New feature: unit acceleration

[alre]

In SVN version of the game, units don't move at full speed right after they are told to, but take a brief acceleration time to get to full speed. If told to turn, they start moving after the turn with a speed that is slower when the turn is sharper.

Effectively, this is the same thing as having slower unit turn rates. I think it looks better though, and I'm not sure about how this may have a better effect on how ranged units correct throwing direction (but it looks promising). What puzzles me, is that A24 turn times were unpopular and have been corrected in A25 for that reason. However, they are coming back now with another name, also I'm not sure, but they feel even slower than in A24 to me. The motive for the change, apparently, is to avoid dancing, but I never saw anyone complaining about dancing lately.

Maybe, now that unit acceleration has been introduced in the engine, turn rate and acceleration stats can be tweaked as to make turns overall not slower than they are now, which is what the community has been asking before.

[Stan`]

There was a big debate and disagreement between @wraitii and @bb_ over this. Which in turn ends up being a Single Player versus Multiplayer debate. Or an A23 vs all (wink Player of 0AD) 

https://code.wildfiregames.com/D3523

https://code.wildfiregames.com/D3179

https://code.wildfiregames.com/D3274

 

So far I see no way to conciliate the two.

 

[real_tabasco_sauce]

@Stan`

I think unit acceleration is fine for multiplayer. It could be increased or decreased for infantry, but I think it would add a lot more risk to playing cavalry. If you foolishly dive into some spearmen, you shouldn't be able to turn on a dime and instantly retreat. In general, I think it would allow for more strategy when it comes to out-maneuvering your opponent.

@alre Maybe if unit acceleration and slow turn speeds combine to be too slow overall, turn speeds should be increased in favor of keeping unit acceleration, especially considering the changes in "proposals for formations"

Right now, i've seen more people complain about heroes and other tank units being put on flee when attacked and baiting enemy melee units to give chase. People get very mad when this happens lol. I agree few players complain about dancing, at least the of the formation kind.

[BreakfastBurrito_007]

@alreI noticed this change as well but I have not tested it out. Thank you for bringing up the topic and starting the conversation.

I had envisioned the change to help differentiate cavalry from infantry, as well as improve balance between the two.

Compare swordsmen and cavalry swordsmen, and we realize that cav swords are just fast, strong, swordsmen. Adding differences in movement behavior is the main way to distinguish cavalry from infantry, and I feel if the unit acceleration values are done right then at the end of the day we would have cavalry that have quite noticeable acceleration while infantry are harder to notice.

I predict the following results if unit acceleration is less noticeable for infantry and quite pronounced for cavalry:

• cavalry rushes in the early game will be harder
• palisades will be more effective against cavalry even if they have the same time to kill when attacked by cavalry (cavalry will have to stop)
• cavalry taking a bad fight will be more punished (I hope this would extend to cavalry champions)
• cavalry micromanagement in battle will be less easy

A key remaining question to do with acceleration values is how to assign them to the different weapon classes, such as spear, pike, archer, sling? or should these values be strictly determined by whether or not the unit is riding a horse?

@alre did you find the acceleration rates too slow when you tried them?

 

[alre]

because of slower turn times, cavalry was harder to use in A24, and noone liked that. Noone would cav rush, noone would use cavalry except for cav archers.

2 hours ago, real_tabasco_sauce said:

especially considering the changes in "proposals for formations"

what's that?

[real_tabasco_sauce]

7 minutes ago, alre said:

what's that?

 

[chrstgtr]

12 hours ago, alre said:

In SVN version of the game, units don't move at full speed right after they are told to, but take a brief acceleration time to get to full speed. If told to turn, they start moving after the turn with a speed that is slower when the turn is sharper.

Effectively, this is the same thing as having slower unit turn rates. I think it looks better though, and I'm not sure about how this may have a better effect on how ranged units correct throwing direction (but it looks promising). What puzzles me, is that A24 turn times were unpopular and have been corrected in A25 for that reason. However, they are coming back now with another name, also I'm not sure, but they feel even slower than in A24 to me. The motive for the change, apparently, is to avoid dancing, but I never saw anyone complaining about dancing lately.

Maybe, now that unit acceleration has been introduced in the engine, turn rate and acceleration stats can be tweaked as to make turns overall not slower than they are now, which is what the community has been asking before.

Seriously, again? Wasn't this the primary thing that made about half the player base quit last time? 

All that aside, doing massive changes like this break the progress made in balancing units

[wowgetoffyourcellphone]

Maybe try it out bruh?

[BreakfastBurrito_007]

1 hour ago, chrstgtr said:

Seriously, again? Wasn't this the primary thing that made about half the player base quit last time? 

All that aside, doing massive changes like this break the progress made in balancing units

I think if more players had seen/tested the a24 turn rates before release people would have realized how detrimental they were. I think acceleration should only be non-negligible for cavalry. For infantry it can be there, but it should not feel different.

I wish I knew how to test the attack-ground code because I am very excited about it and its balancing implications, these are both things I want to try out this weekend. I agree that there has been great progress in balancing units, the main exception being the meatshield superiority that we see in a25 and a few freak units. This is precisely what I think attack-ground can solve.

 

Edited October 26, 2021 by BreakfastBurrito_007

[maroder]

One might also want to read the discussion/ watch the videos on the commit

https://code.wildfiregames.com/rP25953

Edited October 26, 2021 by maroder

[LetswaveaBook]

13 hours ago, real_tabasco_sauce said:

Right now, i've seen more people complain about heroes and other tank units being put on flee when attacked and baiting enemy melee units to give chase.

I want to try this with Elephants.

