at least uses a lot of these tweaks as a base. Actions speak louder than words, and thankfully we have a great community member like Stolid here who actually made a working mutator. No reason not to try it now.

Also, someone please put this on a hub

[BigFatErik]

I tried it for a few minutes against bots. Dodging feels snappier and more enjoyable - sliding feels a little quicker for some reason too.

The "sprint" feature that's already in UT4 is absolutely bananas though. I put a couple bots on novice, and they were zipping around the map like sonic the hedgehog. I tried it too, and I was getting around FAR quicker than normal dodging could take me, especially up ramps and such.

[Stolid]

Edit: Having tested it now my initial impressions are

1. Dodge is snappier though the cooldown between dodges did seem a little too long.
2. Sprint is insane, it's like warp-speed now, really fast.

Snappier dodge makes movement feel crisper and more precise. I didn't notice any difference with ADAD spam but that's more of a problem when the guy you're shooting at does it and I was just testing solo. Never really tested jumping/mantling either, simple bunny hopping type jumps felt fine but I didn't try anything like the flak crate in Tempest (the one you could jump onto with hardcore on but not with it off).

Edit2: Tested jumping on another map and found it breaks a bunch of jumps. Jump height is lower, some jumps are no longer possible, the entire map would have to be rescaled vertically to fix it. Also lift jumps and lift+boots jumps weren't quite as high either (the jumps were still makable but more difficult to pull off). Regular boots jumps (without lift assist) seemed almost the same.

Testing was done on my own DM-Codex remake ported at 2.34375 scale and my original map DM-Octagon (With official movement Octagon has several just makable "easy" mantling jumps as well as a bunch of lift jumps, boots jumps, lift+boots jumps and other more complex combo trick jumps).

Coming from UT4 the dodge delay will indeed seem long, but it simply matches that of UT99.
Sprint is indeed very fast now, I buffed it quite a bit just to see where the point is where people would actually use it. I have no problem changing that again.

Jumping is actually considerably higher (~25%) now than it was, if you could make jumps before that you can't make now it's entirely due to assists the game gave you. A UT99 map scaled by 2.34375 would indeed be able to give you that problem, as 2.29 would replicate the exact size. A box will be ~2% too high. This is not so much a problem of the movement as one of the map scale and movement scale not matching up.

[RPGWiZ4RD]

I like the slide. I'm not personally sure about the dodge height but speed-up dodge is needed. The ADAD spam I think might be slightly bit too nerfed, a slight boost would still be welcomed but I think slight bit weaker than currently in the build wouldn't be possibly a bad idea. The jump height is interesting in one way but I'd feel slight air control boost would make it more useful, jumping has always been very weak evasive move in combat because of how predicteable it makes you and this added height makes you actually more predictable but if you would raise the air control a little bit it might start be worth using.

The reduced dodge height makes you feel very heavy again and limits the wallrun possibilites further. I'd want to go the opposite even of making dodge->wall run useable in more than those places of being next to an edge or a downwards slope where added dodge height can help.

Sprinting I'm not sold on, it kinda affects the scaling of maps making you run from end to end much faster and makes the game feel overall quite small but it's kinda good to have a reason for not constant dodge spam in there too. Just don't think there's any golden middlepath which only makes it a pro, it's both a con and pro.

Nice to see some movement brainstorming/testing/focus action though.

[MoxNix]

Jumping is actually considerably higher (~25%) now than it was, if you could make jumps before that you can't make now it's entirely due to assists the game gave you. A UT99 map scaled by 2.34375 would indeed be able to give you that problem, as 2.29 would replicate the exact size. A box will be ~2% too high. This is not so much a problem of the movement as one of the map scale and movement scale not matching up.

If jumps were 25% higher than before it wouldn't be a problem, in fact with 25% higher jumps it wouldn't even require mantling to make these jumps anymore.

I tested with the exact same map at the exact same scale both with and without your mutator. These jumps are makable with current official movement (with mantling) but not with your mutator. To be precise the jumps I'm talking about are normal unassisted (other than by mantling) jumps that are exactly 200 uu up from a flat floor.

And before anyone suggests I'm crazy or it's just a problem with my map take a look at DM-Temple. Most of the boxes/crates in Temple are exactly 200 uu high too. You can jump onto them directly from the floor without this mutator (no wall-dodge required either) but the same jumps cannot be made with the mutator.

[Flikswich]

I'm not a huge fan of the movement values for various reasons as it feels a little excessive to have such fast dodging in a game with multi walldodge. Also it feels a bit too horizontal to me and sprint is a little, err ridiculous.

