Author Topic: Reallistic Combat Model for Native  (Read 150678 times)

Offline Fisheye

  • Blues Brother
  • Guildsman
  • ****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #30 on: September 28, 2007, 01:06:13 PM »
Geh, I'm not that invested in BoW anyway. It was always supposed to be a 5-minute lark, for some bizarro reason it's now the 4th most popular download on mbrepository...

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #31 on: September 28, 2007, 05:53:05 PM »
Geh, I'm not that invested in BoW anyway. It was always supposed to be a 5-minute lark, for some bizarro reason it's now the 4th most popular download on mbrepository...

I know that feeling, oddly enough.  The entire RCM project was originally just trying to get the blades and horses in ONR to behave right.  After two months of trying to explain my ideas on forums, I built the first fully playable version in two days, having never programmed in Python before.  (Told Fujiwara to give me the source code and 48 hours, I would make it work or die trying.)  Now it's the standard for the major mods, and I feel like half the M&B mod community are waiting on me to get this right.

But now that the RCM mods include half of the modded games, I have to be a little more committed.

How are we going to run down this damage bug?  I'm still seeing swords and such reading way too high.  After that rather cheap work-around, the arrows seem playable, but not really "right".  Other stuff is still much too high.


Offline Fisheye

  • Blues Brother
  • Guildsman
  • ****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #32 on: September 28, 2007, 08:01:56 PM »
I'm realising that the weapon mashup in Native can be either realistic, or balanced, but not both.

Plate armor is just too good, and rightly so. Swords and arrows don't do jack against heavily armored enemies. Half the so-called "weapons" in the arsenal are completely out of date. The troop AI is not smart enough to switch weapon types depending on enemy armor and weapon penetration.

I think the Pope (forum dude, not Benedict) got the right imagery when he said it was like a T-34 tank rolling onto the field against Napoleon's troops. Too far out of period, mashup doesn't work. That's why the RCM is best for the historical-type mods (or anything with a consistent tech level, like TLD which doesn't have plate).

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #33 on: September 28, 2007, 10:29:28 PM »
RCM-Native has always been a developers' tool.  It was never balanced for long term playability, nor has any claim been made of such.

However, as a developers' tool, it can be extremely useful to see what kinds of equipment interact in what ways.  Long, relatively light swords SHOULD be ineffective against really heavy armor ... if they are not, then we have a problem with the stats.

Any substantial mod will have to design the equipment level to reflect this.  This is true no matter what model you use.

Don't forget, however, that balance can be developed in other ways.  Mesoamerica has excellent game balance, in spite of one side having top-end Spanish steel from the very high point of Gothic weapons and armor, and the other side being neolithic.  That's because the neolithic side has a LOT of troops, and they're tough, fast, and as well-armed as neolithic people can be, while the Spanish suffer from low manpower and apparent supply shortages (i.e. not all of their troops are as well armed as one might think) that render them quite vulnerable.  ONR, also, draws a deliberate contrast between samurai and peasant equipment - like a T-72 rolling onto the field against a bunch of modern farmers with squirrel-rifles.  Balance there is developed by the high cost of equipping and maintaining samurai.

So if you're working up Band of Warriors, or whatever, consider non-conventional ways of developing balance.  It can be very rewarding.

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

Assuming, of course, that I can ever in my lifetime get the bloody formulas figured out.  They seem to make no sense at all, considering what I knew about the previous ones.  This is an absolute programming nightmare.

Everybody ... please, help.  I need to know how the base damage formula works.  If you don't know, collect a bunch of numbers and guess.

Damage used to be random between 50% and 100% of rated value, then minus armor (calculated the same 50-100%).  Now, either that has changed, or huge bonus factors have been added for various reasons ... I can't tell which.

Until this is decoded, THE RCM .89x IS NOT READY FOR USE.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #34 on: September 29, 2007, 05:05:52 AM »

Eureka!   (That's Greek for "Yikes!")

I found the melee damage bug.

It seems that, starting with .890, strikes now have different base damage.  A quick swing has a max damage equal to the rated number.  A swing that is drawn fully back (i.e. the button held down until the weapon stops moving) does 125% of rated, and a vertical drawn back fully does 130% of rated damage.  Thrust attacks don't seem to be affected.

When combined with even a couple points of powerstrike, to compound the effect, damages can easily go up to 160% or more of rated damage.  With several points of powerstrike ... 160% of 130% ... you get the idea.

Previously, damage was deducted for less-than-perfect attacks ... not added for heavy ones.  In short, it means all the swing-damage weapon values in the RCM are 130% of what they should be.

I'm going to try to have this fixed by some time tomorrow (local time).  I may have to work all night to do it.


Everybody waiting on this ... stand by.  Going to need some people to help test it, again, soon.  Maybe we'll get this fool thing to work soon enough, and the major mods can start porting their stuff to .89x ...looks like there might be hope, now.  The Power Draw thing is still a cheezy work-around, but if that's the worst thing wrong, I'll feel better.


Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #35 on: September 29, 2007, 09:26:44 AM »
(sorry about the double post, but announcement threads tend to get that way)

Everybody:

Alpha 3 of RCM-Native for .89x

http://rapidshare.com/files/59091695/RCM_Native_89x_alpha3.zip.html

This version is playable.

The numbers look strange, I know.  The whole calculation system was redone - in .89x, a fully extended attack is 125 to 130% of listed damage (before the previous base calculation of random .5 to 1.0 of base damage).  That means everything that was 40 before is 30 now, in order to make the final numbers come out the same.

Bow damage is still a little odd ... the higher power draw bonus, plus the fact I can no longer hide points from the power draw by assigning them to the arrow, is creating a computational problem.  It's mostly functional, but it means that you generally have no need to move up to a larger bow until your power draw goes beyond the maximum of the previous bow.  It's the best I could do with the tools at hand.

