LRMs, Clustering, and Head Hits

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

Jade_Rook

General
Moderator
46 Badges
Feb 23, 2018
2.013
202
  • Magicka 2
  • Magicka
  • Stellaris: Distant Stars
  • BATTLETECH: Season pass
  • Shadowrun: Dragonfall
  • Shadowrun: Hong Kong
  • Surviving Mars: First Colony Edition
  • BATTLETECH: Flashpoint
  • Stellaris: Megacorp
  • Imperator: Rome Deluxe Edition
  • Imperator: Rome
  • Surviving Mars: First Colony Edition
  • Shadowrun Returns
  • Age of Wonders: Planetfall
  • Age of Wonders: Planetfall Deluxe edition
  • Age of Wonders: Planetfall Season pass
  • Age of Wonders: Planetfall Sign Up
  • Stellaris: Lithoids
  • BATTLETECH: Heavy Metal
  • Stellaris: Federations
  • Crusader Kings III
  • BATTLETECH
  • Warlock: Master of the Arcane
  • Mount & Blade: Warband
  • Stellaris
  • Tyranny: Archon Edition
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Stellaris - Path to Destruction bundle
  • BATTLETECH - Backer
  • Surviving Mars
  • Age of Wonders III
  • Age of Wonders
  • BATTLETECH - Initiate of the Order
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Surviving Mars: Digital Deluxe Edition
  • BATTLETECH - Beta Backer
  • Stellaris: Ancient Relics
  • Age of Wonders: Planetfall Premium edition
  • Age of Wonders: Planetfall - Revelations
I originally posted this in the Head Hit: Pilot Injured thread. Reposting by request and with additional clarifications.

TL;DR version

LRMs use a special clustering mechanic that causes missiles to clump together, but also prevents them from hitting the head unless the cluster lands on the head. If the cluster does hit the head, there is a chance that additional missiles will also hit the head, but the odds are very low.

SRMs do not use the clustering mechanics. Instead they are treated like firing a separate weapon for each missile, each with a separate chance to hit the head. That means that an SRM 6 gets 6 chances per volley to hit the head.

Machine guns work the same way as SRMs. Each of their 5 shots is independent and each has a chance to hit the head.

Regardless of how many weapons hit the head in a single mech's activation, it will only cause one pilot injury. Ammo explosions and losing side torsos are also included in this, but falls are resolved separately. As examples, if a mech loses a side torso from an ammo explosion and the pilot is hit by another weapon at the same time, they will only receive a single injury. If that same round of firing was also enough to knock the mech over, either from stability damage or loss of a leg, the pilot will take a second injury from the fall.

Long Version

I ran some LRM clustering tests to check how the clustering mechanic and head hits work.

For LRMs the first missile determines the center of the cluster. If it hits the head, then additional missiles can hit the head, but there is only 1 pilot wound. If the first missile hits anywhere other than the head, there is zero chance for other missiles in that cluster to hit the head. Each LRM rack has about a 1% chance of generating a head hit per volley. I tested this with 1,600 missiles fired across 16 volleys, half with standard hit tables and half with modified hit tables to guarantee head hits.

I have not tested indirect LRM fire to see if it uses a different hit table. I also haven't tested the accuracy of the hit tables to see if the RNG is producing more head hits than expected. There is some evidence that the actual hit rate is closer to 2% rather than the 1.2% predicted by the hit tables. Thank you to @Diedel for additional testing.

Clustering Explanation

All hit locations are determined by the hit locations tables in the CombatGameConstants JSON file. These provide the weight and probabilities of hitting each location.

This is the standard front hit location table. This shows the odds of hitting a location on a mech if you attack it from the front arc.

front hits.PNG

LRMs only use the standard tables to determine the first missile hit location which determines the center of the cluster. The location hit has it's weight multiplied by 8 and the adjacent locations are multiplied by 4, with one exception. The head is not multiplied if the cluster lands on it and is set to 0 if the cluster isn't centered on the head.

