New forum has arrived so seems the right time to bring up an interesting finding. The TLDR of this is collision attacks are currently used primarily to provide chariots with functionality but they have applications for a lot of monstrous units even when they remain in sustained combat. You can increase combat performance significantly for anti-infantry focused units like Warsphinxes, the Dragon Siblings, Jabberslythe, Bastilodons, Stegadons, etc. without having to increase WS or effecting performance against other unit classes.

I first observed this phenomenon by noting the Rogue Idol has 6 collision attacks, with a refresh of 10 (so 6 collision attacks per second) and that it's performance into infantry was notably better than units with similar anti-infantry units. The biggest difference I noticed was the Rogue Idol regularly hits for it's listed WS value, while these other units tend to deal less than their listed WS per attack.

So I decided to test the impact of adding collision attacks to these units, and while initial damage on the charge was broadly unchanged what I noticed was the overall length of combat for these units decreased substantially. Below is a comparison of a variety of units from live where they either have no collision attacks, or a profile which is unoptimized. I added collision attacks with a max targets and max cooldown value of X, where X is the units WS/100 rounded up (notably this is also the strategy that best optimizes SE chariot performance, though some of those units need WS increases to be viable). No other adjustments were made to the units combat stats. The target was chaos warriors with GWs and the only active micro was the initial attack order.

We can see above the addition of collision attacks, or better optimizing the collision attack profile, provides a significant increase in the rate of damage application. The Warsphinx, Jabberslythe, and War Mammoth end up routing their targets at a much lower total damage level because their damage application increase had a significant impact on leadership. Given the damage dealt is about 70% of the value from live in about half the time in combat we can intuit a roughly 30% increase in damage dealt over time.

For units like Zhao Ming (Miao was not tested formally but the same games are observed in testing), the Stonehorn, the Bastiladon, the Elemental Bear, and Yuan Bo we see a less aggressive decrease in combat time with similar damage numbers being hit when combat ends. Yuan Bo is something of an exception, but that's because he has a some farily significant animation bugs which can impair his performance against small targets, which are detailed in a thread on the old forum. Suffice it to say a few of his animations only apply damage at the very end of his hitbox which causes them to miss targets closer to Yuan. Collision attacks partially address this by improving some of his animiations contact but are not entirely sufficient to fix the issues.

This result is the result of increased consistency in damage application, primarily observed in sustained combat with the initial charges being unchanged. Many of these units on live waste a significant amount of damage based on their splash profiles, resulting in roughly half of their WS being applied. Collision attacks, which appear to be active during at least some of the combat animation in sustained fights, do a much better job of applying damage. On the two units which already have collision attacks on live (Stonehorns and Mammoths) increasing those values to apply damage more efficiently improves the units performance into infantry without impacting their performance against other unit classes. While possible to adjust the splash profile to potentially improve performance collision attacks seem to be an easy tool for complex and long animations, which may be useful for monster design going forward.

For a couple of units here the performance gain is aggressive enough that if implemented it would need to coincide with nerfs. The Jabberslythe in particular I think would benefit from a reduction to MD and a shift to collision attacks, making the unit more offensive but also making the unit more vulnerable to infantry and monster attacks. The Warsphinx I think is reasonable with these gains but additional assessments of performance would be appropriate. The rest feel broadly find considering cost.

Finally this solution does not work for every SE, it appears to be limited to creatures with long animations that send a significant number of infantry into the air. When tested with the GUO no performance gains were observed, where increasing WS by 100-150 showed marked performance gains against infantry while still leaving the WS below Giants, who are far cheaper. For duelists and bipedal units WS increases tend to be more effective buffs than collision attacks, and I think a fair number would benefit from buffs comparable to the Giants +100 WS, with a few like the Bloodthirsters and Shaggoths needing more aggressive gains. Also as an additional note the Shaggoth unit has a bug where it doesn't properly utilized the sync attack system, while Kholek does, which can cause it to overperform against targets with long flashy animations.

