LRMs, Clustering, and Head Hits

[Jade_Rook]

SRMs and Machine Guns roll location per projectile and each weapon gets multiple rolls. 3 SRM 2s should be no different than 1 SRM 6 in terms of head shots. If anyone has evidence to show that this is not the case, please post it. I will gladly update my information if it is shown as incorrect.

I created this post to share my findings from verifiable research into how the game mechanics work. I welcome others to share their own testing that we can all better understand how the game works.

EDIT: I don't have time to test everything myself, but I can give advice and offer suggestions.

If the hypothesis is that multiple small SRM racks will cause more head hits than a single large rack, this can be tested. I would set it up much like my initial tests. Mod the SRM 2s to fire 30 missiles per volley (probably want to reduce damage to 1 and remove instability). Also create an SRM 6 that fires 90 missiles per volley. Create a mech that mounts 3 SRM 2s and 1 SRM 6. Fire just the SRM 2s or just the SRM 6 at a target dummy and record what parts get hit.

I predict that the results from firing the 3 SRM 2s will be statistically identical to firing 1 SRM 6. If the head hit chances aren't modified, the test should show a rate of about 1.23 missiles per 100 SRMs fired, regardless of number of SRM racks.

 

[xaoshaen]

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.

There's confirmation bias and then there's blind faith. Faith not simply in what someone told you once, but in your interpretation of what they actually said. When your reaction to objective data is "ROFLMAO", then you've ceased participating in rational discourse.

Cluster weapons do have only one chance to hit the head. SRMs and MGs are not cluster weapons. Their behavior in game is not consistent with such a claim. The evidence is not only in this very thread, but it's available to you with a minimum of testing inside the game.

 

[Max_Killjoy]

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.

Spoiler: Front Hits Table

View attachment 363762

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:

Spoiler: Cluster Centered on Head

View attachment 363769

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:

Spoiler: Cluster Centered on Center Torso

View attachment 363773

Spoiler: Cluster Centered on Left Leg

View attachment 363774

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.

Just to confirm -- LRMs use "clustering" for hit locations regardless of whether it's a normal shot, a Precision Shot, or a called shot on a downed mech, correct?

 

[HonorKnight]

Just to confirm -- LRMs use "clustering" for hit locations regardless of whether it's a normal shot, a Precision Shot, or a called shot on a downed mech, correct?

Yes, the called shot part only affects the initial missile (or the first one that hits), then the remaining missiles in the salvo ignore the called shot and use the clustered table. So at high tactics when called shots are really powerful, LRMs are still not particularly good in called shots.

 

[HonorKnight]

Cluster weapons do have only one chance to hit the head. SRMs and MGs are not cluster weapons. Their behavior in game is not consistent with such a claim. The evidence is not only in this very thread, but it's available to you with a minimum of testing inside the game.

This is wrong by the way and completely disproven by the OP itself. See this chart from the OP showing multiple head hits from a single (artificially large) volley:

 

[Jade_Rook]

Just to confirm -- LRMs use "clustering" for hit locations regardless of whether it's a normal shot, a Precision Shot, or a called shot on a downed mech, correct?

This is correct. As HonorKnight says, called shots do bias where the cluster will land, but nothing further.

Cluster weapons do have only one chance to hit the head. SRMs and MGs are not cluster weapons. Their behavior in game is not consistent with such a claim. The evidence is not only in this very thread, but it's available to you with a minimum of testing inside the game.

This is wrong by the way and completely disproven by the OP itself. See this chart from the OP showing multiple head hits from a single (artificially large) volley:

Cluster weapons both do and do not have a single chance to hit the head. If the cluster is centered on the head, there can be multiple head hits (but they will only cause a single pilot injury). If the cluster is not centered on the head, there will be no head hits. Each launcher only gets a single chance to land a cluster on the head (so they have one chance), but they can cause multiple hits if the cluster does land just right (multiple missiles from a single volley can hit the head). It is a bit of an odd distinction and I can see how it can be easy to jumble up the wording a bit.

Each LRM rack gets 1 chance on normal hit tables (possibly modified by called shots) to determine where the cluster will land.