This is what the hit location table looks like if the cluster is centered on the head:

head cluster.PNG

Notice that the head hit odds actually went down for subsequent missiles. That is because the head isn't multiplied, but it is still adjacent to the torsos, so those do get their odds multiplied. The odds for the initial hit are still 1 out of 81, but don't expect to be getting multiple LRMs to hit the head from a single launcher.

As further examples here are the Center Torso and Left Leg cluster tables:

center cluster.PNG

leg cluster.PNG

Also of note, called shots only affect where the cluster is centered. They don't modify the cluster hit tables. An LRM 20 with a called shot to the CT still only expects 9 or 10 missiles to hit the CT.

Testing Methodology

I tested clustering and LRM head hits in the single player skirmish against target dummies, Atlai with 1,000 armor in every location but only 2 medium lasers.
I modified an LRM 5 to fire 100 missiles per volley to get clear clusters and set the damage to 1 per missile with no stability damage. I didn't want to knock them over. I was just using armor damage to show hit locations.
I took 4 Kintaros firing only their single LRM 5 each round. I did two rounds with unmodified hit charts and had no head hits.

Results with unmodified hit charts:
unmodded hit charts.PNG


Where each cluster landed should be clear and even over 800 missiles there were no head hits.

After that I modified the forward hit charts to guarantee that I would get some head hits.
hit table.PNG


That gives a 40% chance for any attack to hit the head. I actually had to do 9 shots on this test because the 8th shot got the 4th head head and killed the pilot. With the mech down I couldn't check the armor. This also shows that only 1 pilot injury is generated per volley, regardless of how many missiles hit the head.

Results:
modded hit charts.PNG


As you can see, the volleys that weren't centered on the head had zero head hits, even with the increased head hit chance.

I also did a quick SRM test with the modded hit tables before switching it back. I only fired 2 volleys, but with 44% head hits on the first shot and 41% on the second, it was pretty clear that they are not using the clustering mechanic that LRMs are. SRMs roll location per missile with no modifications for clustering, meaning each missile, not each missile rack, has about a 1% head hit chance.

srm test.PNG


Machine guns work the same. I just did a test with 100 shot, 1 damage machine guns on unmodified hit tables. There was no clear clustering and each of the 3 shots I fired had head hits. Out of 300 total rounds, there were 7 head hits for a hit rate of 2.3%. That is actually higher than I would expect as the rate should be 1.23% (1 out of 81).

MG test.PNG


EDIT: Keeping head hit and pilot injury separate can be tricky.

EDIT2: Adding details on how LRMs function with called shots. TL;DR They still use the clustering mechanics. The called shot should be weighting where the cluster lands, but has no further impact.

Testing Methodology

I used Kintaros shooting at Atlas target dummies in skirmish. The Atlai had no weapons and max armor. The Kintaros were loaded with knockdown SRMs (0 damage, 50 stability damage) and an LRM 100 with no change to damage. I made the absolute minimum changes I need to test this. I would knock down an Atlas with 2 of the Kintaros, record its starting armor values, fire with just 1 LRM 100, recording hit odds and called shot odds as reported by the game, then counted how many hits occurred at each location.

Background Info

Here are the cluster hit tables for the center torso, right side, and right leg. These are just the standard prone hit table, as located in the CombatConstants JSON, multiplied by the cluster hits modifiers (x8 for center of cluster, x4 for adjacent, and x1 for all others, with 0 for head).


Results

Here are the results from 10 shots. I mostly called the CT and 3 of my 4 mechwarriors had Tactics of 6+, so they had the first called shot skill bonus. I did call shots on a couple other locations to get some more info and to avoid coring the Atlai too quickly. Each Atlas carries 320 CT armor + 160 CT structure. After 2 volleys at full damage they were going internal and the third CT cluster finished each one off.

