AC softcap and low level hardcap discussion

[DeathsSilkyMist]

Originally Posted by Goregasmic [You must be logged in to view images. Log in or Register.]

I'm still a bit confused he couldn't see why people thought there was a hardcap. Further down in the 2014 thread:

Yeah I am confused about that too, but there are multiple possibilities:

1. He didn't want people to know, so he denied it. The devs don't give away all their secrets.

2. Perhaps it was added later. The 2014 post was Velious beta I believe, so the 289 worn ac clamp might not have been active or added at the time.

3. The way Haynar was testing AC might have involved guide tools, cheats, a separate program that runs the code outside the client, etc. Sometimes when testing with those methods, you get incorrect results because the code being run is different from what the players are using. So he may have thought it was correct.

[Jimjam]

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

Yeah I am confused about that too, but there are multiple possibilities:

1. He didn't want people to know, so he denied it. The devs don't give away all their secrets.

2. Perhaps it was added later. The 2014 post was Velious beta I believe, so the 289 worn ac clamp might not have been active or added at the time.

3. The way Haynar was testing AC might have involved guide tools, cheats, a separate program that runs the code outside the client, etc. Sometimes when testing with those methods, you get incorrect results because the code being run is different from what the players are using. So he may have thought it was correct.

If AC was hardcapped until very end velious (and only soft capped after this for melee) it could be that a) the classes we're testing on (hybrids and priests) are not considered melee for these purposes or b) there is a era based timelock which Haynar was testing outside of, but hasn't been properly activated on r/b/g so we're still getting pre-softcap era results?

[DeathsSilkyMist]

Originally Posted by Jimjam [You must be logged in to view images. Log in or Register.]

If AC was hardcapped until very end velious (and only soft capped after this for melee) it could be that a) the classes we're testing on (hybrids and priests) are not considered melee for these purposes or b) there is a era based timelock which Haynar was testing outside of, but hasn't been properly activated on r/b/g so we're still getting pre-softcap era results?

That is a good point about timelocked content. It sounds like the 289 worn AC clamp was removed sometime in Velious, but I am not 100% sure. It is possible there is a bug where the timelocked content that enables the 289 worn AC clamp is still active.

Testing this with a shaman shouldn't matter. The post says there is a 385 worn AC clamp for the four cloth casters. All other classes use the 289 worn AC clamp.

[Goregasmic]

4- He's incredibly pedant about the use of softcap/hardcap terms. He's technically right but from a practical standpoint it is basically a hardcap for most players, especially in this 2hander era.

[TytosOfEight]

The relationship between AC and minimum hits is not immediately clear. While minimum hits generally increase with AC, there is noticeable fluctuation between AC 217 and 411. But this variability may stem from limited data points rather than a lack of underlying effects (i.e., a lack of statistical power).

In contrast, the trend in damage reduction is more straightforward. Eyeballing the data suggests no meaningful difference between AC 217 and 386. However, there is possibly a significant difference between AC 200 and 217, and again between AC 386 and 411 (when a shield is used).

[Goregasmic]

Originally Posted by TytosOfEight [You must be logged in to view images. Log in or Register.]

The relationship between AC and minimum hits is not immediately clear. While minimum hits generally increase with AC, there is noticeable fluctuation between AC 217 and 411. But this variability may stem from limited data points rather than a lack of underlying effects (i.e., a lack of statistical power).

In contrast, the trend in damage reduction is more straightforward. Eyeballing the data suggests no meaningful difference between AC 217 and 386. However, there is possibly a significant difference between AC 200 and 217, and again between AC 386 and 411 (when a shield is used).

What was tested in the ranger thread is you have D20 distribution, D1 being min hits and D2 max hits. The more AC you have the more Max hits get sent into the Min hit bucket, while D2 to D19 is mostly flat around 4-5%. When Max hits are flattened to around 4-5% too, that is when you reached the "soft cap" and you won't go any lower. So far DSM found it seems the only way to go below the 4-5% floor is adding a shield and you can potentially bring that floor to near zero depending on RNG and the mob you're facing.

[Snaggles]

Thanks for that DSM! Appreciate you taking the time and it’s something to mull over.

[CrazyPro]

shield AC wasn't even a thing until luclin, why hasn't this been fixed yet?

[bcbrown]

Uploading some spreadsheets for three datasets: DSM level 5 vs level 5, my toons vs Shiel, DSM 60 shaman vs level 50. I find this is a more convenient format for looking at the data. Not a lot of conclusions yet from me. Min/max ratio seems noiser than damage/hit. Shield ac definitely has an effect.