Anyway, this one is playable ... although we are still hunting bugs.  Everybody try it out.

Offline Fisheye

  • Blues Brother
  • Guildsman
  • ****
  • Posts: 427
  • My scalp's so bright, I gotta wear shades.
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #36 on: September 29, 2007, 10:51:19 AM »
Ron, the RCM version of Band of Warriors only got 120 downloads and I got absolutely no feedback on the thread.

Meanwhile the non-RCM Band of Warriors currently has 1200 downloads. I guess the masses have spoken on this issue.  :(

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #37 on: September 29, 2007, 06:25:13 PM »
Ron, the RCM version of Band of Warriors only got 120 downloads and I got absolutely no feedback on the thread.

Meanwhile the non-RCM Band of Warriors currently has 1200 downloads. I guess the masses have spoken on this issue.  :(

Several problems there ...

One, you used an unplayable alpha version.  I suspect that if I had downloaded that for the first time, I would not have thought much of it either. 

Second, you posted it on the Taleworlds board, where the RCM is largely unknown, and those who do know it have only seen the retrofits to native (hardly long-term playable) or some of the crude copies floating around.  Most who saw that post probably had no idea what you were talking about, and those who did probably figured they didn't want to mess around with some kind of a test version.  Plus 90% of the downloaders probably never saw the thread - they just checked the Repository for whatever was new and didn't sound like an internal alpha test of some kind.

Update that version to the latest - alpha 3 - and announce it over here in the new / announcements thread.

I mean, the survey on TLD placed the RCM at preferred roughly 3 to 1 over .808 Native, even in spite of the major balance issues.  The Mesoamerica survey came out a ratio of about 3 to 2 between RCM and some unspecified future version "in-between", with almost no votes for .808 native at all.  It depends on who you ask... people who actually play with the RCM mods for more than two minutes are usually immediate converts.  On the other hand, those who speak out against it most loudly usually admit that they never really played any of the mods for more than a few seconds, and are basing their argument on math that they don't understand.

It's not what you do, it's how you do it...

Offline mfhberg

  • Apprentice
  • *
  • Posts: 49
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #38 on: September 30, 2007, 09:50:37 AM »
Just little adjustments for Native RCM

1 - Strength rating for Brigandine (I use 8)
2 - Damage for Morning Star (much heavier than the Military Pick so I up it by two).

Downloaded and am testing the RCM for BOW, but that heroes not showing up for battle bug gets in my craw.

mfhberg
« Last Edit: September 30, 2007, 09:59:10 AM by mfhberg »

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #39 on: September 30, 2007, 10:13:37 AM »
Just little adjustments for Native RCM

1 - Strength rating for Brigandine (I use 8)
2 - Damage for Morning Star (much heavier than the Military Pick so I up it by two).

Downloaded and am testing the RCM for BOW, but that heroes not showing up for battle bug gets in my craw.

mfhberg

Get the alpha 3 for native until Fisheye gets BoW updated.  (Unless that's done ... I didn't check.)  As far as I know, he used the alpha 2 numbers, which were - thanks to changes in the formula - all 30% too high on melee damage and so quite unplayable.

The lack of strength requirement on some armors and weapons was a Native bug ... I didn't add them, out of fear that some units might be unable to use their armor or weapons.

I'll check on morning star and military pick next time I think of it.  Seems like the morning star was rated too low in TLD as well ... I seem to repeat that mistake quite often.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #40 on: October 01, 2007, 09:18:15 PM »
RCM for Native 0.892 Semi-Final.

http://rapidshare.com/files/59655420/RCM_Native_0892_semi-final.zip.html

(Semi-final because I'm not going to bother putting it on the Repository until we're sure there won't be any more bug-fix releases from Taleworlds.)

This is 100% playable, as far as Native goes.  Thanks to everybody who helped me debug it.  The download includes the source code and module.ini, for anyone using it in other mods (which is really what this is about ... I don't really expect that many people to play Native that much).

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

The Native bugs of a number of armors and weapons having no strength requirements - those were left, since changing them could result in some troops being unable to use their equipment.

The bows are a little freaky ... you don't generally benefit from moving up to a larger bow until your power-draw exceeds the limit of the smaller one.  The slight benefit in arrow speed does not make up for the loss of accuracy caused by stiffer requirements.  This is not entirely unrealistic, but it's a little odd for the game.

A lot of changes to AI in M&B .89x - for one, troops actually target horses now.  Makes it a little tougher to use unarmored horse.  I didn't really do that ... I have no control over the AI.  I actually like it, but I didn't do it.

And, as always, questions and bug reports ... send them in, I'll deal with them as I can.

JeffBag

  • Guest
Re: Reallistic Combat Model for Native
« Reply #41 on: October 06, 2007, 12:22:06 PM »
Hi, mind if you upload it to repository or maybe filefront? I can't download from Rapidshare thanks to my genius ISP. Thanks.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #42 on: October 06, 2007, 01:15:37 PM »
Hi, mind if you upload it to repository or maybe filefront? I can't download from Rapidshare thanks to my genius ISP. Thanks.


RCM_Native_0892_semi-final.zip:
http://files.filefront.com/RCM+Native+0892+semi+fi

JeffBag

  • Guest
Re: Reallistic Combat Model for Native
« Reply #43 on: October 06, 2007, 10:02:53 PM »
Thanks a lot, but the link doesn't seem to work?
« Last Edit: October 06, 2007, 10:07:02 PM by JeffBag »

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Reallistic Combat Model for Native
« Reply #44 on: October 06, 2007, 10:24:15 PM »