NOTE: Each shot of 100 LRMs was dealing about 200 damage to the CT. Two shots was putting an Atlas at about 100 structure left on the CT. If that seems like a lot, remember that I was firing 100 LRMs in each volley and with full damage each (no cover/brace). The standard complement of an assault LRM boat is 60 LRMs (4x LRM 15 Stalker, 3x LRM 20 Highlander). That is taking the full firepower of 3 assault mechs just to go internal on a max armor Atlas.

There was one instance where I called a shot on the arm and the cluster missed, hitting center torso instead. That Atlas had already taken 2 clusters to the CT and was destroyed. Unfortunately, I did not record all of the damage, but I do know that it would have take at least 60 missiles to destroy the CT and I don't think there were many, if any hits to the arm that I called the shot on. I did get the info for the last mech I destroyed and am only missing the exact number of missiles that hit the CT (but I do have the minimum).


They are all fairly reasonably close to their appropriate cluster hits table. The most interesting result was when I did a called shot on the head. Naturally, the cluster didn't land on the head (it was only a 2% chance) so there were no hits to the head. If called shots from LRMs bypassed the clustering mechanics and used the called shot tables, I would expect the hits to be fairly evenly distributed across the mech.

Instead it landed on the right torso. The hits are very clearly weighted in favor of the right side. They are perhaps a little heavier on the CT than I would expect, but some variance is normal. (heh, normal... unintended statistics pun...)

I agree that the cluster hits tables do resemble the called shot tables. They both weight a specific location more than the others. The use of the clustering tables is probably clearest in looking at the instances when the cluster does not land on the called location. It would also be clearer with Called Shot Mastery as that would give 90% odds for the cluster to hit the CT... but only 55-65% of the missiles would hit the CT.

Conclusion
LRMs use the clustering mechanics, even when fired with a called shot. They deal a lot of damage per ton, but it is unfocused and generally bad at destroying fully armored locations.
 
Last edited:
All done with direct fire. I will probably get around to an indirect fire test sometime this week.

I doubt it is different as the top attack table is, unlike in the beta iirc, never used.

I might as well dump my result in here to have everything together.
My test is a bit different, it is one Locust with 10 weapons (medium laser, LRM20,SRM6, MG) against 4 locust with lots of armor (one costs 680 mil), as the MG does not work at 0 damage for some reason. Hits are always from the front with 95% hit chance. Test is over when all mechs are down, which is 17 head injuries (6+6+6+8 guts). I did grab the info out of the logs, as i am not really interest in injuries, but general head hit chances, so one salvo of 10x20 LRMs can hit more than once, even if it would be only one injury. O have never seen an LRM hit the head more than once, though it seems to be possible, but with only 20 it seems to be pretty rare. The first line is the weapon and the number of head hits overall, the second line is the number of "shots" fired and the percentage of head hits for that, the third line is an approximation of salvos and the percentage for that. I would like to grab more data out of the logs i kept, but the log does not track head injuries right, it list side torso ones also as "head hit".

lrm 19
10513 0,1807286217
525,65 3,6145724341

srm 36
1468 2,4523160763
244,6666666667 14,7138964578

mg 26
970 2,6804123711
194 13,4020618557

ml 19
670 2,8358208955
670 2,8358208955

lrm
CenterTorso 2253
Head Damage 19
LeftArm 1132
LeftLeg 761
LeftTorso 2087
RightArm 1194
RightLeg 889
RightTorso 2178
10513 0,1807286217
525,65 3,6145724341
srm
CenterTorso 274
Head 36
LeftArm 169
LeftLeg 154
LeftTorso 261
RightArm 186
RightLeg 126
RightTorso 262
1468 2,4523160763
244,6666666667 14,7138964578
mg
CenterTorso 176
Head 26
LeftArm 118
LeftLeg 73
LeftTorso 170
RightArm 129
RightLeg 85
RightTorso 193
970 2,6804123711
194 13,4020618557
ml
CenterTorso 146
Head 19
LeftArm 76
LeftLeg 70
LeftTorso 97
RightArm 86
RightLeg 66
RightTorso 110
670 2,8358208955
670 2,8358208955
 
That would explain very well why i got the impression that especially locusts and other enemy mechs with MGs score head hits so often.

