Armor Training - Action Penalty / Maneuver Penalty

[Asrial]

I'm wanting to put this research in a good spot.

So the long standing belief is that to train off the maneuver penalty on armor it's "Skill Bonus = 20 * RT Penalty" and so that's why you end up with "8 ranks for double leather" and "80 ranks for MBP/hauberk" and such.

I feel that this is wrong.

First though, there is an important distinction that has to be made.

At some point "maneuver penalty" became a thing and became separated from "action penalty." Based on my research, a recent comment by GM Contemplar, and GM comments through the years.. I believe that the only thing that exists is action penalty.

Here's the comment by Contemplar on Discord (7/24/19) when I was asking for clarification about whether maneuver penalty and action penalty are two separate things or not:

"As I said before, I did look into the code and the Maneuver and Action penalty appears to be the same thing. Just used different terms, which could be anything. So to appease your concern I would say they are the same until proven otherwise."

Another piece that supports this is that armor over training, with the intention to reduce the action penalty, is based on "Skill Bonus = 20 * RT Penalty - 10" training.

So...

1) Action Penalty and Maneuver Penalty are the same thing.

2) Training for this penalty is "Skill Bonus = 20 * RT Penalty - 10."

Here's my research that shows this...


SMR2 via 309

Unencumbered, 69 armor, hauberk, defensive = [SMR result: 36 (Open d100: 24)] = 12/13

Unencumbered, 70 armor, hauberk, defensive = [SMR result: 31 (Open d100: 21)] = 10/11

Unencumbered, 80 armor, hauberk, defensive = [SMR result: 42 (Open d100: 31)] = 10/11

309 has this thing where the results aren't always the same. Without any changes one shot may be a difference of 10, another shot 11. However, the takeaway is that the "armor penalty" of the target goes down at 70 ranks and remains unchanged at 80 ranks.


Defensive Strength

69 RANKS: 330

70 RANKS: 348

80 RANKS: 348

120 RANKS: 352 (-15 instead of -18 AP; this DS is correctly predicted)

The takeaway here is that the action penalty affects your DS through the evade part of the formula. If you're under trained then the penalty is even worse. Again, just like with 309, there is a change at 70 ranks and no change at 80.

[GSIV Rogue]

I'm wanting to put this research in a good spot.

So the long standing belief is that to train off the maneuver penalty on armor it's "Skill Bonus = 20 * RT Penalty" and so that's why you end up with "8 ranks for double leather" and "80 ranks for MBP/hauberk" and such.

I feel that this is wrong.

First though, there is an important distinction that has to be made.

At some point "maneuver penalty" became a thing and became separated from "action penalty." Based on my research, a recent comment by GM Contemplar, and GM comments through the years.. I believe that the only thing that exists is action penalty.

Here's the comment by Contemplar on Discord (7/24/19) when I was asking for clarification about whether maneuver penalty and action penalty are two separate things or not:

"As I said before, I did look into the code and the Maneuver and Action penalty appears to be the same thing. Just used different terms, which could be anything. So to appease your concern I would say they are the same until proven otherwise."

Another piece that supports this is that armor over training, with the intention to reduce the action penalty, is based on "Skill Bonus = 20 * RT Penalty - 10" training.

So...

1) Action Penalty and Maneuver Penalty are the same thing.

2) Training for this penalty is "Skill Bonus = 20 * RT Penalty - 10."

Here's my research that shows this...


SMR2 via 309

Unencumbered, 69 armor, hauberk, defensive = [SMR result: 36 (Open d100: 24)] = 12/13

Unencumbered, 70 armor, hauberk, defensive = [SMR result: 31 (Open d100: 21)] = 10/11

Unencumbered, 80 armor, hauberk, defensive = [SMR result: 42 (Open d100: 31)] = 10/11

309 has this thing where the results aren't always the same. Without any changes one shot may be a difference of 10, another shot 11. However, the takeaway is that the "armor penalty" of the target goes down at 70 ranks and remains unchanged at 80 ranks.


Defensive Strength

69 RANKS: 330

70 RANKS: 348

80 RANKS: 348

120 RANKS: 352 (-15 instead of -18 AP; this DS is correctly predicted)

The takeaway here is that the action penalty affects your DS through the evade part of the formula. If you're under trained then the penalty is even worse. Again, just like with 309, there is a change at 70 ranks and no change at 80.


Awesome work on this. Did you do any SMR v2 research with the armor overtraining (120 ranks vs. 80 ranks)?

For reference, here are Midgar's SMR v2 armor research results:


The most shocking thing of all, though, was when I put armor on. At 200 armor use and rank 5 armored evasion, I then put on a set of MBP (So -7 AP). My SMR2 defense dropped by.....0. I then tried it without armored evasion (-12 AP), and my SMR2 defense dropped by...0! So then finally, I dropped my armor use skill from 200 to 80 ranks, giving me -20 AP in my MBP, and my SMR2 defense dropped by an incredibly massive.....ZERO!!!!

[gilchristr]

Even if there is no difference between 70 ranks vs 80 for SMRv2, how do you know there no difference for SMR or CML, both of which are still in use (and hidden).

What is with the obsession about trying to save a few TPs, why not just train to 80 ranks. It gives you more redux, plus then you cant get bent over (like GSIVrogue's third wife) on a hidden SMR or CML roll that might be differnet than SMRv2.

The one thing you post may be relevant to is overtraining armor. If overtraining benefits kick in at 120 vs 130 ranks, then I could see stopping at 120 ranks. Can you test overtraining cuttoffs?

[GSIV Rogue]

Even if there is no difference between 70 ranks vs 80 for SMRv2, how do you know there no difference for SMR or CML, both of which are still in use (and hidden).

