SPECIAL and Realtime

Even with the prospect of Fallout 3 still not being in development (or being in the very beginnings), i thought i'd create a thread discussing the reasons why SPECIAL wouldn't work very well under realtime. Honestly, i know very few instances where it causes inconveniences or problems; i do not know the full extent of the problems, but would like to bring this up and discuss it with people who know them, so we could enumerate a fair amount of them; maybe the thread can generate enough good points so it becomes a stickied thread, or a reference for Bethesda (this, assuming the developers manage to take some medication for their headaches and resume reading the site).

Let's start by assuming Fallout is transposed into a first-person, realtime model.

If you're playing a game as a first-person shooter, then what happens to a character's agility? The character isn't in control, the player is. And in this case, it's useless, because you have the player's agility controlling it. If you're playing a game as a first-person shooter, then what happens to a character's Perception? Again, it's useless, because you have the player's own perception determining wheter he hits things or no. When you play any other first-person shooter, like Quake, you hit things because you aim yourself, not because the character has an 'X' value in a skill that determines his chances to hit. And if you have this happening, what happens to those two attributes? They become useless. At that point, it should be noticeable that if it goes the FPS route, its no longer SPECIAL, it's SECIL.

If you go outside character attributes, then you also have skills. What happens to your weapon skills in an FPS? Weapon handling skills in an FPS are handled by the player; again, not by the character. Small Weapons, Heavy Weapons, Energy Weapons - all working based on the player's skill to effectively use weapons, specially with Perception out of the picture.

You could try to make it like Deus Ex. You could have the character need to upgrade his weapon skills to better use them... but there was a reason why they were removed, it's because they removed the immediacy of real time, first person combat. FPSers are usually about twitch reflexes. If you have a system ruled by players' skill (FPS), and you try to place barriers to the players' skill (ie, the need to increase skills to use things), then like i've stated above, it crumbles, because the system combines two very different systems.

Even if you simply place it in realtime but do not change the PoV, again, this causes problems. What happens to things like Sequence and Action Points? In realtime, there is no place for a sequence, because everything is happening simultaneously. And action points are basically gone as well. Combat also ends up being hectic and will risk requiring the addition of pausing. Which begs the question: if it requires pausing, then why not just include turns and give the players options to speed up the process trough elements like, say, speed sliders and concurrent turns?

Now, the above was taken from a recent convo between me and AlanC9 at the Obsidian boards. My point is very streamlined (honestly i have no patience to expand on the matter, given current circumstances and the fact that i've written too much about this in the past and at one point it gets tiresome), but i think presents a few of the problems with the change.

If anyone wants to present more points as to why SPECIAL should not go realtime, post them. Hell, if you think it should go, post those points as well. I'm interested in knowing the full extent of the problems, along with any theories on how it might work.

 

I don't think this is limited to SPECIAL, really.
Any deep stat system is broken by 'true' real time (to avoid all the stupid arguments about KOTOR, or NWN or BG or whatever Fuck the Monkey system a random poster pulls out of his ass, I mean real time in the sense of you press keys/mousebuttons, and the character acts with no rounds/turns/phases/invisible twinkly thinks/whatever to break up the action in any way), there simply isn't enough time to insert calculations to accurately reflect a character's skill.

And before anyone jumps in- Deus Ex was very shallow in terms of things that were laughingly called 'stats', and simply degraded weapon effects according to the level you were below the maximum skill.

Round based 'realtime' like the Bioware games, is a slightly different thing (it isn't actually real time, since it pads extra time and animations around the actual actions), but still distinct from turn based. This style could potential be made to work with SPECIAL, but personally, I don't think it would be as fun.

 

well, I obviously haven't played that troika vampire game yet, but the system they describe could obviously be adapted to the SPECIAL system as easily as it could any other simple mechanism set.

The obvious points, By adjusting internal weapon accuracy based on the charecter's stats, it is quite possible to make accuracy partially dependant on those stats (and skills), not wholly since the player still has to aim (assuming this is a FPS and not NWN).

