Combat Damage Model (RCM) - merged threads

[Ron Losey]

RCM for TLD3.1 http://www.wupload.com/file/2662215892/RCM_TLD31.rar

(Finally got both the mod and the new M&B enhanced version to download ... my Internet connection is REALLY bad these days.)

Guys, bad news ... your equipment stats are all over the board, they look like you took them from a random number generator.  Sticks effectively do more damage than swords.  Bows will penetrate a steel helmet but have exactly zero percent chance of stopping an unarmored man, even if you hit him square in the chest at point blank range.  Any attempt at getting "game balance" like this will be utterly futile, and tweaking a few numbers will just make it more absurd (if that is possible) but equally futile.  It's a hopeless mess right now.

Now, I don't have a great deal of time to help these days, but I'll do what I can.  I can make the changes (if you send me the Python code) or I can tell you what to use for baselines and you can work on it ... either way is fine by me.  Either way, this is going to have to be done as soon as possible ... unless somebody was really trying for that comic book/D&D "I don't mind having an arrow stuck through my liver because I have HIT POINTS!" feel to the combat sequences.  (Or the movie look where the tin-man armor was apparently made of paper.)  Because that's surely where the numbers are sitting right now.

Sorry about that ... but I did say in previous posts, personal mail, and whatever, that somebody should tell me when they're ready to do the equipment stats, before making a release.  Because obviously, item stats based on Native are not likely to give a feel of intensity, or believability, or much of anything else besides that nagging feeling that certain critical elements of the program were written by monkeys and/or kindergarten children who watch too much television.

So tell me how you prefer to handle this, and I'll try to work within your work flow/timetable/whatever.

[GetAssista]

Hey, Ron!

We remembered about your proposition of developing proper RCM for TLD all the way, don't worry You did not get any message prior to the release because TLD weapon/armor stats did not arrive to a finalized state yet. It's a huge work that would have shifted the release date significantly, and we did not want that. You are right, current weapon starts are messed up somewhat, even related to standard MB damage model. We will continue tweaking them with upcoming patches, in the direction of usual M&B damage model.

I'll PM you necessary txt and py files on weekend, and you would have complete freedom in choosing their stats. I personally have no doubt in your abililty to deliver an excellent damage model. We, sadly, had no time for that.

[Ron Losey]

Thanks for the vote of confidence ... although, in all honesty, I don't think anyone was going to question my abilities on the subject.  Of course, I'll still be nervous about it.  Then again, my way of making sure I got them right is to see if a knife duel in the game causes flashbacks ... maybe I'm going at this the wrong way.

Anyway, roger that.  Awaiting the files.  Not sure how long it will take, as you noted ... but hopefully not too long.  I developed the original RCM for ONR in 48 hours ... so it can't take all THAT long, unless I just get really sidetracked.  I do not, however, care to repeat the 48 hour programming marathon, so don't expect that kind of performance again.

And for the record, they're messed up more than "somewhat" ... they are totally screwed up beyond all comprehension.  It will be a total "start over, from the top" project.  But that's OK ... that's the way the RCM rescale works.

I've read all of Tolkien's stuff, but I don't claim to be an expert on it ... I may have to collect opinions on the desired "feel" of certain items.  Or, regardless, there may be bugs in that regard.  I make no promises on that front ... we'll just have to iron those issues out as we go.

Standing by for those files.

[Merlkir]

Feel free to ask Tolkien-related questions, I'm sure the "magical/mythical" nature of some of the items will prove a challenge.

[Triglav]

Cool that you're on it.

Keep in mind several things.

Our basic division of armours, not saying it's a good one, but to signify why some stuff is as is was. Orcs generally not beyond 25, men around 30, max 40, and elves and dwarves up to 50.
This to make battles balanced where we wanted them to go, so that orcs are cannonfodder, easily killed, elves and dwarves and Gondor knights are Terminator tanks, hard to beat...you get the idea. So when you change stuff, kinda try to devise it so that an orc counts for half a man and a third of an elf roughtly. We like our low level orcs to be one hit killed. They could pack more of a punch than they do now though.

Also price/quality match. It is important to keep weapon diversity and quality. So that starting characters have something to look forward to that their lvl1 weapon is shit and they have to work hard for the next and next up to top quality eventually.

So before you go actually changing numbers would be good if you devise a formula to encompass all that TLD wants to achieve, then we can check it over, then go into implementation.

[MadVader]

What I always felt was missing was more stat diversity and subtle stat tweaks that make items distinct (also names.. more distinct and less generic names). There should be some upgrade paths for different quality items that belong to a single faction, or stat trade offs for similar items.
(How do you know which are more valuable? Look at who is wearing them.)

You should note that some of the items are already tweaked to be more appropriate, but some have more generic stats. I suggest a more evolutionary approach and not ripping everything to shreds, but that depends on what we want to do with the item stats.

P.S. I was pestering GA to do this for ages, and he actually did a lot of work, so I find the OP tone offensive. Even if you think you are the God of Item Stats, you simply don't talk down people like that.

[Northcott]

It would be nice to see weapon lengths closer to the visual representation of the model: there's things like the Rivendell sword with a 94 length -- but the model is significantly shorter than the Lorien longsword, which has an 87 length.

Light armour costs just as much as heavy armour... and speaking of heavy armour, the Dunedain Ranger leather armours are not only expensive and provide very little protection, but weigh more than some medium-heavy armour sets -- 19 weight on it.

As for realism... that's always a tricky thing in these games. We (in the grand, M&B-loving community sense) are attempting to abstract some pretty complex physics along with wild elements of chance with really basic numbers. It never quite gets realistic. Nor should it, necessarily.  Whether or not it's entertaining to play is the most important thing, with the notion of capturing the feel of the source material perhaps coming a close second.

