[For bigtino] Thread about heavy armor

bigtino said:

So again, if I see a target has heavy armor, I'm going to choose to exchange for damage, and the result is going to be considerably more expected damage than if I had fired at a lightly armored target and chosen not to exchange. The choice isn't tactical mastery, either - it should be, in-character, pretty easy to visually determine whether a target is wearing light or heavy armor, and the choice is obvious after that is determined.

Click to expand...

Let's see, this is going outside of my comfort zone when it comes to mathematics, but I'm pretty sure that the chance to hit exactly one attack out of two equals 100% - (chance to miss both) - (chance to hit both), in which case the following should hold true.

If I have high ground against my target, and I see that they are wearing a kevlar vest, I will shoot at them twice. If I don't have Deadly Strike, my expected total damage for the two attacks is 11,12. If I do have Deadly Strike, the expected total is 14,47.

If I have high ground against my target, and I see that they are wearing a battlesuit, I will shoot twice and exchange two of my attack dice for one damage die. If I don't have Deadly Strike, my expected total damage for the two attacks is 8,35. If I do have Deadly Strike, my expected damage is 11,16, which is actually slightly less than if I had not exchanged and just attacked twice, for a total of 11,21.

So, if I have high ground, I can expect to do 11,12 points of damage against the target in light armor, or 8,35 points of damage against the target in heavy armor. If I have Deadly Strike, the numbers are 14,47 opposed to 11,16 or 11,21. Whether I have Deadly Strike or not, I can expect to do about 3 more damage to the target in light armor, since there is no reason for me to exchange attack for damage against them. It doesn't matter how much more damage without exchanging I do against light armor than I do heavy armor, or how much more I deal to heavy armor than I deal to light armor when I do exchange.

Dalamar said:

Using these numbers, you've posited that in almost all scenarios, attackers will have Deadly Strike. I find that an odd supposition, seeing as there are numerous other exploits a character could have.
Likewise, this presumes that the attackers are skilled (2d6), equal Grade to the defender, and wielding one of the best rifles in the book.

Click to expand...

Well, I was just using your numbers/examples, which you had initially posited showed a clear advantage for heavy armor *shrug*

I did apply the math slightly wrong, in my example, counting the deadly strike damage twice. The expected damage should actually be 8.75 vs. kevlar and 9.75 vs. battle armor. Closer, but still doing more damage to the heavy armor.

Conceptually, heavy armor should offer more protection against an unskilled combatant right? But if you take someone who keeps all the same stats in your example minus deadly strike and attacks twice without modifiers, they expect to deal 7 damage to the person in kevlar compared to 6.78 damage to the person in heavy armor. That's a pretty negligible difference for paying several thousand credits more.

If you make your sample attacker less experienced with rifles, the loss of damage will be applied equally to both defenders. The likelihood to hit will skew further toward hitting the heavily-armored one because of their -4 defense. I haven't mathed it out, but I'd expect that to make the situation worse.

All of this is just to say that heavy armor probably needs more SOAK attached to it so that the defense penalty is negated.

I adjusted my spreadsheet so that I can actually change numbers around. No ability to directly adjust attack and defense dice pools just yet, but I can at least tweak the SOAK provided by armor and the average damage of the attacker's weapon. So here are some variants of the basic calculations, and the base sheet with a couple more situations added (Cover, Aim+Cover, and defender wearing a Forcefield).

As raspberryfh suspected, losing the skill bonus to damage did not help the battlesuit too much, though the numbers are much closer now for the standard attack. Once I get around to updating the two attacks sheet, I'll see what the total is for those.

Upping the battlesuit's SOAK by 2 or lowering the kevlar vest's SOAK by 1 gives clearer results in the battlesuit's favour.

I also just noticed that the battlesuit includes and integrated military scanner. I presume that factors somewhat in its cost, which is actually twice as much as the equivalent OLD armor (platemail, SOAK 8, 1000gc).

I have very much enjoyed this thread and the one about AP ammo. There is just time to squeeze some small changes into v1.2 based on both threads. Thanks so much for the input!