Agility determines movement speed, accuracy and dodging in SPECIAL, all of these things can be altered in an FPS as well by giving higher movement speeds to charecters with higher AG, this would also in and of itself increase dodging ability in an FPS, but the game could also include a mechanism for a charecter dodging if a bullet "hits" the charecter and that could also be made dependant on AG.

Action points are simply an absraction to indicate how quickly the charecter's actions occur and hence how many they can perform in an arbitary ammount of time, in an FPS the rate at which these actions occur (how much time each one takes) can be adjusted directly. Again though, it's less direct because the speed of the player is also a factor.

For my own part, I theought the combat of the first two fallout games was an utter waste of time, I could only command one charecter which removed any trace of tactical thinking and the fights could take all evening. If they insist on not giving me any more units to command (ie actually letting me control the NPC's), then I hope they can at least speed the combat up so I don't spend 10 min watching mutants trudge accross the map before I take another two called shots.

 

A comparison to Deus Ex is valid here. At low skill levels, bullets went all over the place and went straight more or less at high levels. In theory this sounds good, but it completely discouraged the use of ranged weapons unless I could find a spot to aim and try to get a head shot. That brings up called shots. I can't think of a single way that that could be implemented in a real time system if it isn't due to the player's own skill.

Increasing movement speed based on agilitiy just makes it even more twitch based. Dodging anything other than unarmed attacks would be silly.

 

Just make bigger hit boxes for higher skill. But still, it's a twitch game... brr.

 

Damn, this is a long post. I'll try to organize it a bit.

1. Introduction: in which our hero sallies forth

The only area where real-time would make a particular difference is combat, of course, since that's the only area that was TB in the first two FO games. But you're also bringing first-person view into it, so I guess you're asking two questions: How would FPV affect SPECIAL outside of combat, and how would FPV and/or real-time affect SPECIAL in combat?

First question first: I don't think SPECIAL would suffer due to an FPV outside of combat. I'd probably be frustrated with it for other game design reasons, but the world interaction allowed and structured by SPECIAL wouldn't suffer at all from FPV outside of combat. Character interaction might even be enhanced, offering a more natural direct character interface than the talking heads of the first two FO 1 and 2.

Now the next two questions: I think SPECIAL could also survive FPV in combat, if combat remained TB. I've spent a week now trying to imagine an effective TB/FPV combat model, and I'm still not convinced it wouldn't be a horrible train-wreck of a game system. I am convinced, though, that it would be possible to create a clunky, irritating TB/FPV combat model in which SPECIAL would work just fine, albeit with a total lack of elegance.

Real-time suddenly changes the picture, though. SPECIAL becomes untenable in real-time.

2. Character speed and reflexes

First, there's the problem of AG and PE reflecting character reflexes. There are some solutions possible - e.g., conditioning gun accuracy, speed, line-of-sight, and damage on stats - but the only way to avoid making combat effectiveness depend mostly on the player's (real-life) reflexes would be to slow the game down so much that it would be more like "quarter-time" than real-time.

And that's maximum manageable combat speed, mind you; to reflect the speed (AP) differential between a character with AG 1 and a character with AG 10 would require slowing the AG 1 character down to a snail's pace. The obvious alternative would be to slow everyone else down for higher-AG characters, but that would be like playing Max Payne with bullet-time on all the goddamn time. In either case, I'd shoot myself in the face before playing such a game, and I think most videogame consumers would share that high-minded principle.

Real-time, then, would invalidate low-AG characters, which already violates the fundamental openness of SPECIAL and the Fallout setting.

3. Movement potential

Other problems are insoluble in real-time no matter the speed. AG, for example, allows a character to shoot things faster and move faster (i.e., farther). With a real-time FPV, however, it's highly difficult to judge the advantage, or even the possibility, of different movement paths before you've chosen them. Even looking around costs valuable time in which your enemies are simultaneously advancing upon you or shooting you dead.

