Units die too quickly: Melee units have too little armour

[Seleucids]

Spearman and swordsman have 3 hack 3 pierce armour. What is this joke? 

Units die way too quickly in this alpha (and in A26 and in A25). There is no possibility of retreating from a fight. If you retreat, you will loose everything. The amount of damage dealt while turning around and pathfinding outside enemy range is enough to kill the melee unit. Ranged units die almost instantly. The attack damage is disproportionally high compared to the effective health of units, so we see instant vapourisation of entire armies. 

The problem of units dying too quickly is proliferated by forge techs focusing too much on attack damage and not enough shield upgrades. 

 

For comparison, in A23, the melee units had 5 hack 5 pierce armour by default. Fully upgraded units can have up to 8. Meanwhile, the damage techs are not as strong. That allowed reasonable death rates. 

 

How do we fix this? Simple: 

1. Nerf the attack damage upgrades, especially cap the T3 to +1

2. Improve shield upgrades. Increase like +1 +2 +3 each tier. 

3. Increase turn angular frequency. 1000 radians per second is a good value. There is no possibility of dancing given the new pathfinder which stops the unit before it turns, plus the lag which makes dance impossible. A high angular frequency will aid unit pathfinding, prevent units from being stuck and allow retreating. 

4. Default unit armour values:

• Spearman: 6 hack 6 pierce
• Swordsman: 6 hack 5.5 pierce
• Pikeman: 8 hack 10 pierce

[Emacz]

 

11 minutes ago, Seleucids said:

Spearman and swordsman have 3 hack 3 pierce armour. What is this joke? 

Units die way too quickly in this alpha (and in A26 and in A25). There is no possibility of retreating from a fight. If you retreat, you will loose everything. The amount of damage dealt while turning around and pathfinding outside enemy range is enough to kill the melee unit. Ranged units die almost instantly. The attack damage is disproportionally high compared to the effective health of units, so we see instant vapourisation of entire armies. 

The problem of units dying too quickly is proliferated by forge techs focusing too much on attack damage and not enough shield upgrades. 

 

For comparison, in A23, the melee units had 5 hack 5 pierce armour by default. Fully upgraded units can have up to 8. Meanwhile, the damage techs are not as strong. That allowed reasonable death rates. 

 

How do we fix this? Simple: 

1. Nerf the attack damage upgrades, especially cap the T3 to +1

2. Improve shield upgrades. Increase like +1 +2 +3 each tier. 

3. Increase turn angular frequency. 1000 radians per second is a good value. There is no possibility of dancing given the new pathfinder which stops the unit before it turns, plus the lag which makes dance impossible. A high angular frequency will aid unit pathfinding, prevent units from being stuck and allow retreating. 

4. Default unit armour values:

• Spearman: 6 hack 6 pierce
• Swordsman: 6 hack 5.5 pierce
• Pikeman: 8 hack 10 pierce

Or try historical

HP was increased. Melee has alot more resistance than ranged, but = life.  Again people have all these ideas but dont want to test them out

[real_tabasco_sauce]

3 minutes ago, Seleucids said:

There is no possibility of retreating from a fight. If you retreat, you will loose everything. The amount of damage dealt while turning around and pathfinding outside enemy range is enough to kill the melee unit.

Keep in mind https://gitea.wildfiregames.com/0ad/0ad/issues/6946, which means some of your retreating units must stop after you give the order. So try to use just one order for your retreat to avoid this.

6 minutes ago, Seleucids said:

1. Nerf the attack damage upgrades, especially cap the T3 to +1

2. Improve shield upgrades. Increase like +1 +2 +3 each tier. 

I agree with a more cautious version of one of these.

7 minutes ago, Seleucids said:

4. Default unit armour values:

• Spearman: 6 hack 6 pierce
• Swordsman: 6 hack 5.5 pierce
• Pikeman: 8 hack 10 pierce

This would throw us partially back into the meat shield meta.