Great findings spell.

But i logically fear for balance if we stick to the current system.

I personally believe we have reached the breaking point of

Single collison damage calculation.

Single impact damage system

Single charge system

For all unit types.

And my worry is by improving the collison damage profile we might break the units that do currently have extreme damage out put on a good charge like.rhinox cav will completely over perform.

Though i agree in the absence of any dedicated balance team trying to optimize the collison damage system is the best that can be achieved 

I definitely wouldn't recommend adding collision attacks to units like Rhinox Cavalry, who are generally performing well. The main takeaway is for certain animations on SE's the normal splash profiles just don't work well. Creatures that send infantry flying and fail to get a sufficient number of targets to inflict proper damage seem to do far better with collision attacks, and that's potentially an easier fix than trying to tweak the splash profile to be just right. On the otherside stuff like GUO's work fine, they just need appropriate damage to be threatening.

If CA is going to keep creating SE's with more and more elaborate animations like the Dragon Siblings leaning into collision attacks will produce more consistent and likely simply results when it comes to infantry performance. Obviously it's not enough to fix hitbox bugs entirely like those observed on Yuan Bo (the tail slap is the biggest but the animation where he rapidly moves in an S shape and the dual bite both have issues where they either place the hitbox behind the target or apply an inappropriately small amount of damage, both to SE's and ME units) but using collision attacks does significantly reduce the performance gap seen when compared to non-bugged siblings.

Spellbound1875#4610 wrote:

I definitely wouldn't recommend adding collision attacks to units like Rhinox Cavalry, who are generally performing well. The main takeaway is for certain animations on SE's the normal splash profiles just don't work well. Creatures that send infantry flying and fail to get a sufficient number of targets to inflict proper damage seem to do far better with collision attacks, and that's potentially an easier fix than trying to tweak the splash profile to be just right. On the otherside stuff like GUO's work fine, they just need appropriate damage to be threatening.

If CA is going to keep creating SE's with more and more elaborate animations like the Dragon Siblings leaning into collision attacks will produce more consistent and likely simply results when it comes to infantry performance. Obviously it's not enough to fix hitbox bugs entirely like those observed on Yuan Bo (the tail slap is the biggest but the animation where he rapidly moves in an S shape and the dual bite both have issues where they either place the hitbox behind the target or apply an inappropriately small amount of damage, both to SE's and ME units) but using collision attacks does significantly reduce the performance gap seen when compared to non-bugged siblings.

doesn't rhinox cav have Collison attacks?

dragon sibs dragon forms are useless they are way too clunky to be usefull

i do agree Collison attacks are needed for SEMs what i was trying to say to have two different collision profiles for SE and ME so there wouldn't be cases of issues due to their nature

Does collision attacks interact with max splash targets increasing the number of entities damaged past the base value?

Bastilean#7242 wrote:

Does collision attacks interact with max splash targets increasing the number of entities damaged past the base value? 

Example: Would giving Grimgor collision attacks 7 increase the number of targets damaged by a melee attack animation splash attacks up to base 7 + 7 collision attack targets splashes a bunch of troops with his big swings?

saweendra#3399 wrote:

Spellbound1875#4610 wrote:

I definitely wouldn't recommend adding collision attacks to units like Rhinox Cavalry, who are generally performing well. The main takeaway is for certain animations on SE's the normal splash profiles just don't work well. Creatures that send infantry flying and fail to get a sufficient number of targets to inflict proper damage seem to do far better with collision attacks, and that's potentially an easier fix than trying to tweak the splash profile to be just right. On the otherside stuff like GUO's work fine, they just need appropriate damage to be threatening.