4. Firearms accuracy

PE affects your firearms accuracy, but it's impossible to model that in real-time without (quite literally) introducing some serious clumsiness. As others have pointed out, Deus Ex modelled accuracy by making the crosshairs unpredictably shaky at lower skill levels, but this is a deeply unsatisfying solution for anyone who plays FPS games. (At this point, I'll admit to being a regular FPS player; I'm even in an active UT2K4 clan.) It essentially requires nerfing the traits of one game genre in deference to the traits of another game genre which you've already nerfed. I mean, who comes up with this shit?! My primary motivation to increase my gun skills in DX was to make the game play like it was supposed to, goddamit.

5. Firearms range

PE, and firearms skills in SPECIAL, and guns in the FO world, also involve range limitations. These were effectively nonexistent in DX, and they're functionally absent in MW as well - by which I mean that my bow can fire much farther than my line of sight extends even with the maximum draw distance. Never mind maximum range limitations, though; my concern at the moment is with accuracy changes throughout a weapon's range (which also seem to be absent from both DX and MW).

Assuming you're going to commit the game design sin of implementing the "shaky crosshairs" described above, how do you model a high level of accuracy for close targets at the same time as a very low level of accuracy for distant targets? "Easy," you respond: "The crosshairs remain steady when trained on the close target, but go haywire when moved to the distant target." Way to go, champ - except for the fact that the character model, the hitbox, all visual indicators are already smaller for the distant target, so modelling inaccuracy on a distant target would require the crosshairs to actually be less shaky than for a comparably inaccurate close target. The visual difference (in quanta of "shakiness," if you will) between an accurate close target and an inaccurate distant target is then rather small, and therefore difficult to track for the average player. You could always increase the hitbox of the inaccurate distant target in order to allow a more visually-apparent shake in the crosshair, but then we're getting into the kind of "epicycles on epicycles" excreta that made the Ptolemaic solar system such a goddamn mess before Copernicus cleaned house.

Still with me?

Another solution would be to provide floating stats (percentiles, e.g.) or textual indicators of accuracy. Those would provide the needed information, but they'd be difficult for players to process in a real-time environment, especially when playing an AG 10 character moving at rapid speed - precisely the character who should be most able to handle rapid input on the battlefield.

Visual cues are another possibility, of course; many games now use different crosshairs (broad vs. tight, e.g.) to denote expected accuracy. The problem with such approaches is that they provide little granularity. We need a system which can clearly reflect the difference between a PE 1/10% small guns character and a PE 10/150% small guns character, and changing the crosshairs (or any other similar, easy-to-grasp visual indicator) might provide four or five levels of feedback at most.

The lowest-common-denominator alternative is to give no visual indications whatsoever. Morrowind's modelling of archery is a bad example - or, rather, it's a great example of this terrible approach to combat design. I played an archer in MW, but I never really got a sense, when increasing abilities or skills, that my higher stats actually translated to more effective skills in the game world. I was never able to judge whether an arrow would hit or miss, until I hit level 15 or so and my arrows always hit (at which point raising my archery skill only increased damage). The level of feedback was phenomenally poor, and my interest in my character was thereby lessened. Note to designers: that really, truly doesn't help IMMERSHUN.

To recap: Effectively providing visual indicators of accuracy changes throughout weapon range is impossible in FPV, without floating stats or textual indicators which undermine the effectiveness of FPV - but removing all feedback is probably even worse.

6. A lyrical interlude on isometry

(Note to self on reason #73 for making FO3 isometric: it actually allows the modelling of real weaponry, including those with extreme ranges, which become totally impossible in FPV [due to technological constraints on view distance].)

7. My temporary conclusion

I tend to think that SPECIAL could work with an FPV outside of combat. I expect I'd be disappointed with it, at least at first (and especially with NPCs in tow), but it could work. In combat, though, TB - or at least a quasi-TB, phase-like system like BG's - is absolutely essential for SPECIAL to have any relevance in the setting.

I've mentioned only a handful of the problems we can probably imagine with a real-time context for SPECIAL. The more I think about it, the more I'm totally convinced that real-time combat would gut the SPECIAL character system.

