login

Author Topic: .890 and high damage numbers  (Read 7629 times)

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
.890 and high damage numbers
« on: September 27, 2007, 07:17:02 AM »
We have a little problem here, and it's playing havoc with my ability to get the RCM working.

I am seeing damage numbers in the vicinity of twice what they should be.  Something very serious has been changed in the way the damage scale computes.  I haven't got it figured out yet, but I'm working on it.

If anybody has any info on the new formula calculations, or can get such info ... this is just too strange for me.

Offline Winter

  • I am Tek Jansen!
  • Administrator
  • *****
  • Posts: 476
  • MBX Forum owner
    • View Profile
    • Street of Eyes: The Writing of Ryan A. Span
Re: .890 and high damage numbers
« Reply #1 on: September 27, 2007, 07:56:00 AM »
We have a little problem here, and it's playing havoc with my ability to get the RCM working.

I am seeing damage numbers in the vicinity of twice what they should be.  Something very serious has been changed in the way the damage scale computes.  I haven't got it figured out yet, but I'm working on it.

If anybody has any info on the new formula calculations, or can get such info ... this is just too strange for me.

The M&B damage scale is now exponential rather than linear. You can adjust it in module.ini .

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: .890 and high damage numbers
« Reply #2 on: September 27, 2007, 08:10:14 AM »
We have a little problem here, and it's playing havoc with my ability to get the RCM working.

I am seeing damage numbers in the vicinity of twice what they should be.  Something very serious has been changed in the way the damage scale computes.  I haven't got it figured out yet, but I'm working on it.

If anybody has any info on the new formula calculations, or can get such info ... this is just too strange for me.

The M&B damage scale is now exponential rather than linear. You can adjust it in module.ini .

The speed adjustment is variable in the module.ini ... I already had a talk with that.  It's not what I'm talking about.  I even tried setting that to zero, just to eliminate it from the equation and make sure.

With no speed adjustment at all (i.e. both targets stationary), numbers are coming out all different from what they should.

Before the damage formula was random .5 to 1 of rated value.  This is acting more like it's now .75 to 1.25, or maybe .5 to 1.5  - repeatedly WAY too high.  It could be 1.0 to 1.5 minus armor - it's hard to tell without collecting thousands of examples.  Whatever it is, it's NOT what it was before.


Offline Winter

  • I am Tek Jansen!
  • Administrator
  • *****
  • Posts: 476
  • MBX Forum owner
    • View Profile
    • Street of Eyes: The Writing of Ryan A. Span
Re: .890 and high damage numbers
« Reply #3 on: September 27, 2007, 08:35:30 AM »
The speed adjustment is variable in the module.ini ... I already had a talk with that.  It's not what I'm talking about.  I even tried setting that to zero, just to eliminate it from the equation and make sure.

With no speed adjustment at all (i.e. both targets stationary), numbers are coming out all different from what they should.

Before the damage formula was random .5 to 1 of rated value.  This is acting more like it's now .75 to 1.25, or maybe .5 to 1.5  - repeatedly WAY too high.  It could be 1.0 to 1.5 minus armor - it's hard to tell without collecting thousands of examples.  Whatever it is, it's NOT what it was before.



Alright, have you tried messing with the new armour soak values? Also in module.ini:

armor_soak_factor_against_cut       = 0.5
armor_soak_factor_against_pierce    = 0.33
armor_soak_factor_against_blunt     = 0.25

armor_reduction_factor_against_cut       = 1.0
armor_reduction_factor_against_pierce    = 0.65
armor_reduction_factor_against_blunt     = 0.65


You'll also notice in module_items that Native armour values are significantly higher than they were before. Maybe that's what you're seeing?

Offline Fisheye

  • Blues Brother
  • Moderator
  • *****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: .890 and high damage numbers
« Reply #4 on: September 27, 2007, 08:49:22 AM »
Yeah Ron's already tried all that. I think it's the scaling factor that's throwing everything off. So if a knife has base damage 30, and you hit with velocity 1, it's doing damage 30k where k is some arbitrary number.... which changed since 0.808 so we have to figure out.

By the way, exponential is k^x which grows much faster than x^k (polynomial) in particular x^2. The new scheme grows with the square of the velocity, not exponentially with velocity.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: .890 and high damage numbers
« Reply #5 on: September 27, 2007, 09:15:26 AM »
Yeah, I've already changed all of that.  Several times.  I've even set up arbitrary numbers and ran test cases on target dummies and such.  If it was that simple, it would just be a matter of time to do the math, and I wouldn't be concerned.  I even tried setting the velocity bonus numbers to zero, and I still got numbers that were sometimes double what they should be.

The base formula is different.  I need to know exactly HOW it is different.  I suspect that, instead of using between 0.5 and 1 of rated damage, it's now using something like between 0.5 and 1.5, or some such variation.

And since there are like a half dozen mods waiting on me to get this right, I needed to know about 12 hours ago.

If somebody can't come up with a clear response pretty quickly, it looks like I may be spending my entire week of vacation (Chinese National Day, 1 October) collecting statistics and trying to figure out how they add up.  I may pull out my hair.  And the teams of a half dozen mods may grow more than a little impatient as well.