I'd like to add a new item to your list of solutions:

5. Infantry speed values.

• generally, let melee infantry be faster than their ranged unit counterparts (except pikemen).
• This allows for potentially interesting balance considerations for ranged units

This is all stuff that we can play around with in the com mod, but I'd like to try and tackle the capture situation first.

[Seleucids]

template_unit.xml

The pathfinder has been fixed for you

 

simulation/templates/template_unit.xml	2025-03-11 15:48:17.774030190 +0000
@@ -62,7 +62,7 @@
     <Anchor>pitch</Anchor>
     <Floating>false</Floating>
     <FloatDepth>0.0</FloatDepth>
-    <TurnRate>14</TurnRate>
+    <TurnRate>100</TurnRate>
   </Position>
   <RangeOverlayManager/>
   <RangeOverlayRenderer/>
@@ -126,7 +126,7 @@
     <WalkSpeed>9</WalkSpeed>
     <RunMultiplier>1.67</RunMultiplier>
     <InstantTurnAngle>1.5</InstantTurnAngle>
-    <Acceleration>35</Acceleration>
+    <Acceleration>1000</Acceleration>
     <Weight>10</Weight>
   </UnitMotion>
   <Visibility>

 

Edited 22 hours ago by Seleucids

[Seleucids]

5 minutes ago, real_tabasco_sauce said:

This would throw us partially back into the meat shield meta.

I think there is nothing wrong with meatshield meta. 

But you can try to increase armour of ranged units as well, so that everyone is strong. All in all we just want units to live a bit longer. Right now, if you make 1 mistake, you have lost it for your team. 2v1s are especially problematic, there should be some kind of retreat possible or at least give all players enough time to wait for their ally to arrive. Meatshield or not, please please just no more vapourising units !

 

8 minutes ago, real_tabasco_sauce said:

agree with a more cautious version of one of these.

Suggest values please

 

 

[chrstgtr]

Personally, I think units die too fast and have died too fast since a24, which removed the auto health upgrades from phasing up.

I think the obvious improvement is just an overall unit health buff. It makes battles last longer, which is more satisfying, and makes retreating a bit easier, which I agree is needed  

Also, agree with @real_tabasco_sauce re melee units needing a speed buff. Glad you’ve come around to this idea. This is more about melee/range balance, though. 

[real_tabasco_sauce]

3 minutes ago, Seleucids said:

14 minutes ago, real_tabasco_sauce said:

agree with a more cautious version of one of these.

Suggest values please

I was thinking like 1-2-2 or 2-2-2. I see now that you mean 1-2-3 should be the values (thought you wanted to increase each tech by this much). So, what you suggested might also be fine.

[real_tabasco_sauce]

1 minute ago, chrstgtr said:

I think the obvious improvement is just an overall unit health buff. It makes battles last longer, which is more satisfying, and makes retreating a bit easier, which I agree is needed  

Yeah, I think this is the safest option (least knock-on effects).

[borg-]

I'm honestly fine with it, I don't think units die too quickly now, it seems to be in line with what 0a.d has always been. If units take longer to die, could we have bigger performance issues as well?

Furthermore, lag has been a painful part of 0a.d for years, games already take much longer than they should, these changes would make it even worse.

Edited 22 hours ago by borg-

[Seleucids]

Great, so what health values are we suggesting? 

Let's try:

150 health for melee inf

75 health for ranged inf

Cavalry can stay as they are, for now. 

[Seleucids]

8 minutes ago, borg- said:

If units take longer to die, could we have bigger performance issues as well?

Actually not at all. Longer to die = less death updates -> less need to call entity corpse / mirage replacements which task performance + less need to call death animations (also performance issue). There will be less stuttering as well. 

I think the performance issue you mean is constantly having 200 units on the screen. This will keep the fps low for longer. However, right now the min fps is the same if not even lower when you are at 200 pop, so it doesn't really change much. The quick death approach is just banning units from the map to reduce lag. In any case you can just limit max pop to 175. 

 

To summarise, rate of death has not much impact on performance. 

Edited 21 hours ago by Seleucids

[real_tabasco_sauce]

7 minutes ago, borg- said:

could we have bigger performance issues as well?