 

Are you perhaps a reader of Mr. Steven Brust's Viscount of Adrilankha series?

With regards to the point on the targeting hitbox increasing in size with skill accuracy, doesn't this seem rather counter-intuitive from a presentation perspective? It would seem to me that the better my character is shooting at range the smaller the hitbox would get, simply because it would indicate my increased capacity for aimed shots. If the hitbox gets larger as your accuracy increases how can you make 'called shots' Fallout-style whent he box covers a greater area of your target? The other thing that bothers me about this idea is that I've never liked the phony way in which some games let you shoot off target, in other words shoot outside the perimeter of the target, yet still hit it because your 'character skill' is somehow correcting your poor aiming with the crosshairs. I am no FPS sharpshooter by any stretch, though I enjoy playing the games, but a cheap victory to me is no victory at all. So my question is, how would that situation be resolved in a realistic manner?

As an aside, I think first-person combat and real-time are inherently incompatible with SPECIAL and therefore completely foreign to Fallout under any appropriate circumstances. Some franchises are suited for a variety of genre games because the setting is generic enough and structurally open enough to allow such design crossovers, but Fallout, I believe, is not one of them.

-Reklar
(a Fallout/RPG fan)

 

I'm worried about lock picking and traps skills due to how horribly they were handled in Morrowind. Adding depletable tool sets for skills makes no sense if your at a high level for said skill. If your skilled with those tools they shouldn't break after 5 uses. Now that i think about, this post really has nothing to do with the subject at hand.

 

Okay, for the last time. There was a combat speed slider. In both games IIRC. If you couldn't be bothered to try out the options menu or just RTFM, it's really no one's fault but your own.

 

I played around with the speed slider in both games but I still wound up reading a novel during most of the combat (hubbard's Final Blackout make a nice complement).

 

A sudden feeling of deja vu.

Yeah, true RT is unworkable. I think a phased system could be made to function.

This could be made workable by building slack into the combat rounds via animation (KotOR/NWN). Faster characters have less slack.This probably isn't viable in first-person, but it would work in OTS. You'd still probably have to cut some of the extreme cases -- what's the rate of fire ratio between a low AG and a high AG+perks character? 6-1?

The problem with this argument is that it isn't about implementing the movement rate itself, it's about implementing TB decision-making. Not really on-topic.

But accuracy doesn't have to be based on your skill. It could work like KotOR, or an IE game. This model only breaks down if the game really can't tell who you were trying to shoot. I don't see this as insurmountable; friendly-fire incidents are a Fallout tradition, no?

 

I'm not sure what you're suggesting. I can imagine a system wherein the targeting reticle automatically snaps to the proper area on a target's body if your skills are high enough, but that's the only way I can interpret your statement for a real-time, FPV combat system. KotOR and IE games aren't valid comparisons because they didn't feature FPV, nor did they allow called shots (well, other than the D&D "Called Shots" - I can't even recall if those were in BG2).

I get the feeling we're talking past each other on this one, but hey, it's an internet tradition, no?

 

Possible solutions

Agility = How slow or lagged the controls are. At worst, the controls would have about a half second delay. Enough to piss you off, but make the game playable.

Perception. Major fogging. at 5 no fogging. above that, minor details, like a few bullets on the floor, become highlighted.

Accuracy, Targetting reticule like rogue spear.

 

A low-AG character should be viable throughout the game. Ain't nobody gonna sit through laggy controls for the entire game, even if it's only confined to combat, so I'm doubting the appropriateness of that solution. It also doesn't address the fact that high-AG characters should be able to move more quickly - not simply that their reflexes should be faster, but their speed should be significantly faster than a low-AG character (who, again, should nevertheless be a viable and satisfying character choice). They should be able to perform more actions in a set duration of time, even if that action is movement, and laggy controls don't have anything to say about that.

