Login

russian armor

Suppression mechanic: How does it REALLY works ?

17 Sep 2019, 22:07 PM
#1
avatar of elchino7
Senior Moderator Badge

Posts: 6313 | Subs: 1

Cruzz
Too much effort. Don't even have an accurate calculation for suppression in general, only guesstimates based on testing.

Mr Smith
I domt know how suppression works precisely to give you the answer.


Cruzz
How Suppression Works:
Penalties for suppression and pin


Penalties for firing on suppressed/pinned units

Suppression value required for suppress: 0.2
Suppression value required for pin: 0.6
Suppression value at which unit recovers from suppress: 0.15
Suppression value at which unit recovers from pin: 0.5
Suppression modifiers against cover: 0.1 green, 0.5 yellow, 1 nothing, 1.5 red.
Suppression recovery rate: 0.008 per second, multiplied by 0.5 in red, 1 in nothing, 2.5 in yellow, 5 in green
Units gain a 50* multiplier to recovery rate after 7 seconds of being out of combat.

Suppression is a squad value. Individual models don't feature into the equation except for the guy the suppressing weapon is firing on, whose cover will be used to determine the penalty applied to both regular suppression and AOE suppression. Suppression is always applied regardless of whether the bullet hits or misses. AOE suppression is only applied on hit.

Now let me explain why this is utterly broken. Because aoe suppression takes its modifier from the thing that its firing on, this means that any squad in front of green cover being shot at will make stuff behind the green cover suppress extremely fast as well through aoe suppression.

Recovery works in combat, though is clearly interrupted by some events like receiving and receiving damage. As such while it's fairly easy to keep units not in cover suppressed just by firing at them occasionally because the natural recovery speed will take quite a while to remove suppression, against units in green cover you will have to be actively doing something to them in order to stop them from recovering.

Incremental accuracy is probably +15% for every 2 or 3 models in the search area in addition to the one being fired upon. It's definitely not per each model. Takes too long to test to give a certain answer.


Fixed a bug where firing at non-infantry targets would suppress nearby infantry. Now suppression platforms will only deal AoE suppression if the primary target is an infantry model.



Q:It's not supposed to be relevant if a shot hits or misses?
A: It matters for AOE suppression which only seems to apply on hit.


Does that mean every shot fired will actually apply suppression to its target?

Yes suppression on the targeted squad will be applied whether you hit or miss.


Incremental accuracy

It certainly doesn't seem to be per entity, because making a massive stacked blob and firing at it with say an M20 (12% incremental on 20% far accuracy) will still have it miss most shots. For the time being I've been assuming that it's being calculated like this as results seemed mostly in line with my very limited testing:

while(targets/2 >= 1){
accuracy = accuracy * incremental;
targets = targets/2;}


17 Sep 2019, 22:12 PM
#3
avatar of elchino7
Senior Moderator Badge

Posts: 6313 | Subs: 1

Recompilation from old post, some values might be outdated though.

That been said, after some recent tests, some things kinda contradict or doesn't correlate from what we know so far.

Basically, i was testing suppression using the following variables.

MG42, range 20, models with invulnerability so they don't die, behind Conscript sandbag (heavy cover).

At that range, vet0, 4/5 model squads (Pio, CE, IS, Volks, Rifles) were not suppressed in the first mag (6 bursts). 6 model squad (Conscript/Guards) and 7 model squad (Cons) were suppressed in 6 and 5 burst respectively.

What i found surprising, later, is that doing the same test with vet 3 vs vet 0, it would throw other results as well (takes longer to suppress).

On top of that, testing 7 squads of conscripts with 1 model, will make suppressing the targeted model takes ages as opposed as previous results (like 15 bursts and only 1 squad been suppressed).
18 Sep 2019, 03:13 AM
#4
avatar of BlueKnight

Posts: 149

I honestly find all the information in this topic very interesting. Thank you for digging the info and sharing it. I always had a feeling that lower RA/model squads like Tommies are more forgiving when getting a burst from MG42 than conscripts, as if they would take longer to suppress but it could be just my feeling and I could be wrong. Your test results are intriguing though.
18 Sep 2019, 13:42 PM
#5
avatar of Stug life

Posts: 3971

wait vickers get no debuff if pinned :romeoHype:
18 Sep 2019, 14:03 PM
#6
avatar of Maret

Posts: 573

It means that bigger squads got more suppression than smaller? They will be faster recover from suppresion than smaller squads? Or recovery applies to entire squad and don't have nothing common with size of squad?
18 Sep 2019, 16:15 PM
#7
avatar of elchino7
Senior Moderator Badge

Posts: 6313 | Subs: 1

wait vickers get no debuff if pinned :romeoHype:


This is from quite some old time, not sure where to check for suppression/pinned effects. The post is prior to UKF release.


jump backJump back to quoted post18 Sep 2019, 14:03 PMMaret
It means that bigger squads got more suppression than smaller? They will be faster recover from suppresion than smaller squads? Or recovery applies to entire squad and don't have nothing common with size of squad?


