King's Blade and weapon balance issues

[Corvo]

For discussion's sake, this is my data:

Barding Wordweaver (4-6-4, no favoured), weapon+shield, weapon skill 3, battle skill 1, max hope per combat 4, + Keen reward.
against an Orc chieftain (armour 3d), all options set as standard.

I ran the combat with the “lots”* times, times 7 for every weapon (21 times total).
It resulted in a pretty tight spread.

Victories for the Hero:
wielding a sword: between 45,1% and 46,9%
wielding an axe: between 46,1% and 47,8%
wielding a spear: between 41,3% and 42,6%


When I ran the same hero against 2 Messengers of Lugburz (I was looking for a tough fight against armour 2d), these were the results (here I ran the combat “lots” times 12, four times for weapon):

Victories for the Hero:
wielding a sword: between 60,4% and 62,5%
wielding an axe: between 57,5% and 60,6%
wielding a spear: between 60,7% and 61,3%

*@Elfcrusher: how many times is “lots”?

[Corvo]

@Michebugio

I was making some trial combat Woodman vs Orc Guard. Weirdly, the Spear was always the better weapon. Then I noticed that the simulator write "Orc Guard, armour 2d+0" (instead of 2d+4)
So maybe it's just an error (we already know that the Spear is the better weapon against 2d armour)

[Michebugio]

Victories for the Hero:
wielding a sword: between 45,1% and 46,9%
wielding an axe: between 46,1% and 47,8%
wielding a spear: between 41,3% and 42,6%

Victories for the Hero:
wielding a sword: between 60,4% and 62,5%
wielding an axe: between 57,5% and 60,6%
wielding a spear: between 60,7% and 61,3%

Yes, these are definitely coherent values for Keen weapons with the statistics I posted before, though I would have expected slightly higher ratings for Keen spear.

I think we may now safely conclude that giving a "default Keen" to all weapons is a good alternative to my house rule, since it has a significant effect on weapon balance

[Glorelendil]

When I ran the same hero against 2 Messengers of Lugburz (I was looking for a tough fight against armour 2d), these were the results (here I ran the combat “lots” times 12, four times for weapon):

I think it's a good idea to start with fights that are close to 50:50 to get the best sense of the impact of individual variables.

*@Elfcrusher: how many times is “lots”?

Read the mouseover tooltip. It's 10,000 divided by the number of enemies. So 5,000 for two enemies.

[Glorelendil]

@Michebugio

I was making some trial combat Woodman vs Orc Guard. Weirdly, the Spear was always the better weapon. Then I noticed that the simulator write "Orc Guard, armour 2d+0" (instead of 2d+4)
So maybe it's just an error (we already know that the Spear is the better weapon against 2d armour)

Yup. Simulator is not adding attribute level to favoured armor; not a bug, just haven't implemented it yet. (That's why the single fights are important...picking through them is the only way to figure out what's not working.)

[Michebugio]

I was curious about armor effectiveness and how my rule or Corvo's may affect armored characters.

I just ran some (Lots, actually!) simulations with a Barding (By Hammer and Anvil), Weapon Skill 3, Battle 1, no Hope expenditure, Open stance armed with a Sword and a Buckler against an Orc Chieftain. First I gave him a Hauberk and a Helmet, and the results were:

without Michebugio's rule: Win rate 28.5%
with Michebugio's rule: Win rate 31.5%

Well, 3% more is not a great deal. My rule doesn't impact heavy armors as much as I hoped.

Then I just removed the Hauberk and the Helmet: the Barding is now naked against the Orc Chieftain, armed with only his Sword and Buckler! The results were:

without Michebugio's rule: Win rate 33.5%
with Michebugio's rule: Win rate 36%

Do I really see nonsense here, or are there any good reasons why a NAKED character is actually a better fighter than a fully armored one??? This is going off topic, by the way...

[Glorelendil]

Do I really see nonsense here, or are there any good reasons why a NAKED character is actually a better fighter than a fully armored one??? This is going off topic, by the way...

It's possible my code is bugged, but we've explored this before (there's a whole thread about it) and I think what you're seeing is the result of becoming Weary after a single hit. Try it with medium armor and see if win rates are higher.

[Michebugio]

Try it with medium armor and see if win rates are higher.

Done: win rates nicely and steadily increase from 28.5% to 33.5% as the armor gets ligher...

[Evening]

No no no, my rule was tested for all scores of weapon skill from 2 to 6. The percentages you see in my previous posts are the average values based on separate results for each weapon skill from 2 to 6, not just for 3d skill. The scores already take into account the skill progression, so the average is for the whole character's "history", let's say.

You lost me here. Are you talking about on the simulator?

[Michebugio]

You lost me here. Are you talking about on the simulator?

No, I was talking about statistics: "tested" means simply evaluating the % to score a Wound for each weapon skill. Sorry about the confusion, I didn't even have in mind Elfrcusher's simulator yet!