Your suggestion for PE makes a little more sense, partially because you calibrate the baseline at 5 rather than 1. (Your AG suggestion, btw, couldn't be calibrated at 5, so there goes that solution.) Are you imagining this fogging only in combat, though? It doesn't seem to make sense that you'd be super-nearsighted in combat but would have 20/20 vision the rest of the time - nor is it likely that anyone would willingly play 60 hours of a game which inflicted 20/200 vision on you throughout your playing. And don't forget that MW sports one of the most CPU-dependent graphics engines I've ever seen; if FO3 uses a FPV and that engine is similar, what about the large number of players who will have to enable significant fog simply to get playable framerates?

Help me out with the reticle (not "reticule"; yeah, I'm a spelling fascist) for Rogue Spear. What's the system there?

Here's a new point to think about: one of the fundamental questions in this discussion is the type of feedback you want to offer your players as they interact with your gameworld. Should it be primarily statistical/textual or primarily visual/iconic?

Most CRPGs rely heavily on stat- or text-based feedback (enemies are "Wounded," you have an 85% chance of hitting your target, etc.), but BethSoft's recent games rely much more heavily on visual feedback (when they provide feedback at all). An integral part of SPECIAL is a broad array of skills that, frankly, makes MW's character system look kinda grade-school, and that begs the question: do you really want to be forced into relying primarily on visual feedback for all of those SPECIAL skills, or might it make more sense to stick with stat-based feedback (which we already know goes very nicely with SPECIAL)? If you realize that yes, after all, you really should stick with primarily textual or statistical feedback, then do you believe the game is still a natural fit for real-time and/or FPV?

 

on the agility thing, jumping, climbing, going prone and back would all happen quicker. Maybe give you a bit less damage at higher levels. Why should playing a cripple be fun? its not steven hawking's deer hunter, its fallout. the weak are killed and eaten.

Same goes for perception. at 1, the player should practically be blind. 5 is average. 10 is godly. 1 should be blind. and i did mean all aspects of the game.

the reticle. well, depending on your and choice of weapon, your reticle would be a circle, or a few points. the more you shot if your accuracy was low, the bigger the circle would get, until you regained control.

 

If I was designing an FPS special in five minutes:

Strength is how much damage a melee attack does and carrying capacity.
Perception is accuracy ranged and viewing distance.
Endurance is how much damage you can sustain and how much fatigue you have for sprinting/jumping.
Charisma isn't a major combat attribute.
Intelligence isn't a major combat attribute.
Agility is top speed and rate of acceleration.
Luck is Luck.

There are FPS games out there already where you can choose between strong and slow heavy weapon classes and agile and wimpy scout classes.

 

Indeed. And a venerable one.

I was thinking of handling called shots more the way an NWN special attack is handled. Bind a particular key to each type of called shot, use that key to attempt that particular called shot.

This would be rougher in OTS than iso, since you're also using the WASD keys for movement.

 

The problem is that we're not talking about "an FPS special." We're talking about Fallout, and the venerable SPECIAL system, and all of the stuff you proceeded to describe has already been defined in several games and a whole lot of design documents. This topic isn't about gutting SPECIAL; it's about whether SPECIAL can work in real-time and/or FPV.

Have you even played the game being discussed by this thread, Quigs?

Part of the point of the whole FO setting is the open-ended character development and usefulness. You should be able to play a diplomat (with high IN and CH, low AG and PE) as well as a stealthy sniper (with high AG and PE, perhaps rock-bottom CH). Your thoughts seem to be stumbling towards a combat-heavy Morrowind clone.

We might disagree on these points, but I think they're kinda the assumptions in this thread.

 

yeah, ive played fallout. beaten em more times then I can remember. the game never made enough checks though. no checks for above 9, and the only thing i can remember for a low stat is that you cant speak well. (other then the obvious modifiers, like low perception and strength penalties)

1 is the lowest possible stat you can pick, right? so a 1 in perception terms would be a deaf blind man, with no sense of smell, touch, or taste, and can only grasp the obvious.

1 in endurance, and your that guy who bit his own tongue off in the movie VII

so yes, its *possible* to play such an example, it shouldnt be just as easy to play a moderately good character then it is to play some strange blink monk.