Bigger model squads receive more suppression. 4/5 been around equal and 6 been noticeable higher.
Recovery is independent of model count. Been on cover or not fighting is important.


This thread is less about been informative and rather see if people are curious enough to talk about it and finally see if we can reach a definite conclusion about how this crucial mechanic works at all.

One of the cores mechanics of the games, and no one knows how it really works at all.
18 Sep 2019, 20:50 PM
#8
avatar of Hannibal
Senior Moderator Badge

Posts: 497

The squad size dependency is really interesting. Have there been any major chances in the suppression mechanic since the release?
This finding suggests that either direct suppression is dealt on a per model basis or that AoE suppression spreads from the targeted model to the surrounding models (also of the same squad), after which the suppression values get added up to the squad suppression value. Since 7 men Conscripts have more models that eat the AoE suppression, suppression is quicker.

One question though on the 1-model squads:
Does it make a difference if the squad got "bolstered" or not (so one model in a 6 men Con squad compared to 1 model left in a 7 men Con squad, similar for Tommies)?
Also does it make a difference if the natural squad size is small or large (for example 1 model of 7 men Con vs 1 model of Major)?
18 Sep 2019, 20:57 PM
#9
avatar of Hannibal
Senior Moderator Badge

Posts: 497

Also the suppression recovery is weird.
Assuming that these values are still the same in the live game or at least similar, recovery in green cover would be

5*0,008/s = 0,04/s

Now suppression of the MG42 is 0,012 (see Vipper's post) per burst I assume. Or per second of burst? Well, I don't know. Still combinded with the reduced RA and suppression of squads in green cover, the MG42 should basically never be able to suppress any squad in green cover.

Now the numbers might be totally outdated. But at least from that we can assume that the suppression model is way, waaaaay more complicated than just shooting at targets and adding a fixed value very time until the suppression bar is full. Or maybe there is an easy solution and I just don't see it.
18 Sep 2019, 21:12 PM
#10
avatar of Sander93

Posts: 1962 | Subs: 3

Now suppression of the MG42 is 0,012 per burst I assume. Or per second of burst? Well, I don't know..


It's 0.012 per bullet. It fires 17.6/16/14.4 bullets per second (rate of fire) depending on the range.

So for example at mid range, it deals (0.012 * 16 =) 0.192 suppression per second. At least that might just be the base value, as other things might further influence it according to some sources.
18 Sep 2019, 21:40 PM
#11
avatar of Hannibal
Senior Moderator Badge

Posts: 497



It's 0.012 per bullet. It fires 17.6/16/14.4 bullets per second (rate of fire) depending on the range.

So for example at mid range, it deals (0.012 * 16 =) 0.192 suppression per second. At least that might just be the base value, as other things might further influence it according to some sources.

Good to know. Can I find this rate of fire stat in the attribute editor?

But still, against a green cover squad this would be 1/10 (see elchino's post), so 0,0192 suppression per second, and only if all bullets hit. Or did I miss another thing here?
18 Sep 2019, 21:42 PM
#12
avatar of murky depths

Posts: 598

Normal suppression applies even if bullets miss, going off what was mentioned before.

And it's .192 not .0192, or put in another way, 4.8x (or 480%) the rate of recovery in green cover.
18 Sep 2019, 22:05 PM
#13
avatar of elchino7
Senior Moderator Badge

Posts: 6313 | Subs: 1

Ok results after a roughly first attempt at modding 101.

Cause i don't know how to copy paste 2 units so they appear at the same time:

-Test subjects were MG34 (tuned up to 2.00 accuracy on all ranges) and MG42 (acc at 0.01)
I purposely chose this 2 because they should have similar suppression values overall.

Same test as i did before:

Test range map, range 20, conscript sandbag, 7 models con vet 0, invulnerability.

MG42: just before i got tired of counting, what used to take 5 burst at vet 0, it took around 30 burst (5 reloads) before the conscripts got suppressed.

MG34: while i didn't test before, it only took 4/5 burst to suppress the same unit. I also tried Obers (4 models) vet 3 and they manage to do so in 5/6 bursts as well. Same with 7 model vet 3 Cons and vet 3 Guards.


On the open, 35ish range, the MG42 took 1 FULL burst to suppress and 4 total to pin.
On the open, same range, the MG34 took 1/2 - 3/4 burst to suppress and only 2 to pin.
18 Sep 2019, 22:25 PM
#14
avatar of Vipper

Posts: 6401 | Subs: 1

...
Now suppression of the MG42 is 0,012 (see Vipper's post) per burst I assume. ...

I can not take credit for the post it Stuglife
19 Sep 2019, 02:39 AM
#15
avatar of Doomlord52

Posts: 556

Guessing the formula used is going to be pretty hard; there's a TON of variables to consider. We might be able to get a good approximation, with enough work, but there will still (probably) be issues. That said, has anyone on the patch team asked if any dev knows (or can find) how it really works? That's probably the easiest (and best) way to go about this.

Here's what I've found, using a similar setup to elchino7.