On a serious note, if the feature makes movement seem more realistic and smooth, it is the way to go. Then after that comes balancing. What I felt in A24 is that:

1: I never made a diff, so nothing really changed. If you want a change, it seems being able to make a diff is very useful.

2: Unit pathing was overhauled, but there was no balance testing after that. So something was implemented and "the balance team"(if such a team existed) wasn't aware of the consequences. Now that skirmishers have good pathing, they are OP and my view is that archers were OP in A24 because other units lacked good pathing.

[alre]

43 minutes ago, maroder said:

One might also want to read the discussion/ watch the videos on the commit

https://code.wildfiregames.com/rP25953

yeah, this patch slows down movement, because introduces acceleration time, without increasing max speed.

on the other hand, unit pushing sped up movement, because it avoided "stop bumps" and made it possible to have units that actually push other units in the back making them run forward, which was previously absolutely impossible. "Balancing team" wasn't really aware of how deep the consequences this would have had.

In general, I wouldn't mind the game to be more slow. I'd like booming to stay this fast more or less, but battles are currently quite a bit too fast in my opinion. This however, is a complicate matter, and acceleration per se is a simpler one: for how it is implemented, it just means that turning is slower, and since A24, we know what that means.

6 hours ago, BreakfastBurrito_007 said:

I think if more players had seen/tested the a24 turn rates before release people would have realized how detrimental they were. I think acceleration should only be non-negligible for cavalry. For infantry it can be there, but it should not feel different.

I wish I knew how to test the attack-ground code because I am very excited about it and its balancing implications, these are both things I want to try out this weekend. I agree that there has been great progress in balancing units, the main exception being the meatshield superiority that we see in a25 and a few freak units. This is precisely what I think attack-ground can solve.

If it was implemented just for cavalry, it would still be problematic. The main complaint about slower turning rates in A24 was precisely about cavalry. To make cav more strategic and realistic, acceleration is not enough.

[Stan`]

Just to mention it, I had a lot of good feedback about turn rates in A24 from casual players and some YouTubers which seem to be totally against what's represented on those forums.

[wowgetoffyourcellphone]

18 minutes ago, alre said:

If it was implemented just for cavalry, it would still be problematic. The main complaint about slower turning rates in A24 was precisely about cavalry. To make cav more strategic and realistic, acceleration is not enough.

When trample is implemented, the acceleration and turn rates will hurt cavalry less, as they will still be doing (trample) damage a they're turning.

[maroder]

I don't mind acceleration in general, but the problem is that the current pathfinding of large units and formations doesn't fit well to it. 

Large units are frequently obstructed by other units and have to stop and turn around and accelerate again and are obstructed again and have to stop... ect.

For me, this makes them more annoying to navigate. This could however be mitigated if unit pushing would also include standing units, so that larger (heavier) units could just push obstructions away.

And it's a similar story for acceleration and formations. If they would have a smoother movement and would shift less, the acceleration wouldn't be an issue for me.

Edited October 26, 2021 by maroder

[alre]

37 minutes ago, Stan` said:

Just to mention it, I had a lot of good feedback about turn rates in A24 from casual players and some YouTubers which seem to be totally against what's represented on those forums.

yet they messed up the game balance and meta. we already decided to revert that change for a reason.

[wowgetoffyourcellphone]

2 minutes ago, alre said:

yet they messed up the game balance and meta. we already decided to revert that change for a reason.

You decided to not bother balancing it.

[BreakfastBurrito_007]

26 minutes ago, wowgetoffyourcellphone said:

When trample is implemented

sheesh, talk about differentiating units!

Is this in the pipeline? how would it work?

[maroder]

3 minutes ago, wowgetoffyourcellphone said:

You decided to not bother balancing it.

Guess the problem is also that any balancing discussions only makes sense if the people actually test the development version. Discussing balance based on the last release is mostly obsolete and /or misleading.

[alre]

well I did test Valihrant turn rates mod at the time, and I liked most of it. I'm not sure what do you mean when you say I didn't bother.

[Stan`]

I think he means that the realistic values could have stayed if the rest was balanced against it.

[maroder]

2 hours ago, Stan` said:

Just to mention it, I had a lot of good feedback about turn rates in A24 from casual players and some YouTubers which seem to be totally against what's represented on those forums

Yes causal players are very much underrepresented here. 

But even if they would voice their opinions more publicly, the question is still if it is possible to find a balance between what causal players want and what competitive players want.

Turn rates/ acceleration is just the tip of the iceberg.

[chrstgtr]

2 hours ago, Stan` said:

Just to mention it, I had a lot of good feedback about turn rates in A24 from casual players and some YouTubers which seem to be totally against what's represented on those forums.

I get that. But people are just positive when an alpha is released. there was also a ton of positive feedback when a25 came out for quicker turn speeds and I haven’t heard anyone complaint about them about the quicker turn rates.  The difference with a24 was there was also a ton of negative feedback when a24 was released. And that positive feedback for a24 slowly disappeared in a few weeks.  That negative feedback during testing resulted in ‘quicker’ turn rates (that some felt was unrealistically fast) and yet those turn rates were still clearly a mistake. 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 

Edited October 26, 2021 by chrstgtr

[Gurken Khan]

26 minutes ago, maroder said:

Yes causal players are very much underrepresented here.

I try to help out in that regard.

I absolutely hated the turn times in a24 and I'm absolutely not looking forward to them potentially re-introduced with acceleration 'feature'.

[a 0ad player]

Feels like quicksand. Would require a lot of balancing (eco, rushing, fighting, ...). Would break cavalry/rushing without rebalancing, which is generally in a very good place right now.