What is with the obsession about trying to save a few TPs, why not just train to 80 ranks. It gives you more redux, plus then you cant get bent over (like GSIVrogue's third wife) on a hidden SMR or CML roll that might be differnet than SMRv2.

The one thing you post may be relevant to is overtraining armor. If overtraining benefits kick in at 120 vs 130 ranks, then I could see stopping at 120 ranks. Can you test overtraining cuttoffs?

Great points about CML and SMR - but according to this the overtraining benefits for hauberk appear to kick in at 120 ranks as his research shows:

https://gswiki.play.net/Armor_Use

[Asrial]

Awesome work on this. Did you do any SMR v2 research with the armor overtraining (120 ranks vs. 80 ranks)?I glanced briefly through the thread and caught snippets of there potentially being diminishing returns. Over training is definitely not worth it if your goal is SMR2 defense.


Unencumbered, 80 armor, hauberk, defensive = [SMR result: 42 (Open d100: 31)] = 10/11

Unencumbered, 125 armor, hauberk, defensive = [SMR result: 54 (Open d100: 44)] = 10/11

Unencumbered, 135 armor, hauberk, defensive = [SMR result: 29 (Open d100: 19)] = 10/11I think what I noticed with this is that it was easier to get the "10" results instead of "11" results (which is a benefit) at 125 ranks compared to the 70 ranks testing. In all my 309 testing... armor doesn't seem to play a huge part, and if the penalty is only going down by a couple of points with over training then I don't see it really having an important impact.


Unencumbered, 0 armor, no armor, offensive = [SMR result: 143 (Open d100: 94)] = 49/50

Unencumbered, 0 armor, brigandine, offensive = [SMR result: 94 (Open d100: 23)] = 71


Even if there is no difference between 70 ranks vs 80 for SMRv2, how do you know there no difference for SMR or CML, both of which are still in use (and hidden).Yeah, that's really the biggest problem. My biggest defense to that is basically, "prove it." Which no one can. ..but it can't be proven beyond doubt that it's not either. I personally think this is just something that diverged a really long time ago and, since there was no way to prove it, has become set in stone almost.

I have an old post saved where a guy (Divid) did 500 stand tests in hauberk at 70 ranks and 500 at 80 ranks and his data actually had him failing more often in 80 ranks than 70. ..but who knows what a roa'ter maneuver is based on. The easy coding method would be to have a single penalty stat for armor, that's linked to your character, and all maneuvers pull from that stat when they need to do something based on armor.

Regarding over training... that SHOULD be concrete as there's this note from Coase (https://gswiki.play.net/Armor_Use#Overtraining_Benefits)...

"For example, an augmented breastplate requires 90 ranks of Armor Use to train away its RT adder. It has an Armor Group (AG) value of 5 (skin = AG 0, robes = AG 1, soft leather = AG 2, hard leather/scale = AG 3, chain = AG 4, plate = AG 5). Its base action penalty is 25. So, if it is worn by a player with 140 ranks of Armor Use, the action penalty will drop to 21, to 17 at 190 ranks, and to a minimum value of 13 at 240 ranks."

[GSIV Rogue]

I glanced briefly through the thread and caught snippets of there potentially being diminishing returns. Over training is definitely not worth it if your goal is SMR2 defense.


Unencumbered, 80 armor, hauberk, defensive = [SMR result: 42 (Open d100: 31)] = 10/11

Unencumbered, 125 armor, hauberk, defensive = [SMR result: 54 (Open d100: 44)] = 10/11

Unencumbered, 135 armor, hauberk, defensive = [SMR result: 29 (Open d100: 19)] = 10/11I think what I noticed with this is that it was easier to get the "10" results instead of "11" results (which is a benefit) at 125 ranks compared to the 70 ranks testing. In all my 309 testing... armor doesn't seem to play a huge part, and if the penalty is only going down by a couple of points with over training then I don't see it really having an important impact.


Unencumbered, 0 armor, no armor, offensive = [SMR result: 143 (Open d100: 94)] = 49/50

Unencumbered, 0 armor, brigandine, offensive = [SMR result: 94 (Open d100: 23)] = 71

Yeah, that's really the biggest problem. My biggest defense to that is basically, "prove it." Which no one can. ..but it can't be proven beyond doubt that it's not either. I personally think this is just something that diverged a really long time ago and, since there was no way to prove it, has become set in stone almost.

I have an old post saved where a guy (Divid) did 1000 stand tests in hauberk at 70 ranks and 1000 at 80 ranks and his data actually had him failing more often in 80 ranks than 70. ..but who knows what a roa'ter maneuver is based on. The easy coding method would be to have a single penalty stat for armor, that's linked to your character, and all maneuvers pull from that stat when they need to do something based on armor.

Regarding over training... that SHOULD be concrete as there's this note from Coase (https://gswiki.play.net/Armor_Use#Overtraining_Benefits)...

"For example, an augmented breastplate requires 90 ranks of Armor Use to train away its RT adder. It has an Armor Group (AG) value of 5 (skin = AG 0, robes = AG 1, soft leather = AG 2, hard leather/scale = AG 3, chain = AG 4, plate = AG 5). Its base action penalty is 25. So, if it is worn by a player with 140 ranks of Armor Use, the action penalty will drop to 21, to 17 at 190 ranks, and to a minimum value of 13 at 240 ranks."

Thanks for the follow up. Great posts.

[Asrial]

During this I also discovered that full leather has a -1 AP and not the 0 it's been listed at for the past 20 years :D

[gilchristr]

Did you make sure to take off your leg greaves when you observed that -1AP?