Author Topic: Balance issues for 2.3.1  (Read 43821 times)

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Needs Rebalancing?
« Reply #15 on: June 18, 2007, 02:16:13 AM »
You've "never felt that horses are too slow", ... "Top speed feels about right but I'm no rider though."   ???  What kind of statement is that?  Like someone who says a flight simulator doesn't feel natural, but he's never been in an airplane...  Or the guys who keep saying what bows and arrows should be like, and then admit they have never used a bow and arrow in their lives.  (I've killed quite a bit of stuff with arrows, and so have a lot of other people.)

Well, I have ridden some horses, and I had both a lot of technical data and a lot of input from some good horsemen when coming to the conclusion that the Native horses are about HALF the speed of a pretty basic horse.  It took a while to come to that calculation.  Even then, I tried not to add any really strong horses ... that would be even worse (like 3 times the native horses).

As for the damage, a lot of medicine, hunting stories and combat data say that humans (and most anything else) will likely go down and twitch if you stick a Japanese yari through their leg.  In fact, one or two good jabs with a stupid pocketknife will put most people down.  (Average fatal stabbing in North America only penetrates two inches past the ribcage.)  Compare that to test cuts with heavy blades that show a good cut will literally slice a human body in two.  The statistics used in the RCM are about as accurate as you can get in a simple statistical model... but if in doubt, I rated them low - certainly not too high. 

The armor tests were also based on real tests against real armor.  No matter what you see in the movies, it's hard to stick a knife through a coat of maille.  I've tried.  The statistics do a pretty good job of describing the chances of a particular blow doing damage, and if so, how much.  It actually came out much more realistic than I originally expected.  Initially, some of that was probably accidental, or just lucky guesses ... but you can do the test cuts for yourself and confirm this, if you really want to.

Note that "lethal" is a funny concept.  The RCM numbers were based on incapacitating wounds.  A wound that could kill later but that has little immediate effect is "lethal" but not combat effective, while a cut nerve or tendon will put anybody on the ground even if the wound is relatively minor.  Also, wounds that will put a person into shock from pain and bleeding within two or three seconds are debilitating, even if they don't seem to hit anything critical.  I just figured to let the "surgery" skill sort out which wounds were eventually deadly and which ones could be patched up.  Don't confuse a deadly wound by hospital standards with a debilitating wound in combat, or you would get some very funny numbers.

What the RCM numbers go against is common video game/movie logic.  There is a body of totally screwed-up data left over from early D&D pen-and-paper games, combined with some Hollywood and some general misunderstanding - and it forms an opinion of what combat "should" look like in that fantasy world.  That's the problem in Native M&B, why nothing seems to balance - it's based on a body of absurd data, and no number of minor tweaks will make it less absurd.  That was why the ONR team first got involved in this combat rebalance ... and it just grew from there.

But, as previously stated, wait and see - the proof is in the final product, after all.

Offline Merlkir

  • Administrator
  • *****
  • Posts: 5742
    • View Profile
    • My DeviantArt online Gallery
Re: Needs Rebalancing?
« Reply #16 on: June 18, 2007, 02:41:20 AM »
As the horses may be slower than the real ones, I think the speed is about right for MaB. It's just fine for the AI to handle. If you set the horse speed too high, the AI changes direction a lot and doesn't actually ride in (even a curvy one) a line. For the suvival of a rider I think Charge is more important. Variag Cataphracts have very high charge and last quite long even when they ride into a mass of infantry accidentally. (they usually ride through)


The real problem for me was that the orcs have high Agility (they have to be fast), but that also makes them very fast with their weapons. They basically outsmart any human swordsman. (well, elves and high tier gondorians are fast as hell too, but with the orcs I meant low tier troops) Also their weapons do some serious damage.
I nerfed the orcs down (via the old man) and it's more playable..
I'm still waiting what the RCM comes up with though.
Here's my gallery: http://merlkir.deviantart.com/

I'm now painting and drawing commissions. I'll paint portraits, pets, girlfriends, favourite characters..you name it. Send me a PM if interested ;)

DaBlade

  • Guest
Re: Needs Rebalancing?
« Reply #17 on: June 18, 2007, 02:58:18 AM »
I agree with charge being more important. That's why Rohan units do better after thy get armored horses; they are slower but have better charge.