I want to contribute to this conversation (and the one on AP ammo, but that's another thread), and I think I have a few things to point out that haven't yet been mentioned. Full disclosure, I deal with statistics questions similar to this relatively frequently. I'm not a statistician and anyone who is will shudder in horror at my methods, but I ran a simple simulation and found some interesting results I think others may find worth thinking about. Until I get python to export my results in a convenient graphical format, all I can do is try to describe them (anyone know why matplotlib keeps saving blank figures?). When my code works I'll gladly share it for others to critique and modify.

A few people have done some detailed spreadsheets and calculated average damage outputs in various scenarios. I really don't want to rain on anyone's parade and please don't think I'm trying to denigrate or belittle you or your results, but we're doing a statistical analysis here, and the damage roll forms a distribution that has some average value but also a shape. The average doesn't tell the whole story. There is a lot to be learned by looking at the shape the distribution of damage rolls, and one of the principle things is also the range of likely damage values. The average in a given scenario (say you're using Deadly Strike and Aim, but they're in cover) is a value that depends on if they're using kevlar or a battlesuit (I forget the exact results off-hand), but there are a few things to note. I compared all results across sixteen different scenarios (combinations of Aim/High Ground, Deadly Strike, trade for damage, and Cover), and found that statistically we can't actually say much because the range of possible damage results is quite wide. There's often a difference in the average damage, but that difference often doesn't mean a lot. I'm still unsure about a few aspects to the analysis (I need to talk to a few people who know more statistics), but this is what I'm feeling.

A few other thoughts:

First, there is very often a significant chance that the attack deals no damage. This can be either because the attack misses or because the damage gets SOAKed. Which armour type gives the better chance of taking zero damage depends on the situation. This is obvious and has been mentioned. I bring it up for completeness.

Second, for attacks that do deal actual damage, without exception a target wearing kevlar will take more HEALTH damage on average than one in a battlesuit. Again, obvious and nothing new.

The parameter we're interested in is, how often and under what circumstances does wearing a battlesuit mean the wearer takes less damage than wearing kevlar, and should the costs/performances of these armours be modified to improve this?

From a test combat tonight, I think that doubling the SOAK value of heavy armor probably increases its power too much. Increasing by 1.5x may yield better results, which I'll have to experiment on further.

The math here is out of my depth, but in the hope of a simple solution I'll throw in an idea.
What kind of difference would it make if the Defense penalties were halved, like the shield bonuses were in the errata? With a -1 for medium armor and a -2 for heavy would things improve?

dekrass said:

The math here is out of my depth, but in the hope of a simple solution I'll throw in an idea.
What kind of difference would it make if the Defense penalties were halved, like the shield bonuses were in the errata? With a -1 for medium armor and a -2 for heavy would things improve?

Click to expand...

I think that particular thing is a perfect opportunity to bing back the armor skilll reducing defence penalties. The defence penalty is reduced by -1 for each die in the armor skill (that means you don't really need a light armor skill, but I think that works thematically -- wearing a synthetic weave shouldn't require skill; wearing powered combat armor should).

I finally got my code to do things correctly. I assumed attacks from someone with a 5d6 attack pool using an assault rifle (2d6+2 damage). The target has 14 range defense and either kevlar or a battlesuit. I simulated 10.000 attacks in every combination of Aim, Cover, Deadly Strike, and trade for damage. If anyone wonders about other base values, it's really easy to change.

I think the easiest way to present this is via damage quantiles, that is, how likely are you to do at least _ amount of damage? Over all the scenarios I simulated, I found that you're more likely to take at least 1 damage in a battlesuit than in kevlar, but you're more likely to take at least 5 damage in kevlar than in a battlesuit.

If you're more interested in the individual results from the sixteen different situations rather than the overall results, I offer those as well. The two blank spots are where the attacker trades for damage against a target in cover, which obviously is guaranteed to fail.



For those unfamiliar with how to read these types of graphs, the horizontal axis represents a minimum damage amount, and the vertical axis is how often an attack will do at least that amount of damage.

I'd be curious to see what the numbers look like with a base defense of 18, since a defense of 14 is pretty easy to hit with 5 attack dice. We've already established that if you're very likely to be hit, heavy armor should SOAK more damage. The real question is where the tradeoff lies for soaking damage versus completely avoiding damage.

That said, it's interesting to me that when you combine all those different circumstances, heavy armor is more likely to take a small amount of damage 60% of the time. For the remaining 40% of all attacks, it looks like the difference in damage is 1-2 points (less than the actual soak difference of the armors).