If CA is going to keep creating SE's with more and more elaborate animations like the Dragon Siblings leaning into collision attacks will produce more consistent and likely simply results when it comes to infantry performance. Obviously it's not enough to fix hitbox bugs entirely like those observed on Yuan Bo (the tail slap is the biggest but the animation where he rapidly moves in an S shape and the dual bite both have issues where they either place the hitbox behind the target or apply an inappropriately small amount of damage, both to SE's and ME units) but using collision attacks does significantly reduce the performance gap seen when compared to non-bugged siblings.

doesn't rhinox cav have Collison attacks?

dragon sibs dragon forms are useless they are way too clunky to be usefull

i do agree Collison attacks are needed for SEMs what i was trying to say to have two different collision profiles for SE and ME so there wouldn't be cases of issues due to their nature

To the first, no they do not they apply both collision damage and make attacks the way cavalry do.

I'd agree on the dragon siblings being pretty clunky, hence the suggestion of utilizing collision attacks to improve infantry performance. For SE stuff some under the hood changes are needed.

To the last point collision attacks don't seem to apply to targets larger than infantry, I haven't noticed any differences in performance versus large targets with these changes. It essentially just saves CA time trying to make complicated animations applicable against infantry since collision attacks seem a lot more forgiving when it comes to spreading damage.

Bastilean#7242 wrote:

Bastilean#7242 wrote:

Does collision attacks interact with max splash targets increasing the number of entities damaged past the base value? 

Example: Would giving Grimgor collision attacks 7 increase the number of targets damaged by a melee attack animation splash attacks up to base 7 + 7 collision attack targets splashes a bunch of troops with his big swings?

No collision attacks and splash attacks are mutually exclusive, you use one or the other to distribute damage. In your example Grimgor would only be able to apply his WS to 7 targets at once, though he'd be an exceptionally bad collision attacker because his animations don't create situations where he could distribute damage on contact. Generally if a units attack animation itself is what launches targets (GUO sword slam is a good example) then normal splash attacks are a superior choice. It's when units have big animations that push infantry around without attacking that collision attacks see a significant gain. The Elemental Bear's turning animations are a good example where infantry can be displaced outside of it's splash attack by the animations leading up to the attack. With collision attacks the displaced infantry still receive an attack properly.

The other benefit beyond controlling for animation quality is since collision attacks are applied per second it can boost damage against infantry in particular, which given their massive health pools relative to other unit classes helps avoid instances where damage values are either too low to use against infantry or too high to be safely used against large targets. The Chorf trains are a good example of this in pricinple where their bugged additional attacks produce decent performance against infantry and massively overperform into cavalry. Instead of adding additional attacks optimizing collision attacks would have improved performance against infantry without hurting cavalry.

Spellbound1875#4610 wrote:

saweendra#3399 wrote:

Spellbound1875#4610 wrote:

I definitely wouldn't recommend adding collision attacks to units like Rhinox Cavalry, who are generally performing well. The main takeaway is for certain animations on SE's the normal splash profiles just don't work well. Creatures that send infantry flying and fail to get a sufficient number of targets to inflict proper damage seem to do far better with collision attacks, and that's potentially an easier fix than trying to tweak the splash profile to be just right. On the otherside stuff like GUO's work fine, they just need appropriate damage to be threatening.

If CA is going to keep creating SE's with more and more elaborate animations like the Dragon Siblings leaning into collision attacks will produce more consistent and likely simply results when it comes to infantry performance. Obviously it's not enough to fix hitbox bugs entirely like those observed on Yuan Bo (the tail slap is the biggest but the animation where he rapidly moves in an S shape and the dual bite both have issues where they either place the hitbox behind the target or apply an inappropriately small amount of damage, both to SE's and ME units) but using collision attacks does significantly reduce the performance gap seen when compared to non-bugged siblings.

doesn't rhinox cav have Collison attacks?

dragon sibs dragon forms are useless they are way too clunky to be usefull

i do agree Collison attacks are needed for SEMs what i was trying to say to have two different collision profiles for SE and ME so there wouldn't be cases of issues due to their nature

To the first, no they do not they apply both collision damage and make attacks the way cavalry do.

