New feature: unit acceleration

[Stan`]

4 hours ago, chrstgtr said:

This really feels like a failed experiment at this point, and I don’t understand why we are rehashing it when there is so much other stuff that the player base actually wants 

We are “re-hashing” it because @bb_ introduced acceleration.

Also because sadly the “player base” isn't entirely represented on those forums. 

[Micfild]

Hi!

I was reading the diff and i have a question and a few observations.

Question:

What is the main goal of introducing acceleration? (is it to bring unit movement closer to reality or to smooth out movement in sharp angles or something else?)

Observations:

They way acceleration seems to work in the code is:

A variable in defined in the templates called <InstantTurnAngle>. For any turn angles smaller than this variable, movement is the same as in A25. For values bigger than this angle (sharp angles), the unit is completely stopped (speed = 0) and then accelerates towards the new direction (while turning). It seems that this because of this sudden stop (which can happend quite a lot in the battlefield) movement seems jank and slow.

So, in order to balance out the mechanic it seems that the variables at our disposal are:

WalkSpeed

RunSpeed

InstantTurnAngle

Acceleration

And this Speed = 0 when The angle is bigger than InstantTurnAngle.

================================================

I don't have the SVN, so i can't test this out myself, but maybe that speed = 0 thing is the culprit for the current behaviour and setting it to something like Speed = WalkSpeed/2 might be a good compromise. Again can't be sure.

 

[LetswaveaBook]

1 hour ago, Micfild said:

I don't have the SVN, so i can't test this out myself, but maybe that speed = 0 thing is the culprit for the current behaviour and setting it to something like Speed = WalkSpeed/2 might be a good compromise. Again can't be sure.

I once did a Unity tutorial John Lemons haunted Jaunt where you did exactly movement like that. Rather than turning on the spot, units(or characters) would move forward and turn while moving. Depending on the Speed&turn rate, you didn't need much space to turn around. Also considering unit pushing, I assume opposing units wouldn't really block units from turning. I have to admit I never tested it on SVN so I could be easily wrong about that and I am an SVN noob with an outdated PC.

Maybe also allowing units to turn small angles(like 5 degrees) without needing to stop to turn would be nice.

https://learn.unity.com/project/john-lemon-s-haunted-jaunt-3d-beginner

[Micfild]

7 minutes ago, LetswaveaBook said:

Maybe also allowing units to turn small angles(like 5 degrees) without needing to stop to turn would be nice.

They don't stop for such small angles. The current <InstantTurnAngle> for units in general (not ships or siege) seem to be 1,5 (i think radians) so about 86 degrees. Which means that any angle below that they should behave the same as A25 (if i'm getting this right).

[alre]

3 hours ago, Micfild said:

Hi!

I was reading the diff and i have a question and a few observations.

Question:

What is the main goal of introducing acceleration? (is it to bring unit movement closer to reality or to smooth out movement in sharp angles or something else?)

Observations:

They way acceleration seems to work in the code is:

A variable in defined in the templates called <InstantTurnAngle>. For any turn angles smaller than this variable, movement is the same as in A25. For values bigger than this angle (sharp angles), the unit is completely stopped (speed = 0) and then accelerates towards the new direction (while turning). It seems that this because of this sudden stop (which can happend quite a lot in the battlefield) movement seems jank and slow.

So, in order to balance out the mechanic it seems that the variables at our disposal are:

WalkSpeed

RunSpeed

InstantTurnAngle

Acceleration

And this Speed = 0 when The angle is bigger than InstantTurnAngle.

================================================

I don't have the SVN, so i can't test this out myself, but maybe that speed = 0 thing is the culprit for the current behaviour and setting it to something like Speed = WalkSpeed/2 might be a good compromise. Again can't be sure.

 

I tested it, but I didn't look at the code. I was under the impression that smaller turns the unit would also slow down, but less. I must have been confused by the actual turning time, that was already there. Maybe the best result could be achieved if for all turns that are sharper than <InstantTurnAngle> rotation time could be taken down near 0 (seeing a horse spinning around itself is not the best one can ask), but after the turn then acceleration applies?

I wouldn't think acceleration looks quite good on infantry though.

Still, other units like ships and chariots seem like they may very good use of acceleration.

I would love to see this feature used to the best, but yet again I must insist that it doesn't break this hardly achieved balance (I agree with @a 0ad player here). Once again, I think that single/multiplayer is largely a false dichotomy.

[Stan`]

53 minutes ago, alre said:

Once again, I think that single/multiplayer is largely a false dichotomy.