Probably less costs actually, since there would be less frequent range queries in typical fights, but my guess is it wouldn't be very noticeable.

[chrstgtr]

8 minutes ago, real_tabasco_sauce said:

Yeah, I think this is the safest option (least knock-on effects).

Agree. All the other stuff impact unit balance instead of actually address the underlying concern that all units die too fast. 

[chrstgtr]

7 minutes ago, Seleucids said:

Great, so what health values are we suggesting? 

Let's try:

150 health for melee inf

75 health for ranged inf

Cavalry can stay as they are, for now. 

This doesn’t make sense. Again, this changes unit balance, which isn’t what’s being discussed.

Just do something like +10% health across all units. Or make it an auto upgrade like it was before with phasing 

[Seleucids]

25 minutes ago, chrstgtr said:

Just do something like +10% health across all units. Or make it an auto upgrade like it was before with phasing 

This also seems fine but 10% wont make much of a difference. 30% is a better starting point. 

The units that die too fast are infantry and especially ranged infantry. Cavalry doesn't have this problem and therefore they don't need health increases. Note that each ranged cav has double the health of a ranged inf with more armour. Each melee cav is  60% stronger than its inf counterpart. OP. 

Auto-upgrade with phases might be fine. 

[Obskiuras]

I suggest a shield wall formation:

The shield wall formation would be diferent from phalanx formation. The shield wall formation would give a bonus against projectile damage, the phalanx formation would give a bonus against melee damage. Many players use and abuse ranged infantery, witch exterminates melee infantery. The shield wall formation will make the melee infantery more viable, in a historical period marked by the use of strong infantery, and will make it easy to invade enemy territory without losing more than half of your army in the process.

 

[Atrik]

Units toughness are at a good spot now (exception for champ cavs). It's better now with the melee re-balance then it was in vanilla a26. Retreating should have a non-null cost, and it's not that costly rn.

[Emacz]

why cant it be balanced where melee/ranged have same health, but a lot different in resistances? so ranged die quicker that way.

[Obskiuras]

4 hours ago, Obskiuras said:

I suggest a shield wall formation:

The shield wall formation would be diferent from phalanx formation. The shield wall formation would give a bonus against projectile damage, the phalanx formation would give a bonus against melee damage. Many players use and abuse ranged infantery, witch exterminates melee infantery. The shield wall formation will make the melee infantery more viable, in a historical period marked by the use of strong infantery, and will make it easy to invade enemy territory without losing more than half of your army in the process.

 

What do you thing @Stan`?

[Stan`]

I think something like that was attempted in milleniumad. Not sure it was finished.

[Classic-Burger]

Ideally, units should be able to hold out for a long time so they can be better controlled. In other words, better handling.

In terms of combat I want slow pacing.

And this applies to all soldiers.

Retreat is important.

Otherwise the game becomes a mess about spamming units. It becomes a cycle of dying and training and very little maneuverability.

Look at melee combat in other strategy games.

[Classic-Burger]

12 hours ago, Obskiuras said:

I suggest a shield wall formation:

The shield wall formation would be diferent from phalanx formation. The shield wall formation would give a bonus against projectile damage, the phalanx formation would give a bonus against melee damage. Many players use and abuse ranged infantery, witch exterminates melee infantery. The shield wall formation will make the melee infantery more viable, in a historical period marked by the use of strong infantery, and will make it easy to invade enemy territory without losing more than half of your army in the process.

 

 The problem is that the formations are useless without the bonuses that have not been implemented.

[Classic-Burger]

12 hours ago, Seleucids said:

This also seems fine but 10% wont make much of a difference. 30% is a better starting point. 

The units that die too fast are infantry and especially ranged infantry. Cavalry doesn't have this problem and therefore they don't need health increases. Note that each ranged cav has double the health of a ranged inf with more armour. Each melee cav is  60% stronger than its inf counterpart. OP. 

Auto-upgrade with phases might be fine. 

More fitness training technologies.