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Needs Rebalancing?
« Reply #18 on: June 18, 2007, 03:04:02 AM »
The speed to maneuver ratios are what screws up the AI on horses.  Check out the numbers in ONR, and then observe the actual horses ... they don't look unrealistic.  Where the M&B AI has a problem is if you set the maneuver numbers substantially higher than the speed - then they go everywhere, like a wounded snake.  That's bad - you want to avoid that sort of thing.

A high charge value can be used to prevent a running horse from stopping every time it gets near a guy on the ground, yes.  Higher speed will also do it - charge damage also gets speed bonus.  You have to be sure not to increase both at the same time when testing these things, or your changes can multiply and come out way more than expected.

A lot of the orc weapons are going to come out slower in the RCM version.  In native, huge axes were only slightly slower than daggers.  ???  With that change, the speed of some of the orcs will be lost - they can't get reach, speed, and damage all in the same weapon.  That should help balance out the issue of low-end orcs being too strong/fast/whatever.

The RCM changes will be so numerous that's it's hard to predict how the final outcome will balance.  First, however, a lot of people who have only played Native-based models and have no sense of tactics will immediately run into a group of hostiles and get killed.  Then they will whine, like they think it's too hard and only kryptonite should hurt them.  Then somebody will finally build up a huge army and a lot of impressive equipment and then decide it it too easy.  This is the normal progression, and I'm expecting it.  Also, somebody will whine because their player is no longer the only character on the field who is capable of hurting someone.  That's also normal.

But in final assessment, I'm expecting good results... and so are a LOT of other people who have been asking for this since the first of the year.

Aethelred

  • Guest
Re: Needs Rebalancing?
« Reply #19 on: June 18, 2007, 03:51:28 AM »
Er, an honest statement of personal opinion based on my experience of how horses function in M&B? I do have some experience of horses, I've seen 'battle' horses charging in real life. Faster 'realistic' horses look strange in M&B, and the AI handles them (even more) oddly. Hence I don't feel that it's crucial to speed up the horses since it plays a bit odd and what you say is slow to me seems fast enough to be effective as a cavalryman. I'm not saying I'm against horses being faster, it just never occured to me as a problem. Just to clarify out of interest did you compare the top speed of horses in M&B to the top speed of a combat horse with a heavily equipped man on it? I'm not assuming you didn't, just want to know if there's really such a disparity.

Statistics aren't the most reliable way of defining what is lethal. You may quote that statistic about fatal stabbings - that doesn't necessarily mean the majority of stabbings of that depth are lethal. There's a reason why in the UK a knife blade under 3 inches in length is usually legal to carry around. I can also say that people can survive getting shot in the head and all sorts of other serious wounds. What the high weapon damage seemed to do for me was to exaggerate the statistical certainty of death. I also think the combat model in Mount & Blade plays better with lower weapon damage. High damage weapons just kill far too quickly and easily, since there are no other factors to take into account such as getting your weapon out of a corpse it's been stuck in, having that weapon break or the haft fall off your spear. In ONR you could just bash lmb and kill 1 person every second with some weapons. 1 little, 2 little, 3 little enemies...

I take your point about incapacitating wounds, but that only really works effectively with the player's party. Enemy parties don't have a surgery skill, so every 'incapacitating' wound is lethal, which you can hardly say is realistic. Hence why I say it's an effort at rebalancing rather than the creation of realism which the default M&B system doesn't allow (yet).