Yes of course it's an over simplification of the problem. When I came around ten years ago I saw something like age of empires with a total war style of fighting, with anim sync arrow volleys etc.

I've never been into the grand strategy thing, nor an overly complex economy. The thing I loved the most about Age of Mythology was the mythology strings, and the scenario editor especially any kind of building that was out of the ordinary.

I'm not interested at all in the competitive part as I am the worst player amongst the team member and it's just hopeless. I just enjoy building stuff and looking at the art. So I am a minority inside another one.

 

[Lion.Kanzen]

3 minutes ago, Stan&#x60; said:

I'm not interested at all in the competitive part as I am the worst player amongst the team member and it's just hopeless. I just enjoy building stuff and looking at the art. So I am a minority inside another one.

most of us artists are immersive.

Alexandermb, I among others, we like more to enjoy what we create.

When we see that absurd attempt at AoE IV trying to be an E-sport we feel relegated.

 

AoM and AoE III at least had the courage to create new gameplay.

 

Total war is very nice, but it lacks the build feel. City building is very nice mixed with battles.

 

If you look at the battle scenarios in AoE IV they are just combat. Little city management, little micro and macro is simple, there are many automaticisms.

It became a copy of Blizzard, which are not bad, just that they are very repetitive and routine.

[Freagarach]

8 hours ago, Lion.Kanzen said:

Total war is very nice, but it lacks the build feel. City building is very nice mixed with battles.

Then there is the ANNO series, which satisfies the itch for city building but has horrible battles. ^^'

I've always longed for a combination of the two (ANNO and Total War).

[wowgetoffyourcellphone]

9 minutes ago, Freagarach said:

Then there is the ANNO series, which satisfies the itch for city building but has horrible battles. ^^'

I've always longed for a combination of the two (ANNO and Total War).

I think an ultimate strategy game would allow you to "city build" your capital or home city, while using basic Total War-style city management for your colonies or conquered cities.

[psypherium]

Extra speed is fine just so long as it doesn't reintroduce the RNG inherent in dancing strats. Less RNG is more fair.

Edited October 27, 2021 by psypherium
semantics

[Freagarach]

6 hours ago, psypherium said:

Less RNG is more fair.

AoE-IV Agreed when adding the cruise missiles.

[real_tabasco_sauce]

@Micfild I agree, maybe increasing the angle at which a25 behavior is preserved is good. I do not have the SVN either, so I haven't actually seen the current movement, but I have a rough idea after seeing everyone's thoughts. I feel like the units shouldn't be stopped entirely, with speed = 0, but rather accelerate from a base speed only a little lower than walk speed. I certainly don't think acceleration should be a huge departure from a25.

speed = walkspeed/2 is a good start, but i think maybe even 2/3 would be enough, and this could perhaps depend on the class of the unit (cav, inf, ranged, etc). It does seem logical that a unit wouldn't start at 0 but rather accelerate into a march after a second or so.  I think cavalry should have to accelerate a bit more than infantry to reach their normal speed.

While unit acceleration is controversial in this forum, i think it is a highly flexible change, and could be tweaked to most people's liking. This could also be another way to balance units into the future.

[Micfild]

3 hours ago, real_tabasco_sauce said:

While unit acceleration is controversial in this forum, i think it is a highly flexible change, and could be tweaked to most people's liking.

I agree. I really like the concept and think it would be a great addition to the game, so i'm glad people are trying to implement it. Of course, as with all new mechanics, it will take a while to tweak the values to get it just right.

4 hours ago, real_tabasco_sauce said:

speed = walkspeed/2 is a good start, but i think maybe even 2/3 would be enough

Yeah, i suggested walkspeed/2 to be a little conservative, but to be quite honest, i have no idea what the ideal value would be. Any suggestion would be just mere guessing.

4 hours ago, real_tabasco_sauce said:

I do not have the SVN either

One intereting thing i found is that we can donwload the individual files from the diff and build a mod with it. Unfortunaly i don't know of any way to donwload the entire file structure other than manualy downloading each individual file (which can be quite a bit of work since this diff has a lot of files in it), but it may prove to be a way we can test and play with it without needing the SVN.

[alre]

1 hour ago, Micfild said:

One intereting thing i found is that we can donwload the individual files from the diff and build a mod with it.

nope. that doesn't work with changes to the engine, like this one. c++ needs to be complied, mods need to be not.

[Micfild]

6 hours ago, alre said:

nope. that doesn't work with changes to the engine, like this one. c++ needs to be complied, mods need to be not.

Makes sense. My bad. I thought that just replacing the .h file would do the trick like javascript.