Testing the system of balance

[Phoenix_Dragon]

Nymdok suggested posting up an example of a combat that I later examined using his balancing system, and got vastly different results than had actually happened in-play. My eyeballing of it would have judged it as a challenging, but winnable fight. The math appears to say that it should have been much more dangerous.

The enemy they were facing was the fallen demonic champion of one of the outcast gods. A massive winged powerhouse with enchanted armor and weapons that gave it all sorts of extra fun abilities, including full awareness of everything around him (No back attacks). A breakdown of the relevant stats:
*Attack skill 27. Rounded down from 27.75, after including an average of the SM penalties it would face against the PCs, and including any magical enchantments. Incidentally, I rounded down on the demon's stats, and up for the PC's stats.
*Damage 11d(100) cutting. I actually forgot entirely about the AD when doing the comparison, so it should have been doing a tiny bit more damage.
*Primary defense skill was a sword parry at 20. It had Weapon Master, and was using a Bastard Sword, so it could conceivably get the multiple-parry penalty down to only -1. If forced to defend enough times, it could fall back on its secondary defenses of an "unarmed" parry at 15 (The enchanted arm armor let it parry as a weapon, so no damage), and a dodge of 14.
*30 hitpoints. Which, impressive as that is, leads me to a complaint about the system, that assumes a target is automatically down at -1xHP, which isn't very fair on a HT15 High Pain Threshold demon. He didn't actually go down until -5xHP.
*DR ranging from 4 to 32, with most locations having 20-29. All DR had a level of Hardened, negating all Penetrating Weapon enchantments except, amusingly, the mage's. When I did my calculations, I ran it as the demon having only DR4, which could be achieved without too much difficulty (An extra -2 to ensure that, though I neglected to assign that penalty to the PC's attack skills for the balance system).

Ignored for the sake of this comparison was the demon's empowered hoof-stomp power, which had attack skill 22, and did 11d+2 crushing with both a 1-point burning follow-up (Incendiary attack) and an 11d+2 crushing explosive follow-up. It was too messy to figure it out, since its damage output depended on how tightly-grouped its enemies were, and even a successful dodge would only divide the explosion damage by 3 (Plus the somewhat-armor-bypassing trait of explosions). In the end, this power ended up being the only thing that dealt any damage to the PCs, and not very much.

As for the PCs, they consisted of a shape-shifting dual-wield warrior (The only truely-dedicated fighter of the group), a half-dragon mage (Who apart from a 1-point Curse one second before the demon went down, mainly relied on his named spear), the mage's bound succubus familiar (Had a fire-whip power that could entangle most enemies, but that part was too weak to do anything to the demon), and a young dragon (Who's main offensive contribution was a couple precisely-aimed blasts of his lightning "breath").

Party averages were attack skills of 19, damage of 4d (Almost all cutting, a little burning), defenses of 13 (Rounded up), 16hp, and DR of 4.25 (Completely zeroed out by the demon's armor-divisor, and only counting the chest DR, which was universally the highest). Most of this was boosted up by the warrior, of course. After all the skill boosts, blessing, and SM adjustment, she had a skill of 29, did roughly 7d cutting damage IIRC (Or, optionally, about 5d impaling), had a parry of 17 (The only Weapon Master, and with two weapons), 21hp, and DR8 on every location except some of the face (DR2).

Incidentally, all combatants, including the demon, were capable of flight, with the sole exception of the warrior. The battle took place on a raised path in a subterranean chamber. The ceiling was about 30 feet over this path, giving everyone room to fly if they desired. Immediately off the path was an abrupt fall, ranging from 20-50 feet or so, with all sorts of jagged rocks and spikes. There was no natural light, but the PCs brought in enough that the whole chamber was quite well-lit, and between that, the demon with Dark Vision, and most of the PC party having some level of Night Vision, I didn't bother giving any lighting penalties. The only damage the PCs took was from the largely-unavoidable explosive attack. The only time that one of the PCs might have been hit, it was negated by luck (Another thing it seems the system has no easy way to account for).

[Nymdok]

SO the lines IM getting from this are....

1 Bad Guy

A:27
DMG: 11d(100) cut
Defense: 16
HP:30
DR: (Roughly) 25 (Hardened)

4 Good Guys
A: 19
Dmg: 4d Cut
Defense: 13
HP: 16
DR:4


So Demon on Offense:
27 v 13 => 15 v 7 => 80%
11d(100) vs DR 4 => Expected 38.5
80% of 38.5 = 30.8 or 31 x 1 attack = 31

PCS on offense
19v20 (DAing would hurt us) = 10%
4d Cut vs 25 DR = 0 (6x4<25)


Your guys prevailed in this combat?

I must be missing some data.

Lets consider the Warrior alone.....
After all the skill boosts, blessing, and SM adjustment, she had a skill of 29, did roughly 7d cutting damage IIRC (Or, optionally, about 5d impaling), had a parry of 17 (The only Weapon Master, and with two weapons), 21hp, and DR8 on every location except some of the face (DR2).

Attack: 29
Dmg: 7d cut / 5d Imp
Defense : 17
HP: 21
DR: 8

NOw we look again:
Demon on Offense
27 v 17 => 15 v 11 => about 40%
Dmg 11d(100) vs DR 8 = 38.5 again.
40% of 38.5 = 15.4 cut => 23 avg per attack vs Warriors 21 hp. 3 Hits to hp<HP-1

Warrior on Offense
29 v 16 => 15 v 9 => 60%
Dmg 7d cut vs DR 25 => 28.5-25 = 3.5
Dmg 5d Imp vs DR 25 => 17.5 - 25 = 0

3.5 x 1.5 = 5.25 versus demons 30. about 11 hits.