If the cluster hits the Head -> additional missile can hit the head, one pilot injury is inflicted
If the cluster hits somewhere else -> No missiles will hit the head
After the center of the cluster is determined, all additional missiles will group around center with modified hit tables, ignoring called shot effects

 

[Drasius]

This is wrong by the way and completely disproven by the OP itself. See this chart from the OP showing multiple head hits from a single (artificially large) volley:

Can confirm that LRM's can have multiple head hits - I was doing some testing for another thread debunking a claim that the AI wasn't using the same to-hit tables as the player. I set the to-hit values wildly biased to legs in 1 game and just as wildly towards the head in another. Sure enough, in the game where headshots were prevalent (like, >90% of shots would hit the head) due to the altered to-hit table, a single LRM20 salvo headcapped a ... I want to say Blackjack? but it was a while ago.

 

[Ganggreenkhan]

I ain't one of them there math professor's but if I'm reading OP right what he is saying is that all us salty old TT players who have been screaming that head hits happen way to often in this game aren't crazy.

 

[Drasius]

I ain't one of them there math professor's but if I'm reading OP right what he is saying is that all us salty old TT players who have been screaming that head hits happen way to often in this game aren't crazy.

Head hits are set at roughly 1/2 the rate of TT (a bit over 1% in game compared to ~2.7% in TT), but the lack of to-hit malus' and the fact that the base accuracy here is so high that there's a lot more shots landing. As a "salty old TT player", run a quick number crunch on Indirect LRM fire at long range from a 4/5 pilot who walked against a target that moved 5 hexes using TT rules and then compare that to what you see in game.

A mission of 8 OpFor vs your 4 superstars here takes ... ~30 minutes? A bit less if you're playing well and don't have to maneuver a bunch. How many turns would you get through in half an hour of an 8 v 4 in TT? Not bloody many I'd wager.

Not only that, but how often did/do you play campaigns where pilot injuries mattered instead of just pick up games? If you're not used to tracking pilot injuries from game to game in TT, you're not going to care anywhere near as much about any head hits that don't kill the pilot since it won't have any effect on your next mission.

End result is that you take less head hits overall, but more head hits in the same time compared to a game of TT and the ones you do take in game have consequences. So you're only half crazy.

 

[Fauxknight]

If you consider that most player mechs in this game have max or at least higher than normal armor combined with damage reduction from cover or braced then we can extrapolate that they can take more overall hits than in TT. More hits means more chances to hit the head, so even with the reduced head hit chances from TT we still see a lot of head hits.

I constantly have my scout pilots in the medbay because of a stray headshot. My other mechs take plenty of headshots as well, but the cockpit mods easily soak those up. Strangely I've only once lost a pilot to a head hit, I've had stray PPCs and the like breach a mechs head armor on occasion, but I've been able to keep them safe every time except the one...which I think was a melee attack.

 

[Rusticles]

In the Campaign mode; I have lost Dekker three times to headshots, Behemoth two times to headshots, Glitch one time to head shots, and several (4) other "nobody" pilots to headshots from the OPFOR. Considering I am only running 4 Mechwarriors to the OPFORs 8-12 (at most) on a given mission; I have (literally) hit the heads of OPFOR Mechs (causing deaths) less than half as many times. I have actually had 2 OPFOR Mechwarrior kills via melee attacks.

I have played the Campaign three full times and halfway through #4 as of last night. With the ratio of Enemy OPFOR Units being roughly 2.5:1 (average) against me... Headshots are not out of character at current levels. I can concur via my own data from using different weapons loadouts that LRMs do get a single chance per launcher to hit the head. My testbed Mech was a Stalker with 8 LRM 10 Launchers and another test with a Stalker and loadout with 4 LRM 20 launchers (virtually no armor).

I have not done any testing with SRMs or MGs to date. Probably will not due tot he fact that I can attest to the validity and repeatability of the experiment results from the OP. I run a "vanilla" game with zero mods and got virtually identical results to what has been posted here. My test hypothesis was to see if LRMs were or were not clustered.

Nuff said...

 

[Dermenore]

I originally posted this in the Head Hit: Pilot Injured thread. Reposting by request and with additional clarifications.

[...]

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.

[...]

I did some test and they go again this assumption. See here:

https://forum.paradoxplaza.com/foru...h-weapon-balance.1097172/page-3#post-24354739

 

[Max_Killjoy]

I did some test and they go again this assumption. See here:

https://forum.paradoxplaza.com/foru...h-weapon-balance.1097172/page-3#post-24354739

My understanding is that Jade Rook's "modelling" and numbers are pulled directly from the game's configuration files.

 

[Dermenore]

He said he tested for the classic shoot with a target dummy. He didn't say he did the same with called shoot. He didn't say too he looked into game files to read the rules games.