If every single projectile has a head hit chance the overall chance to hit the head is much higher the more projectiles are put into the target.
 
Last edited:
That would explain very well why i got the impression that especially locusts and other enemy mechs with MGs score head hits so often.

If every single projectile has a head hit chance the overall chance to hit the head is much higher the more projectiles are put into the target.
In the late game along with my 2 long range snipers with high stability damage I had a pair of dedicated headshot 'Mechs, loaded out with multiple SRM6's and MGs, with 9 tactics for "Called Shot Mastery" to get called shot head hit chances in the 16-18% range for guaranteed head hits every time. Because who's got time to sit around and knock a 'Mech down for up to 4 turns in a row (after accounting for the 2 injuries from side torso destruction) to get all the injuries you need for high-end pilots that may theoretically have up to 6 total hit points?
 
That would explain very well why i got the impression that especially locusts and other enemy mechs with MGs score head hits so often.

If every single projectile has a head hit chance the overall chance to hit the head is much higher the more projectiles are put into the target.

That is flat out wrong. The devs themselves posted that cluster weapons (ie. multiple projectiles) get a single roll for head shot.

Everything else you read by posters here complaining about head shots is confirmation bias.

p.s. Yes. I have finished the campaign. I lost exactly one pilot to his mech getting blown the frak up. I never lost anyone to headshots. I while I managed to kill a fair number of enemies through injury, it was way more often from ammo explosions, ST loss, and multiple knockdowns to rack up those injuries.
 
Everything else you read by posters here complaining about head shots is confirmation bias.
That's why I asked Jade_Rook to make take his research from page 10 of one of those other threads and put it front and center on this post... but facts and statistics still don't seem strong enough to stem the tide of superstition and anecdotes.
 
Last edited:
That is flat out wrong. The devs themselves posted that cluster weapons (ie. multiple projectiles) get a single roll for head shot.

Everything else you read by posters here complaining about head shots is confirmation bias.
.

To quote a great man, "Yeah, 'cept I'm lookin' at some fair compellin' evidence says" otherwise. SRMs and MGs fire multiple projectiles. Neither of them appear to use the clustering rules. Thus, the more SRM or MG projectiles you fire, the better your chances of landing a head shot.
 
To quote a great man, "Yeah, 'cept I'm lookin' at some fair compellin' evidence says" otherwise. SRMs and MGs fire multiple projectiles. Neither of them appear to use the clustering rules. Thus, the more SRM or MG projectiles you fire, the better your chances of landing a head shot.

It's more weapons that's doing it. If you can mount 3xSRM2's, you'll get more headshots than 1xSRM6 even tho they have the same damage.

Same with MG's. In fact, a Commando running lots of support weapons ends up being the king of headshots.
 
It's more weapons that's doing it. If you can mount 3xSRM2's, you'll get more headshots than 1xSRM6 even tho they have the same damage.

Same with MG's. In fact, a Commando running lots of support weapons ends up being the king of headshots.

Again, the evidence suggests that this is not the case. SRMs and MGs roll independently for each hit, not for each weapon system.
 
It's more weapons that's doing it. If you can mount 3xSRM2's, you'll get more headshots than 1xSRM6 even tho they have the same damage.

Same with MG's. In fact, a Commando running lots of support weapons ends up being the king of headshots.
You seem to be trying to have it both ways... all at once you say that head hits are based on number of weapons not number of projectiles (making 3x SRM2’s better than 1x SRM6 in spite of the researched evidence if the OP and with no supporting evidence of your own) and at the same time you imply that more projectiles is what does it (i.e. machine guns being king)
 
That is flat out wrong. The devs themselves posted that cluster weapons (ie. multiple projectiles) get a single roll for head shot.

Everything else you read by posters here complaining about head shots is confirmation bias.

p.s. Yes. I have finished the campaign. I lost exactly one pilot to his mech getting blown the frak up. I never lost anyone to headshots. I while I managed to kill a fair number of enemies through injury, it was way more often from ammo explosions, ST loss, and multiple knockdowns to rack up those injuries.