Unless there came a crit storm, or some other factors that arent listed here, I dont really see how your guys could have won.

Assuming that the warriors stats were brining up the combat stat line, then that means that the others are below the average, meaning they will have an even HARDER time.

What arent we accounting for?

Nymdok

[Nymdok]

As for -HPx1 and Luck:

The biggest danger from HP -1 isnt really the Death per se, its the Conciousness checks you have to make every round.

With a 15 HT you are more likely than not to make 14 Health checks in a row. However, you do take a penalty for each multiple of HP below zero. Most importantly, your reward for staying concious is that you get to make another check next turn, and are likely to take more damage in the process witch will lead to further penalties.

50% Percent success rate for consecutive HT Checks

Effectve HP 15 => 14 HT CHecks in a row
14 => 7
13 => 4
12 => 2
11 => 1

So if your HT is 15 and your effective HT is 14, you can expect to make about 7 HT checks before you pass out if you take no further damage.

As you continue to take damage though, it becomes less and less likely that you will make it.

For that reason, I took HP -1.

As for luck, Your right, I have no way to account for it easily.

Nymdok

[lexington]

Did they just never get hit? I mean without even going into an analysis it look like the least damage the demon could possibly do would end up as a Major Wound.

[Anthony]

Originally Posted by Nymdok

As for luck, Your right, I have no way to account for it easily.

Subtract the first critical hit? As a note, a fight that was supposed to be hard turning into a blowout isn't all that unusual in GURPS; any system designed around high per-hit lethality and low hit chance will have considerable randomness.

[Icelander]

Originally Posted by Nymdok

[...]
Your guys prevailed in this combat?

I must be missing some data.
[...]

Unless there came a crit storm, or some other factors that arent listed here, I dont really see how your guys could have won.
[...]
What arent we accounting for?

As noted multiple times, your model is useless for modelling actual combat, with actual people making the decisions for the character. This is because it is a naive mathematical simulation that assumes combatants standing absolutely still on level ground hacking at each other's strongest armour.

You essentially have two options here. Either bite the bullet and accept that in order to model GURPS combat, you have to account for several more orders of complexity and start to budget a year or so of serious programming, or, alternatively, realise that even that much work would probably be inadequate and you should just evaluate combats based on rough 'feel'.

[sir_pudding]

I'm not sure how I'd even begin to apply this system to the last two combats in my game. Not all the characters decided to physically go to place that the fight took place, they could have though. Should I have accounted for them? One character is very small, but pilots a formidable machine in combat. Do I use his stats or his construct's? Another character decided to use astral projection to go to the location, but was unable to penetrate through the ward there. How does that get addressed? The first part of the fight involved creatures that could meld through metal, and had attacks with Surge. How do I account for their effects on the construct? The second half was meant to be a mexican standoff, and intended to lead to a chase rather than a fight (which it did), how does this system handle combats in which one side is trying to avoid a fight altogether?

[Nymdok]

Originally Posted by Icelander

As noted multiple times, your model is useless for modelling actual combat, with actual people making the decisions for the character. This is because it is a naive mathematical simulation that assumes combatants standing absolutely still on level ground hacking at each other's strongest armour.

You essentially have two options here. Either bite the bullet and accept that in order to model GURPS combat, you have to account for several more orders of complexity and start to budget a year or so of serious programming, or, alternatively, realise that even that much work would probably be inadequate and you should just evaluate combats based on rough 'feel'.

I'll take option 3 then and see if Phoenix can post up some more details of how the day was won. :)

Again, I reject that its useless.

The trick here is to find out what happened and where the model failed, if at all. From there, it might pick up some improvements.

Nymdok

[Nymdok]

Originally Posted by sir_pudding

I'm not sure how I'd even begin to apply this system to the last two combats in my game. Not all the characters decided to physically go to place that the fight took place, they could have though. Should I have accounted for them? One character is very small, but pilots a formidable machine in combat. Do I use his stats or his construct's? Another character decided to use astral projection to go to the location, but was unable to penetrate through the ward there. How does that get addressed? The first part of the fight involved creatures that could meld through metal, and had attacks with Surge. How do I account for their effects on the construct? The second half was meant to be a mexican standoff, and intended to lead to a chase rather than a fight (which it did), how does this system handle combats in which one side is trying to avoid a fight altogether?

If they're not affecting the combat, no need to account for them.

Dont know the chrachter, but assuming thier tiny size makes personal melee pointless, Id go with the construct.

Meld through metal and attacks with surge means DR =0 if Im not mistaken (and assuming the Construct is made of metal).

It doesnt. Its a combat model not a retreat model :)

Nymdok

[sir_pudding]

Originally Posted by Nymdok

If they're not affecting the combat, no need to account for them.

How could I have known in advance that they'd choose to split the party? The fight was actually a lot tougher than I had originally expected, because they surprised me by doing this. What about the magician? If he had succeeded in his attempts to penetrate the ward, then he'd been a factor? Shouldn't the strength of the ward have been something that should have been accounted for (and honestly here is where I think I balanced badly, logically the Ward could have been any strength as it probably suffered some attempts by the (elemental) creatures to escape, I set it at 10 which proved to be too high)?

Dont know the chrachter, but assuming thier tiny size makes personal melee pointless, Id go with the construct.

Sure that makes sense, with the caveat that he's more likely to use ranged weapons. Another factor is that one of his attacks is a rocket launcher. It's powerful, but one-shot (well 12! shot really) and the reloads are costly. He could use it in any given fight, but isn't likely to.

It doesnt. Its a combat model not a retreat model :)

Any combat could involve a withdrawal though. If your system doesn't account for this, then how do you expect to use it to balance encounters unless you artificially constrain the ability of either side to retreat.