M&B has the infamous shield bug. The AI won't take cover from arrows. You can't duck. All these, and far more, guarantee that realism is a relative and precarious thing. So long as the items play with some sense of balance of utility or function among the weapons, it's all cool.

[Ron Losey]

Merlkir:  Actually, "magic" is easy to do on a RCM rescale.  Once you have a scale for real stats, just figure how much more strength the "magic" properties add.  If Mithril or Dragon Scales are supposed to be impenetrable, just figure them like 200% of steel ... that should make it so that even depleted uranium antitank rounds wouldn't go through.

Triglav:  I understood the basic breakdown in craftsmanship.  I have read Tolkien's work, even if it was a little while back.  I fully intended to represent this, however, in a more logical way ... after all, metal is metal, so the difference between light and heavy armors should be more significant than the craftsmanship penalty/bonus.  Fortunately, RCM numbers give me more latitude in which to work ... I estimate that common (non-"magic" or special material) elven armors will likely go up to 80 or more, so the absolute heaviest orc armors topping out at around 55 or 60 (and most everything else much lighter) will be a pretty serious penalty for them.  That was how I handled the RCM optional package for the last version of TLD, and it seemed to work pretty well.

Madvader:  Wait and watch.  I've done this before.  Triglav, for one, knows exactly what I'm going to do - he was part of the original item data conversions for Onin-no-Ran.  The results are impressive.

Northcott:  You may not be familiar with my work.  Wait ... you will be.  Most of the others here are familiar with my "realistic combat model" conversions.  They're real enough to be considered a "simulator" of simple-weapons combat, even in spite of the little bugs inherent to M&B.  I have to really twist the stats to get it to work, so the final numeric data really looks odd - but it works, to a precision that lets players use common logic to decide what a weapon will do.  If being hit with an arrow would be really bad in real life, it will be exactly that bad on your character when I get done.

The weapon lengths being off ... that's what I was afraid of.  That could take forever to work out.

[WindusAndar]

Dear Ron,

I have a mostly off-topic question to pose to you, if you don't mind.

In the last version of TLD for .808, was the RCM model developed by you implemented in the download package, or did it have to be downloaded, installed and used separately?

Which of these systems would be followed in the current version of TLD?

With regards,

[Ron Losey]

WindusAndar:

If I recall correctly, the optional RCM for TLD .808 was included in the downloads, but you had to manually switch out the files.  That was largely irrelevant, however, since none of the rest of the game was designed or balanced for RCM - so it made the combat more interesting, but the campaign largely unplayable.  There were massive balance issues in the .808 version anyway, and the RCM conversion heavily drew attention to them.  That was a truly unfortunate side effect of trying to do this "optional", and a mistake I would not care to see repeated.

Presumably, according to various communication with the principal programmers, the intention on this version is for me to set up a customized model (base RCM stats plus variations for magic, trolls, and whatever), and then we tweak it from there.  There is little logical reason to try to build and balance a version around Native stats - they're illogical and inherently lacking for game balance, even at best, and absurd and catastrophic to the game at worst.

That said, I am sure the first version will be a little rough ... I was not really involved in the planning phase for the various weapon and armor models, so I will have to try to catch up.  So this will probably look like an initial rough version followed by a whole bunch of minor tweaks released over the next series of patches.  That sort of refining is expected, and as it works out, necessary... especially on something this complex.  I had sort of hoped that some of this could be done before the initial release, to speed the patching process ... but that didn't happen... so we play the cards we are dealt.

[WindusAndar]

Thank you, Ron.

[Shazbot]

Also remember that fodder orcs are for most orcs.  There is no reason why, say, the Black Uruks of Barad-Dur can't be armoured like decently-armoured human troops.

[Ron Losey]

Also remember that fodder orcs are for most orcs.  There is no reason why, say, the Black Uruks of Barad-Dur can't be armoured like decently-armoured human troops.

Exact troop balance comes later.  First we set up the items to reflect reasonable stats, then comes the tweaking of who gets what and how many.

The total difference between armors will be about one weight class per quality distinction, i.e. well-made medium armors roughly matching shoddy but heavier armor.  So the heaviest of the orc armors will likely match what most human troops would have for protective properties, albeit at a greater weight penalty, reflecting that they are compensating for poor craftsmanship with more iron.  (Never underestimate the protective properties of simply wearing more metal.)  Armors that are both well crafted and heavy will have an advantage in protection, as long as you don't define "protection" as "the ability to run away quickly".

(Personally, I find that in real-life deadly force encounters, the ability to run away offers superb protection.  Then again, I don't really need to win an epic war against the dark hordes of the forces of darkness ... just saving my own skin from some dumb but extremely violent would-be mugger is good enough in those situations.)

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

Anyway, I think I can handle it ... let's wait until we have a rough draft before assembling comments, because I suspect that about 98% of this whole line of discussion will be either rendered totally moot or the direction of the commentary will need to go a totally different way. 

Seriously, everybody ... we've gone through this before.  This is not even the first time TLD has been converted to RCM stats.  But it's going to be easier to do the conversion than to explain every detail.  So everybody just wait and watch for a bit ... and then we'll all collect data for the tweaks after a rough draft is out.

[MadVader]

I would have preferred to see item stat rebalancing based on the current (i.e. Native) stat sets and ranges. It worked for Native and other mods, it can work for TLD, "realism" or not.
However, all help is appreciated, even if we end up with two item file variants.
I guess that means we'll still need help with the normal item stats variant, or simply GA will tweak them as we go, and as players report errors and inconsistencies.

[Conners]

Eh? Native worked with its stats... but it didn't seem anything special. Never bothered with anything but the bastard-sword, in Native (you HAVE to use a shield).