Offline Fisheye

  • Blues Brother
  • Moderator
  • *****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: .890 and high damage numbers
« Reply #6 on: September 27, 2007, 09:17:03 AM »
PM Armagan and ask him for the formula before he goes on vacation next week. We should get more people to post here to show we'd like it please.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: .890 and high damage numbers
« Reply #7 on: September 27, 2007, 09:27:50 AM »
PM Armagan and ask him for the formula before he goes on vacation next week. We should get more people to post here to show we'd like it please.

Do you still use the Taleworlds board at all?  If so, could you get that?

I haven't logged on to Taleworlds since ONR moved off ... I don't even remember my password over there.

Offline Winter

  • I am Tek Jansen!
  • Administrator
  • *****
  • Posts: 476
  • MBX Forum owner
    • View Profile
    • Street of Eyes: The Writing of Ryan A. Span
Re: .890 and high damage numbers
« Reply #8 on: September 27, 2007, 09:30:28 AM »
By the way, exponential is k^x which grows much faster than x^k (polynomial) in particular x^2. The new scheme grows with the square of the velocity, not exponentially with velocity.

I guess it shows that I was never any good at maths, eh? :P

Speaking of, I could really use an explanation of the new maths operations in the new module system, particularly for my header_operations documentation effort.

Offline Hellequin

  • Journeyman
  • ***
  • Posts: 254
    • View Profile
Re: .890 and high damage numbers
« Reply #9 on: September 27, 2007, 09:47:05 AM »
The maths ones I can do you, Winter - which ones are you concerned with?

As for the damage model, Ron, I assume you caught the bit where Armagan said he was changing the armour metrics in the new version (cf. here), and that since you mention testing on dummies this isn't the issue here.  Perhaps he's put in a multiplicative effect on Strength or something, to go with this?

Offline Fisheye

  • Blues Brother
  • Moderator
  • *****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: .890 and high damage numbers
« Reply #10 on: September 27, 2007, 10:01:43 AM »
One piece of the puzzle. Fellow worshippers, I bring you the Word from on high:

Hi Fisheye,

You have got it mostly right. The correct formula should be:

Reduced_damage =base_damage - (armor * soak_factor)
Final damage =  Reduced_damage * (1 -  (armor/100 * reduction_factor)) 


http://forums.taleworlds.net/index.php/topic,26790.0.html

EDIT: Math operations discussion split to separate topic.
« Last Edit: September 27, 2007, 12:48:03 PM by Fisheye »

Offline Hellequin

  • Journeyman
  • ***
  • Posts: 254
    • View Profile
Re: .890 and high damage numbers
« Reply #11 on: September 27, 2007, 02:28:15 PM »
Thanks, Fisheye. Quite right.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: .890 and high damage numbers
« Reply #12 on: September 27, 2007, 05:09:33 PM »
The base damage is where it's coming out too high.  Weapons that should be doing 50 or 60 points of damage tops are coming out with numbers like 75 or 114... after anything that armor may have done for the victim of such.

It's not the armor formula ... I eliminated it for testing, to make sure ... and it's not the speed bonus adjustments, which I likewise changed around just to check.

I can eventually isolate it by trial and error, but it will take forever.  I need that base damage formula.



Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: .890 and high damage numbers
« Reply #13 on: September 28, 2007, 02:57:48 AM »
Sent a note to the M&B dev team ... haven't heard back yet.  I did, however, find one of the problems.

It seems Power Draw has been increased to 12%.  Plus, note that it is up to three PLUS the PD requirement of the bow - i.e. a long bow with PD requirement 3 can go up to PD 6, or 72% above listed rating.  It was only like 6% before ... the increase is phenomenal. 

(I only figured that one out when I noticed that crossbows didn't have this problem.)

That means I have to rework all of the bows to account for this.  A mathematical nightmare.

Still can't explain why certain melee weapons do way too much damage.

Everybody stand by ... We'll have a working RCM version soon enough.  Or if not, it won't be for lack of trying.


---------------------------------------------------



Edit:

Here we go again.  Alpha 2.  Check it out, tell me what you think.

http://rapidshare.com/files/58818909/RCM_Native_891_alpha2.zip.html

I think I got the bows back to reasonable.  There is a lot of variation because of PD skill, but at least it's within a reasonable range now.

Also lowered speed bonus again.  It was still too high.

Plus that backwards armor number thing, noted previously.

This is still Alpha test ... help me test it, don't assume it's ready for anything just yet.

And, as before ... we need some eyes on this thing.  The more people who help test it, the sooner we can get it debugged in time for the major mods porting to .89x ...
« Last Edit: September 28, 2007, 04:09:24 AM by Ron Losey »

Offline Fisheye

  • Blues Brother
  • Moderator
  • *****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: .890 and high damage numbers
« Reply #14 on: September 28, 2007, 11:44:11 AM »
Regarding Power Draw:

It's not 70% variation. Since the bows have MINIMUM PD requirement, and the effect is capped at +3 over minmum, every bow only varies from 0-3 points above minimum, so that's a 0-36% range.

I just had a look at the code.

Arrow damage: 33 cut.

Hunting bow: thrust damage 1, cut. (PD req 0)
Short bow: thrust damage 1, cut. (PD req 1)
Strong bow: 10, cut (PD requirement 3)
War bow: 10, cut (PD requirement 4)

You're kidding, right? I cannot believe that this works. NObody in their right mind would use anything other than a hunting bow. The higher PD requirements totally fuck with your accuracy at anything but GODLY proficiency levels.
« Last Edit: September 28, 2007, 11:54:19 AM by Fisheye »