"TestRangeMap" by Tric / Mr Smith (steam workshop)
Two Con sandbag walls 20 "units" appart
MG42 in Green cover, set to invulnerable
Conscript squad in green cover, set to invulnerable

Basically, the idea is to see what the suppression does at 20 range vs. green cover, without having to deal with damage numbers, squad sizes changing due to losses, etc.



I also tested Vet 3 cons, due to their 0.71 received accuracy (vs 1.09 at vet 0), and found that they were suppressed after 9 bursts, rather than 6. This seems to indicated that suppression is influenced (to some degree) by received accuracy, since the 1.53x improvement to received accuracy followed a 1.5x increase in bursts required.

When I tested this with combat engies, they had similar numbers: 4 models at 9 bursts, 2 models at 10 bursts, 1 model at 10 bursts. Combined with Elchino7's tests, it seems clear that a major influencing value is the number of models in a squad, NOT the percentage of a squad left.

Another factor seems to be the MG's accuracy, as per elchino7's testing. Unfortunately, I can't really test this myself, since no stock suppressing unit has a veterancy bonus ONLY giving accuracy.

Then there's the suppression recovery. I don't think the numbers given are correct, though. The problem with the 0.008 x 5 (green cover mult) = 0.04/second recovery speed is that squads would recover (on average) 0.0634 during the break between bursts, which is higher than the MG42's suppression burst of 0.038976. As Hannibal pointed out, that would mean suppressing a squad in green cover would be impossible (it is possible, though). From what I can tell, it also seems that the recovery is a constant reduction in suppression; so it applies itself every "tick" (who knows how fast that is?).

There's also AoE Suppression, which may or may not apply to single squads as well (only 1 model is targeted, so AoE applies to the others), which may explain why larger squads suppress quicker, but doesn't line up (from what I've seen) with how AoE suppression works with other nearby squads.

Lastly (I hope), there's "incremental accuracy", which no one seems to know anything about (how it works, how fast it increments)...
19 Sep 2019, 05:34 AM
#16
avatar of elchino7
Senior Moderator Badge

Posts: 6313 | Subs: 1

This could be a really "oddball" conclusion but MAYBE this is the reason we had armor instead of received accuracy at the release of the game.

Units were not "harder" to hit but rather deflected some bullets damage. This meant that suppression would apply more evenly across cover and veterancy.
19 Sep 2019, 08:07 AM
#17
avatar of Vipper

Posts: 6401 | Subs: 1

....so it applies itself every "tick" (who knows how fast that is?).
...

According to Cruzz:
"this game generally operates on 8 ticks a second, any amount of time is rounded to the closest 0.125"


Another factor seems to be the MG's accuracy, as per elchino7's testing. Unfortunately, I can't really test this myself, since no stock suppressing unit has a veterancy bonus ONLY giving accuracy.

You can probably use thing that increase suppression like, bulletin and fist strike bonuses from camo.
19 Sep 2019, 08:56 AM
#18
avatar of adamírcz

Posts: 675 | Subs: 1

Nice sumarry

Id like to see several things changed:

The "getting shot at from non suppressing units causes supression to be maintained" mechanic should either be removed or at least heavily nerfed

Green cover should completely deny supression even if there is that one moron ahead/out of it (same mechanic as Brit cover bonus could be used, except instead of increasing ROF it would eliminate recieved supression)


Area of supression buffs across the board to further punish blobs
19 Sep 2019, 16:43 PM
#19
avatar of Sander93

Posts: 1962 | Subs: 3

Good to know. Can I find this rate of fire stat in the attribute editor?

I suppose so, probably somewhere in the HMG weapon files, though I do not know where exactly. You can check the stats at https://coh2db.com/stats/, the site is still up to date for ~95% of units.


But still, against a green cover squad this would be 1/10 (see elchino's post), so 0,0192 suppression per second, and only if all bullets hit. Or did I miss another thing here?

As far as I know, accuracy does not matter for suppression. It just takes rate of fire and suppression per bullet (and according to some people perhaps some other variables).
19 Sep 2019, 17:03 PM
#20
avatar of elchino7
Senior Moderator Badge

Posts: 6313 | Subs: 1

As far as I know, accuracy does not matter for suppression. It just takes rate of fire and suppression per bullet (and according to some people perhaps some other variables).


Accuracy does matter for AoE suppression. AoE only applies on hit. This is why, prior to getting fix, hitting sandbags or vehicles was effective to instantly pin any unit nearby.

Now, after testing 1% vs 100% accuracy, it's perfectly clear that accuracy has effects on standard suppression.

Some time later i'll test if i see any difference in 1 model squads.
1 user is browsing this thread: 1 guest

Ladders Top 10

  • #
    Steam Alias
    W
    L
    %
    Streak
Data provided by Relic Relic Entertainment

Replay highlight

VS
  • Oberkommando West flag N1cko
  • Oberkommando West flag Scotch1
  • Soviets flag Frosty
  • Soviets flag Beatricks Swire
uploaded by Frost

Board Info

309 users are online: 309 guests
90 posts in the last 24h
1196 posts in the last week
5218 posts in the last month
Registered members: 68586
Welcome our newest member, fverniced
Most online: 918 users on 27 Oct 2019, 01:03 AM