 

The thing is , Quigs, hardcore FO fans want characters to be viable even if making those characters viable isn't strictly realistic. Some characters should be harder to play than others, of course -- that couldn't be avoided even if you wanted to. But it should be possible to play and win with just about anything.

 

Has anyone here played Farcry ?
Well it involves some sort of realism like constitution meter for running, so you can't run unlimitedly and you need to take time to aim more carefully also crouching gives better aim too.

Those kind of things could be controlled with some of SPECIAL stats.
Like mentioned above it would be similar to Deus Ex.

Perception would give faster targetting speed.
Endurance and agility could affect to speed and running.

Quite like Snuffles already told.

 

You could do that in an FPS with ease, I assumed that everyone in this thread agreed that the non combat side of the game would be as easily creatable in a realtime FPS as it is in third person, are you saying that you think realtime isometric or fps combat modes will magically effect the non combat side of things?

 

I'm with you there, Snuffles.

So, I've been thinking, and I thought of Diablo. See, that game was tile based and real time, and with every move and attack animation taking a certain number of frames, it could as well have been turnbased by simply pausing the game every time the PC had completed their last action. The main difference between SPECIAL time and Diablo time is of course that walking generally took longer time than attacking in Diablo, where in SPECIAL attacking can take as little as twice the time to walk, and as much as seven times, and then I'm not even counting the Bonus Move perk which tilts the ratio in favour of moving even further.

In Diablo both an attack and a move was first initiated, and then the creature got the effect. Swing your sword faster than the mage can teleport away, and you catch him before it happens. In MechForce shooting 'mechs was pretty much an instant affair (possibly with a 1 second delay), and then the weapon would need some time before it could be fired again. Moving or attacking with limbs (punching and kicking) took some time, and the actual attack or move wasn't completed until the time had passed. At least that's the way I remember it.

For SPECIAL in realtime I'd suggest instant effects for attacks and moving, and then a delay depending on the time cost of the action. This is to prevent targets from moving out of reach from melee attacks. You don't want to spend the equivalence of 4 APs on a kick, only to have your target flee 4 hexes while you're waving your foot in the air. Sure, if you have the equivalence of 12 APs, and your target only 6 APs, you'd move at twice the speed and the target would only manage to flee 2 hexes while you're kicking, but even 1 hex would be enough. You'd basically have to break one of their legs to be able to catch up to them and attack them on the run if your attacks are delayed this way. Or chase them down and surround them with your NPC buddies.

There's two problems with this though. Sticking to the current SPECIAL rules, someone could attack 6 times (or even 8 with two doses of Jet) as often as a slowpoke with 6 APs shooting aimed shots with a rifle. So the real time experience would have to be slowed down so that such a speedy fighter could manage to click fast enough. Perhaps a slider could be used to either slow down combat or speed it up, depending on how fast the player wants to click.

There's also the issue of melee, where targets frequently become knocked down and the attacker needs to pursue in order to inflict more damage. I'm thinking that SPECIAL real time would need a few adjustable automatation options, such as "keep attacking the target last clicked on" and "follow targets if needed"/"attack nearby targets before automatically following a target" and "automatically reload weapon when needed".

The second problem is that at least kicking would look damned funny if the kick is instant, but then the kicker has to recover somehow. To a lesser degree this could be said for any melee attack, while ranged attacks could be shown as recoil and readjusting some setting on the weapon during the delay after the actual attack. Bursts can be shown in all their glory, as the shooter keeps firing bullets for as long as it takes to finish the action. Might look a bit silly if their burst only contains a few bullets though, while they by realistic standards might spend enough time to empty a clip and then some. Well, it seems to work fine in the movies so what the hell. A somewhat lesser problem would also be that they'd be shooting for a while after everyone of their targets has already taken the damage. Making the situation something like: "Oh, the super mutant is already spraying bullets? Then it'll be safe to run in front of him!" Of course, with this system the next burst will be seamless and whoever is in front of it when it begins could get hurt.