I'd agree on the dragon siblings being pretty clunky, hence the suggestion of utilizing collision attacks to improve infantry performance. For SE stuff some under the hood changes are needed.

To the last point collision attacks don't seem to apply to targets larger than infantry, I haven't noticed any differences in performance versus large targets with these changes. It essentially just saves CA time trying to make complicated animations applicable against infantry since collision attacks seem a lot more forgiving when it comes to spreading damage.

1. good to know thanks

2. they honestly need new animations

3. Ok i 100% agree

4. do you think it will help game if there was two collision mechanics for SE and ME or a single mechanic will be sufficient for balance  

To point 4 in the abstract sense with sufficient resources it probably would be a benefit, animations that work well against some targets are currently bugged against others. In a context with limited resources I think having a single animation set that works well against everything, or utilizing different systems for different targets is more feasible while producing similar results. I don't know enough about animations to meaningfully determine which better reflects the situation for TWW3.

Thanks for extending the testing on other monsters. The Heirotitan wasn't an option I considered but one where this kind of change is appropriate given how difficult it's been to devise a set of buffs that's both reasonable and effective for it. If you don't mind could you let me know some of the other units you tested? I think compiling a list of units that both need assistance and benefit from collision attacks would be prudent.

In my mod i added collision attacks to the necrosphinx, heirotitan, warsphinx, elemental bear, thorgrim (useful when he does his little charge attack in melee), and tzeentch burning chariots (here its not so impactful in my tests due to its melee animations often make it fly around). i also fixed the collision attacks missing from the khorne hero chariot and doomwheel

Thank you!

It would be good if they add this to the game. 

hellpit is another that gets massive improvements with collision attacks. its most appropriate to apply to any monster that:

1. has big/long/slow attack animations especially lunging/linear ones

2. is a slow expensive lumbering monster with large hitbox

Loupi#8512 wrote:

hellpit is another that gets massive improvements with collision attacks. its most appropriate to apply to any monster that:

1. has big/long/slow attack animations especially lunging/linear ones

2. is a slow expensive lumbering monster with large hitbox

I think hellpit, necro sphinkx , carno ..etc need buffs out side nof this as well , they don't even perform well in to the inteded targets these days.

saweendra#3399 wrote:

Loupi#8512 wrote:

hellpit is another that gets massive improvements with collision attacks. its most appropriate to apply to any monster that:

1. has big/long/slow attack animations especially lunging/linear ones

2. is a slow expensive lumbering monster with large hitbox

I think hellpit, necro sphinkx , carno ..etc need buffs out side nof this as well , they don't even perform well in to the inteded targets these days.

Generally agree with the caveat that considering primary target is important. Duelist stuff like Necrosphinx, carno, etc. I think primarily need more total WS to better threaten monsters and cav.

For a lot of anti-infantry/generalist stuff just having them properly apply their WS to infantry goes a long way towards viability. Some stuff like the Hellpit might need a bit of both given the overall lackluster performance. My take is monsters should be able to use their listed WS value effectively against all targets, with the addition/opitmization of collision attacks being more quality of life change rather than a buff. For a lot of monsters it won't bring them to viability alone, but having the unit card more accurately predict performance into infantry is worthwhile on it's own in my opinion.

I'm hoping to get a complete list of units which benefit from these changes compiled sometime next month, as well as having a clearer idea of the scope of the gains. Ideally that'll help separate which monsters need additional buffs, which are sorted with just collision attacks, and which don't benefit from collision attacks. I'm guessing based on Loupi's additional testing some units like Hydras and the Rotting Leviathan would probably benefit from collision attacks as well but I'd need to test that.

personally i think hellpit should be a generalist with no antilarge, just slightly better stats and HP overall. making it antilarge didnt really make any sense to me. It doesnt reflect the TT, lore or the needs of the roster.  carnosaur needs a bit higher WS for sure as a monster hunting unit