I'm aware it's not easy to penetrate armour. This is one of the reasons I found the weapons a bit overpowered in ONR - against medium armour they very little problem to cause 1 hit fatalities. And yet although armour isn't easy to penetrate, that shouldn't mean you can't be injured. My experience with ONR was that most enemies did 0 damage to me. Getting hit with a heavy weapon anywhere on your body will generally cause a concussive blow, as long as the armour fits closely to the body. Since you used chain as an example, chain mail doesn't give much protection against a heavy blow. Probably more often than not you would be incapacitated without your chain being breached. I'm always a bit suspicious of these weapon tests, most of the ones I've seen demonstrated in front of me or in documentaries are very suspect. Really the only effective test is to have a live (or dead) human put in armour and hit with whatever weapon you want to test. Anything else and it's mostly theory, which people tend to misapply. For example I've seen a program using a scientific test of a bodkin arrow, trying to prove it couldn't penetrate plate armour by dropping it at combat speed on a sheet of steel. The sheet of steel was placed on a solid metal block, hardly a conclusive test. I've also seen archers demonstrating that arrows can pass through steel by firing it at a piece of rather soft looking steel on a wooden board. That one is more convincing, but until someone fires a bodkin arrow from a full-power medieval longbow at a proper suit of armour that isn't empty, it's really not proven. Cutting tests are similarly dubious - the ones I've seen often involve hitting at blocks of soap which are supposed to represent human flesh, but without any bones or stuff involved. Alternatively they hack at bits of meat; that ought to be more convincing but it's usually only a small bit. Only time I've seen someone have a go at a bigger bit is here. Note it takes a fair few whacks even for a zweihander to go through just ribs, so I'm not so sure about cutting people in half.

I don't think the native M&B is quite as ridiculous and based on fantasy movie tomfoolery as you say, the armour is definately more than decorative. Certainly the weapons you're talking about like the giant axes are unrealistically balanced, but on the whole the armour isn't too bad. Slashing an enemy in armour isn't effective unless you have a big speed bonus, which reflects more or less what is realistic - it's not easy to penetrate medium armour or above, but with a good whack you can still hurt the person wearing it. But they definately exaggerated the ability to which you can penetrate armour with piercing weapons - swords can be stabbed through armour too easily. I do actually agree with most of what RCM promises, I just have some concerns that it might be a little overbalanced in the other direction.

By the way running into a group of hostiles in native will get you killed or at least seriously wounded before you can run away! Surely nobody does that and expects to live? And the 'building up a huge army with impressive equipment and finding it too easy' is what I found with ONR. But I don't really expect the same problems with TLD since it doesn't really have any super-heavy armours.

Anyway thanks for working on this, I hope it'll be a nice improvement. ;)

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Needs Rebalancing?
« Reply #20 on: June 18, 2007, 04:56:50 AM »
I hope it will be an improvement too.  However, my personal feeling is that it couldn't be much worse than the Native damage model.  Maybe I'm too negative.

The horse speeds can be confirmed with a simple test.  Average horse with rider (and saddle and whatever), gallop, about 30 MPH.  Distance running horses run about 40 mph, and short-sprint quarter-horses 50 MPH.  Average speed of a running man over rough terrain, 3 MPH.  Now, not wanting to over-rate or unbalance things, I did not model the run.  However, even a gallop should put a horse about 10 times faster than a man on foot.  Accounting for time to accelerate, and assuming any armor to slow them (based roughly off of what happens when you put two people on a horse), I figured that horse speeds ranging from about 11 to 22 would work out.  Having grown up around rodeo ponies, this still feels a little slow to me ... but since it has been years since I was on a horse, and not wanting to unbalance the game, I figured this was conservative.  That's still right at twice the speeds in Native.  You cannot slow a horse more than that by what he carries, without killing the horse... i.e. a horse can still make 30 MPH up to the point when the load he carries will cripple him, then he goes down like a sack of rocks.

As for the surgery skill and enemy parties - if they win, it does not matter.  If they do not win, I presume that you would probably not waste bandages on them.  (You would probably finish off any disabled ones.)  However, theoretically, the surgery skill could be applied to an AI group if they had a hero-type character in the group.