You are the one who is flat out wrong. LRMs use a clustering table so missiles have a higher chance of hitting around the first impact BUT if the first missile does not hit the head none of the following ones can.

LRMs are currently the only cluster weapons.

SRMs and MGs roll individually.

That has been confirmed by examining the JSONs and by player testing as listed in the OP.
 
You are the one who is flat out wrong. LRMs use a clustering table so missiles have a higher chance of hitting around the first impact BUT if the first missile does not hit the head none of the following ones can.

LRMs are currently the only cluster weapons.

SRMs and MGs roll individually.

That has been confirmed by examining the JSONs and by player testing as listed in the OP.

I'm going by the dev post that said that there is only one chance for cluster weapons, which includes missiles and MG's, to make a headshot.

And as for player testing

*ROFLMAO*

Only one player attempted to get an accurate test, and because he was tweaking json files all over the place, his results might be skewed, and even then, he was pretty in line with expectations.

The very short version: confirmation bias is real.
 
I'm going by the dev post that said that there is only one chance for cluster weapons, which includes missiles and MG's, to make a headshot.

And as for player testing

*ROFLMAO*

Only one player attempted to get an accurate test, and because he was tweaking json files all over the place, his results might be skewed, and even then, he was pretty in line with expectations.

The very short version: confirmation bias is real.

Which dev post was that? Because the one I could find said no such thing.

HBS_Eck said:
Agreed on both points. If you hit the head 3 times, blow off the left torso, and blow out the right torso with an ammo crit all in one volley of fire. That's still only going to be one pilot injury.

You can get two pilot injuries in a single volley if you cause a pilot injury from the weapon fire or melee attack and knock the target over. One for the attack, and one for the falling down.

You replied to Honor_Knight and Eck:
No. You get one per weapon. MG's don't get 5 chances for a headshot. They get one. See the dev post above.


Uh. You can get as many pilot injuries as you can get various injury types. So a head shot, side torso removal, ammo explosion, and knockdown would be 4 injuries and kill most pilots.

As a side note, stick ammo in your feet, not side torsos :D

He said no such thing. You probably made the mistake in your second part of the reply (to Eck) as you didn't realise Eck was talking about it happening in a single attack. Which is correct, remove a torso and cause an ammo explosion with the same attack and its still only 1 pilot injury.

Eck confirmed the latter point in the thread:
So I talked to the combat engineer and we looked at the code.

I was wrong about the LRM clustering. I thought that had been removed, but apparently LRMs still have clustering rolls so the first hit location determines valid hits. If the first hit is not a head hit, then subsequent rolls for that rack can not be a head hit. I'll correct my steam guide to reflect this. Good find all. :)

I was right about the only one MechWarrior injury per volley of fire. So if you have a volley of fire that cooks the ammo, hits the head twice, and blows up two torsos, that's lumped up into one injury. And then if that knocks your mech down, it will inflict a second injury.

- Eck

The discussion was started by Eck's tactics guide in which he wrote:
Missile Weapons - These weapons launch explosive projectiles with propulsion systems. Missile Weapons are a great source of consistent damage (since they roll a separate to hit roll for each missile), the damage is spread out over the target. This can be a good thing if you're "fishing" for a head hit or a chance to damage and Crit Seek an exposed location. Long Range Missiles (LRMs) are guided missiles that can cover long distances and, if you have a spotter, be fired indirectly - arcing over intervening terrain. However, they have a significant minimum range which makes them much less effective at close range. Short Range Missiles (SRMs) inflict greater damage as well as Stability Damage, but have a much shorter range and are incapable of indirect fire.

So here he states that all missiles do not cluster and give greater chances of head hits. His correction is that LRMs still use clustering, he makes no mention of SRMs using clustering so the logical implication is that they do not and thus his statement that missiles are good for generating head hits does apply to SRMs.