 

[Jade_Rook]

I could have sworn I ran a test with called shots.

I will retest tonight.

EDIT: Ran the test on my lunch break. Called shots definitely still use the clustering mechanics and definitely aren't gaining the benefits of a called shot on a per missile basis. Out of time right now, will post full stats tonight.

 

[AlyssaFaden]

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

Incorrect.

Don't get me wrong, I sit on the fence on this conversation, BUT I acknowledge something is going on.

Mission start, no shots fired, tank with direct fire weapon. One shot - head hit.
Restarted mission
Mission start, no shots fired, SAME tank, SAME mech ... direct fire weapon: head hit.
Restarted mission
Again, same tank, first shot, at the same mech, ALSO a head hit.

That's not confirmation bias, that's " [Mod edit: Bad language acronym] is going on here?"

 

[HonorKnight]

That's the very definition of anecdotal evidence, @AlyssaFaden. That exact combination of events (three back-to-back headshot rolls) is a 1 in 500,000 chance, true. But 1 in a million chance things happen all the time. With 20,000 players playing at once, a specific 1-in-500,000 event becomes a mere 1-in-25 chance of happening to someone on their next round. Multiply that by the number of rounds played in a day and the number of different types of anecdotes that all are equally alarming, and you end up with tons of anecdotal evidence. Anecdotes make extremely rare things seem common, it's like the old idiom "nobody reports on the plane that doesn't crash", nobody says boo about all the times they didn't get hit in the head, so the vocal few who shout about being repeatedly hit in the head sound like they represent a typical battle, but they simply don't. Controlled large-scale tests clearly demonstrate that head hit chances are in the 1-2% range, not the huge number they are portrayed as by some.

 

[Pherdnut]

Incorrect.

Don't get me wrong, I sit on the fence on this conversation, BUT I acknowledge something is going on.

Mission start, no shots fired, tank with direct fire weapon. One shot - head hit.
Restarted mission
Mission start, no shots fired, SAME tank, SAME mech ... direct fire weapon: head hit.
Restarted mission
Again, same tank, first shot, at the same mech, ALSO a head hit.

That's not confirmation bias, that's " [Mod edit: Bad language acronym] is going on here?"

Wait, so you restart the mission and a tank rolls in and shoots you in the head? Or is this a save point where you never do anything other than repeat the same action with a mech, and then that same guy gets hit in the head by the same tank? That could just be RNG not getting reseeded. You can't save without it being your turn. Make a different move. Pretty sure the game reseeds when you reload outside of missions but I've observed exact same moves leading to exact same results when reloading to a point before taking the first shot at a mech I wanted to capture. Just rotating the camera might undo the seed from your save game.

 

[AlyssaFaden]

Wait, so you restart the mission and a tank rolls in and shoots you in the head? Or is this a save point where you never do anything other than repeat the same action with a mech, and then that same guy gets hit in the head by the same tank? That could just be RNG not getting reseeded. You can't save without it being your turn. Make a different move. Pretty sure the game reseeds when you reload outside of missions but I've observed exact same moves leading to exact same results when reloading to a point before taking the first shot at a mech I wanted to capture. Just rotating the camera might undo the seed from your save game.

Reloaded to an earlier save at the beginning of the mission, thus the loadout I am facing is exactly the same. I didn't repeat the same actions however; tried the mission in different ways, different approaches, different angles. Same thing happened each time: when the tank appeared it would shoot, always at the same mech, always getting a head hit. I should be transparent: on the third instance I *HID* the original target out of harms way, and the tank appeared, shot a new mech, and hit the head. So his first move every time was a head shot. On the third time I was "whatever" and just carried on playing.

It also struck me as some kind of "stuck seed", but I tend to be in the camp of "the RNG is not as random as some people think"

 

[SpaceViking]

Depending on what (pseudo-)RNG they use the seed very well could be "stuck" and always producing a cycle of the same (or similar) numbers. There also could be a flaw in how they handle RNGs.

When I worked on Jagged Alliance 2 1.13 people complained that things that should happen rarely were happening far too often. I investigated and found that their function that checked if actions worked was flawed so that everything happened 1% too often. Basically they checked for an X% chance of something happened by comparing ((pseudo-)random number from 0 to 99) <= X. So for example if X was 1 (i.e., 1%) it would succeed if the (pseudo-)random number was 0 or 1 (which is 2%). Oops! This was actually in the publisher's original release of Jagged Alliance and that code had been used in other of their games as well.