I have always been very careful about making my test cuts accurate.  Any material I use as a test cut is compared to test cuts on animals (dead deer mostly - I grew up hunting, and we processed our own meat ... they seem similar to human structure, and you have to split a few bones in the process of cutting them up anyway).  I try to confirm any findings by at least two or three tests under different conditions, to check the test.  I've seen way too many bad tests too.  Bullet tests are usually more deliberately slanted than blade tests, but either one can be skewed.  That's also why I do a lot of my own.

You mentioned maille - it is deliberately rated lower in the RCM mods than a similar weight of scale or lamellar, because of slightly weaker impact absorption.  I have tested some maille on my own body (deliberately letting somebody beat on me) to confirm exactly how much impact goes through it, compared to other designs.  (I was testing it for construction of modern knife-proof vests, but the tests are equally valid for ancient construction maille.)  Feel free to replicate the tests if you think them inaccurate.  Oddly, the reference to ONR is misplaced in this reference ... Japan's only use of chain armors was on arms or connecting larger sections - Japan never used a "coat of maille" as such.  Japan used hardened leathers which were fair against lighter weapons but weak against real blades, and they used heavy metal lamellars, with little in between.  TLD should not have that odd disparity... maille should make a nice mid-range between the more rigid metal armors and the non-metal ones.

What I mean about the armor being decorative and the weapons a joke is simple ... the weapons would have to be made of rubber for a human to take that many hits and keep fighting.  Then, armor provides less protection than a few extra "hit points" ... like, "because I am tough, you cannot kill me with that hatchet".  That is absurdly unfair.  I mean, who do you know that can take a hatchet in the chest and keep fighting?  The Native M&B numbers say that if you are average, you should be able to take an average of three or four blows to the chest from a hatchet and keep fighting.  There should be no way to bring a healthy man down with two hatchet wounds, and you might still be fighting after as many as six ... without any armor.  What were they drinking?

Realistically, some armor should help ... but without armor, any hatchet hit with any skill or strength behind it at all should have some chance of putting a man down.  (Most people would curl up and twitch after such a wound.)  Without armor, likely nobody could take more than two or three, even if some were not great hits, before shock from pain and bleeding puts them down.  Armor might take all the bite out of it, or only some, depending on the armor and the angle of impact.

How much more absurd can you get than that?  Wait, I know ... arrows that go through plate armor, but an unarmored man can take six of them (including one in the left eye) and keep fighting.  I've seen too many living things die from arrow hits, and standing after a single arrow hit is unlikely without some metal armor to slow it down.  Furthermore, arrows will go through some metal, but they also deflect more often than many weapons - the ones that seem to go through hardened plate armor are insane.  (I've tested arrows against various metals.)

So yeah, "absurd" is still my word for it.  Unless you think you can take three hatchet wounds to the chest and then fight....

Offline Ancientwanker

  • Master
  • *****
  • Posts: 1312
    • View Profile
Re: Needs Rebalancing?
« Reply #21 on: June 18, 2007, 05:40:43 AM »
There used to be a routine to increase the horse speed in tld but i didnt include it this time. It was just a pain to maintain.

Offline Llew

  • Moderator
  • *****
  • Posts: 1399
  • TLD dev team
    • View Profile
Re: Needs Rebalancing?
« Reply #22 on: June 18, 2007, 05:44:29 AM »
He's only there on day one? Oh feck, that sucks, what are you supposed to do if you start far away from the Brigand fort? He should really be in every town if he's going to be there only for 1 day. I hope the number of great hosts of Mordor is 4 by default, or my current game is wasted. ???

--thats a good point now that you can start anywhere, Ill think about that..

Dang LOL! Thats exactly what happened to me!  :-\

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: Needs Rebalancing?
« Reply #23 on: June 18, 2007, 06:00:12 AM »
There used to be a routine to increase the horse speed in tld but i didnt include it this time. It was just a pain to maintain.