[You must be logged in to view images. Log in or Register.]
[You must be logged in to view images. Log in or Register.]
[You must be logged in to view images. Log in or Register.]

I think there's some more nuance in my results against Shiel than we've noticed so far. Looking at average damage per hit I think there's some hints of differences by class. Gonna do some more parses. Running one right now at 198 worn ac plus 61 spell ac for 259 total, empty shield slot. Gonna run another with exact same gear but moving lodi shield from back slot to shield slot. My hypothesis is that "shield ac" exists and is slot dependent, so should be less damage/hit in shield slot than back slot.

After that gonna run a series at like 160, 180, 200 with and without shield ac. Then gonna repeat with cleric and ranger (eventually) to look for any hint of difference between classes. Most of the parses from the last thread were just a couple hundred, too noisy to do much than eyeball the graph and see the same pattern.

[DeathsSilkyMist]

After doing some longer tests, I think I have some good news to report:

=========================================
Shaman 60 207 AC vs level 50 mob 800 hits
=========================================

DV, Count
32, 184
37, 21
43, 34
49, 35
54, 30
60, 33
66, 26
71, 23
77, 36
83, 30
88, 34
94, 39
100, 28
105, 27
111, 30
117, 26
122, 29
128, 34
134, 34
140, 67

Total Damage = 62544

=========================================
Shaman 60 217 AC vs level 50 mob 800 hits
=========================================

DV, Count
32, 188
37, 24
43, 32
49, 30
54, 24
60, 37
66, 29
71, 27
77, 28
83, 36
88, 37
94, 30
100, 23
105, 34
111, 41
117, 33
122, 27
128, 32
134, 28
140, 60

Total Damage = 62141

=========================================
Shaman 60 227 AC vs level 50 mob 800 hits
=========================================

DV, Count
32, 202
37, 27
43, 29
49, 33
54, 30
60, 20
66, 34
71, 26
77, 37
83, 25
88, 32
94, 31
100, 25
105, 31
111, 41
117, 31
122, 27
128, 31
134, 33
140, 55

Total Damage = 61208

=========================================
Shaman 60 300 AC vs level 50 mob 800 hits
=========================================

DV, Count
32, 209
37, 32
43, 33
49, 28
54, 33
60, 33
66, 37
71, 35
77, 27
83, 31
88, 33
94, 28
100, 33
105, 24
111, 37
117, 23
122, 21
128, 32
134, 23
140, 48

Total Damage = 58618

=========================================
Shaman 60 386 AC vs level 50 mob 800 hits
=========================================

DV, Count
32, 221
37, 29
43, 21
49, 30
54, 33
60, 34
66, 46
71, 28
77, 19
83, 27
88, 32
94, 31
100, 27
105, 23
111, 38
117, 36
122, 38
128, 26
134, 36
140, 25

Total Damage = 58631

On longer parses, we start to see a more consistent drop in damage as we increase AC. We also see an increase in Minimum hits, and a decrease in maximum hits.

It looks like the softcap is in effect well past 217 worn AC. The issue is simply that the softcap returns are quite small, which is expected for Shamans and Rangers.

My previous data sets with 400 hits per set were noisy enough to mask the subtle decrease in damage. I am thinking that is what happened in the Ranger thread as well. I don't think they were generally parsing around 1000 hits per AC value.

It is easy to see why people accused Haynar of having a hardcap on AC at high levels. He said he was running 3 hour parses to test his changes, and most people aren't gathering that much data per AC value.

This does show that AC has large diminishing returns past the softcap for at least some classes, which is why people don't feel like it does much. In the data above, going from 207 worn AC to 386 worn AC reduces damage by about 6.5%. It may be a bit higher if I ran even longer parses, as we see the 386 worn AC parse was basically identical to the 300 worn AC parse damage-wise. Considering that the 386 worn AC parse had like half of the max hits and 12 more min hits, that may just be unlucky RNG.

So far the data supports what Haynar's post said about hardcaps and softcaps.

Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]

Uploading some spreadsheets for three datasets: DSM level 5 vs level 5, my toons vs Shiel, DSM 60 shaman vs level 50. I find this is a more convenient format for looking at the data. Not a lot of conclusions yet from me. Min/max ratio seems noiser than damage/hit. Shield ac definitely has an effect.

Thanks for the spreadsheets! Just be aware that some items like orbs do not count as shields, so be sure to test with an actual shield.