The collision changes are absolutely amazing though and I hope epic realize that having a flat base + slighly higher jump is by far and away the best solution as the current capsule does all sorts of bullsiht when you get knocked into a walkway and bounce vertically or slide like some kind of helpless predictable moving target along the side of a surface.

[tigerclaw]

Looking forward to testing this, can you drop a full list of changes here so i can go over it(not got ue4 installed)

I went over the movement to death in the first few months of development, balance is the main issue

One of the problems i couldn't easily get around was the 'floaty' feeling from any kind of 3d movement - you can tweak dodge axis speed and gravity to feel good but really the issue i found is that you glide through the air and fall too slowly

Upping the gravity, combined with a high accel rate so that you can begin to move in a direction isntantly was a good start

Putting the game speed up by even 0.05 was enough to make this feel considerably better

I never found a complete fix to this but was able to make it a lot better in weapons, you don't notice these issues so much in IG which is mainly what i've been watching/playing but in weapons the slow air movement is really noticeable

-.-

[biggamer21]

The dodge is the staple of UT, there's no reason not to make it important along with wall dodges, especially while eliminating the random strafe-fest the game has progressed into. It doesn't generate sound game flow at all.

Good job Stolid, it's nice to see someone who's been around for all the games and able to make a unbiased judgement on what works and what doesn't based on that experience and make something very productive.

Right now the core of the game seems all over the place, starting here like others have said would be a great choice, because let's face it - it can't get any worse.

[Stolid]

If jumps were 25% higher than before it wouldn't be a problem, in fact with 25% higher jumps it wouldn't even require mantling to make these jumps anymore.

I tested with the exact same map at the exact same scale both with and without your mutator. These jumps are makable with current official movement (with mantling) but not with your mutator. To be precise the jumps I'm talking about are normal unassisted (other than by mantling) jumps that are exactly 200 uu up from a flat floor.

And before anyone suggests I'm crazy or it's just a problem with my map take a look at DM-Temple. Most of the boxes/crates in Temple are exactly 200 uu high too. You can jump onto them directly from the floor without this mutator (no wall-dodge required either) but the same jumps cannot be made with the mutator.

That just shows you the "power" of mantling. The default UT4 jump height is 123.7uu, the fact that got you up on a 200uu box means it boosted you up more than 75 units, more than 61%(!) higher than the actual jump height. It dwarfs the already very signifcant 25% higher jump.

If you look at the box right next to the 2 health packs in the lift area on Deck16 in UT99 you'll see it's 64uu high. The jump height in UT99 is 67.3uu so you can just get on top of that in one go. The jump is 5.1% higher.
In UT4 that same box is 150uu high (scale factor 2.34375). The new jump height with this mutator is 154.2uu so you can still just jump on top of it. The jump is 2.8% higher.
The jump is relatively lower compared to that box in UT4 because of that scale factor. If you were to make that map using 2.29 the box would be 146.6uu high and the jump would be 5.1% higher again.

For a jump to not work anymore with this mutator (>154.2) while using a scale factor of 2.34375 the box would have to have been 66uu in UT99 which could be considered too high to be reasonable anyway.

That isn't to say I won't look into making it possible to jump on 200uu boxes again. It depends on how it interferes with the stated goals of this mutator. Even if it won't be possible I don't think its a huge problem, 200uu boxes are almost exclusive to temple at the moment and they aren't a huge part of the gameplay. There are also alternative ways to still get on them.

I'm not a huge fan of the movement values for various reasons as it feels a little excessive to have such fast dodging in a game with multi walldodge. Also it feels a bit too horizontal to me and sprint is a little, err ridiculous.

I scaled the walldodge speed up by the same percentage the normal dodge was increased by. Perhaps keeping it a bit lower would be more reasonable. Alternatively while not currently possible with just blueprints it could be made possible to simply limit the max walldodge speed. Much like the normal dodge speed being limited to 1.11 in the current game as I talked about in the OP.
Walldodging and sprint are some of the few things that aren't directly scaled (and can't be as they aren't in UT99) and should be considered experimental so those in particular are open to changes.

One of the problems i couldn't easily get around was the 'floaty' feeling from any kind of 3d movement - you can tweak dodge axis speed and gravity to feel good but really the issue i found is that you glide through the air and fall too slowly

I simply calculated the dodge height, length and duration in UT99. I matched the duration in UT4 and scaled the height and length by the same factor. There can't be any floating if you go up to the same height and take the same amount of time to fall down as in UT99

[joellll]

Feels nice. I like that sprint is set apart and has a purpose beyond "this one time at band camp sprint accidentally engaged and I moved faster for about 3 feet".