Given a system like this, with turnbased approached from real time rather than real time approached from turn based (like the Infinity and Aurora engines), I think that we could have both options, and have both of them match the rules exactly. It'll be short turns though, which might call for even more adjustable automatation options. "Pause when my target is dead", "pause when my target has moved/been knocked away", "pause when my clip is empty", "pause when I have taken X amount of damage", "pause when I have reached my destination" etc. With enough settings anyone could judge for themselves and their current character/situation how much real time they want. Or have I overlooked something that chucks a wrench into this approach?

EDIT: Been thinking some more, and I suppose that the SPECIAL system would have to be tweaked a bit to create a decent real time experience. Having one move take half a second seems about right, but with the current system that would make the attacks take as much as 3.5 seconds or more (I think some advanced kicks and punches could take 8 APs, 9 APs if aimed), and the fastest attacks would still take 1 second (which is an eternity - constant bullet time, I think someone mentioned that). This might be far too long to feel enjoyable, and a slider for combat speed would just make the moving opponents move too damn quick (Arcanum!). Although it might look pretty cool if a martial artist kicks immediately and then take their time to lower their leg. In fact...it would look damn cool. At least the first few times. After a short while such incessant posing might become annoying, and the crowd will start chanting: "My guy shouldn't stand there with his leg up in the air while someone is blowing chunks out of him with a shotgun!". While Fallout isn't an action game like Diablo, it probably wouldn't hurt sales if the action crowd could be attracted to the game. I mean, we're already stuck with the psychos who use the Bloody Mess perk, right?

If 3-4 AP attacks take about as long time as moving and moving takes 0.5 seconds, 6-8 AP attacks would take at most 1 second (for sniper rifles, bursts and powerful kicks). I think that would be close enough to Diablo's speed to be considered playable, especially with a slider to control the speed. Unfortunately this makes ranged weapons slightly more powerful, since they will have an increased rate of fire compared to moving (melee?) opponents. Perhaps to compensate a moving target for getting shot at twice as much, all shots fired against a moving target could have an automatic miss chance? Might be annoying for a shooter with 95%+ chance to hit who would under the current system rarely miss, but I wouldn't mind seeing movement be useful for avoiding getting hit by bullets. The exact miss chance should probably depend on distance, but end up at 30-60% or so. People getting shot at should then move toward cover, and only stand still to return fire once behind that cover. Yeah, cover...does the current Fallout engine even take cover into consideration?

 

Infinity and Aurora engines handle that with placeholder animations until the real attack annimation goes off. I've heard that Odyssey just breaks the connection between the attack animation and the attack routines altogether, but I haven't verified that myself.

 

Well, I consider both the Infinity and the Aurora engines inferior to Diablo when it comes to satisfying action, and the disconnection between the animations and the actual events is probably the reason. But I suppose that if attacking is going to be slow, someone using melee attacks that cost a lot of APs might do some fancy animation as they prepare for another attack. It's just that I think that waiting for 3 seconds after a 6 AP melee attack is going to become awfully boring. But then again, players who become easily bored might always choose to get as many APs as possible, and thus they might not have to wait as long. Does anyone know how many seconds a SPECIAL combat round is? Just divide by the character's APs and multiply by attack cost to see how long an attack animation would have to last.

I think it could look good though. A fast and powerful melee attack should be almost instant for sake of kinetic energy. Fast punch, slow withdrawal. By freezing the punch (or kick or stab) for a short moment (half the time, max one second) it'll probably look like the character lashed out real fast, and that way the player can see reasonably clearly what happened. Swings could use speed lines to show from where the weapon was swung, and then the character can withdraw their weapon to the ready-to-swing position.

In Arcanum it didn't look too good when someone with a pistol would raise it to the sky after every shot, presumably due to recoil. The pistol would spend more time in the sky than aimed at the target, and the muzzle flares would never actually be aligned to the muzzle, at least as far as my naked eye could see. So freezing the animation seems to be a good remedy for that.