Well, the RCM version, as soon as it's together, will certainly take care of that with a single set of good horse stats that are well tested in other mods.  Since I expect about 99% to use the RCM version, that should eliminate the question.

Speaking of which, AW, I hate to be impatient, but "as soon as it's together" is still waiting on "as soon as I get everything".  And a Great Host of fans are becoming increasingly impatient, so I'm starting to feel pressured here.

Offline Senta

  • Craftsman
  • **
  • Posts: 114
    • View Profile
Re: Needs Rebalancing?
« Reply #24 on: June 18, 2007, 08:02:01 AM »
to OP: i find it quite easy levelling my youths and esquires up. i just bunch them together and let them swing it out with any incoming enemies. plus them not being very tough is more than compensated by upper tier riders. Eorl and Brego guards can eat orcs alive in the ratio 5:1. provided the terrain is adequate.

Offline Ancientwanker

  • Master
  • *****
  • Posts: 1312
    • View Profile
TLD: Balance issues for 2.3.1
« Reply #25 on: June 18, 2007, 08:09:15 AM »
Thread for balance issues...

Offline Ron Losey

  • Master
  • *****
  • Posts: 4418
    • View Profile
Re: TLD: Balance issues for 2.3.1 (with Native combat model)
« Reply #26 on: June 18, 2007, 08:17:57 AM »
General note:

Optional RCM version will be out soon.  To avoid confusion, once it is out, balance issues specifically pertaining to that version should be moved to the "Optional RCM" thread.  This will make sure everyone is at least speaking the same language and/or playing the same game.

No need in creating more confusion.

jefz

  • Guest
Re: Needs Rebalancing?
« Reply #27 on: June 18, 2007, 11:12:43 AM »
Using the old man is still (probably) the best solution to fix the imbalance between orcs and men ~ just lower the orc's iron-skin & power-strike values.  It is simple and effective.

As for the RCM (thank god it's not standard with TLD), my problem with it is that it effectively removes lower-tiers as a threat, which is drastically unrealistic.  For instance, being swarmed by 5 brave peasants would be death to an unhorsed knight.  Sure, the knight will probably fatally wound 2 or 3 of 'em as they rush, but the others will trip or tackle him to the ground --> which is game over, as by that time they've started pouring sand into his helmet or gouging out his eyes, in the meantime 15 more peasants have arrived...  This is how peasants would damage and fight a knight:  get him off his horse and swarm/smother him.

The M&B combat-model just doesn't account for losing one's balance.  M&B addresses this combat inconsistency by allowing peasants to damage armored characters with their meager weapons.  By removing the ability of these peasants and lower-tier troops to inflict damage (as the RCM does), it makes the game like every other BS video-game where the quality of your gear is the only aspect that matters.  While the collective RCM heart is in the right place, the final product is much more surreal than real.

I agree that the horse speed is a little slow, but it's not that far off.  Remember, you're not riding on trails and roads in M&B.  If you've never ridden off-trail in the forest, you don't haul ass if you don't have clear footing (and can be knocked off the horse).  More important than speed, I think the biggest TLD issue with horses are the Porsches (err...wargs) that the orcs are riding.

Rollipeikko

  • Guest
Re: TLD: Balance issues for 2.3.1
« Reply #28 on: June 18, 2007, 12:47:53 PM »
And, because MaB:s system doesn't make difference damage-vice when hittin different spots(except log-ranged attacks), balancing combat is even more difficult, as hit to your toes deals the same amount of damage than what hit to your torso would.

Ebu Deyyus

  • Guest
Re: TLD: Balance issues for 2.3.1
« Reply #29 on: June 18, 2007, 12:51:05 PM »
for me, orc snagas are way too powerful, they get morningstars and some other very deadly weapons!(i am playing uruk hai, and currently fighting mordor by the way)