SPECIAL and Realtime

So suibhne, are you going to answer my questions?

-Reklar
(a Fallout/RPG fan)

 

Reklar, when you say hitbox, do you really mean hitbox? When you say hitbox I think you mean a cube where if you shoot at it, you hit, so the bigger the box, the easier it is to hit, the hitbox should really not change size, it should stay the same size as the model you are aiming at. Do you mean the crosshair? The better you are, the smaller it gets, with your shots always landing inside the crosshair, so if your crosshair is large, then if you put the centre of it over a hitbox, the hitbox will be smaller than it, giving you a chance to miss?

 

I'm not sure if I am thinking of the same thing as you or not. In all the games I can recall playing, except flight simulators, the crosshair is either always inside a hitbox or there is no hitbox. What I am wondering is how you are increasing your accuracy making 'called-shots' if your targeted area increases in size the higher your skill gets? The larger that hitbox gets the more surface area of your target is included and the less realistic it is, to me, that you are aiming for a specific point on the target.

Is that more confusing or does it clarify my question a little?

-Reklar
(a Fallout/RPG fan)

 

Having read parts of this discussion here, on the NMA forums, and at Bethesda I'll add to the unpopular view that S.P.E.C.I.A.L. could be incorporated into a First Person game.

Notice I didnt say First Person Shooter. While you do shoot in Fallout the game isnt a shooter and the perspective adds to the immediacy of the feeling.

Dont get me wrong, I think that FO3 should be TB + ISO + S.P.E.C.I.A.L. but if it has to be FP, S.P.E.C.I.A.L. could work.

Usually games like this use twitch collision detection to determine hits. Why not pull it back to the RPG roots and use PnP dice rolls?

Divide the zones into micro zones that had their own descriptions of lighting and cover.
Use the players agility and luck to further determine the to-hit. Add the armor after the hit.

Albeit for "called shots" you would need some sort of VERY quick to use GUI that should be jog-dialed or scrolled through with one finger. Intelligence could determine the number of skills or called shots you could learn perhaps.

But all of this is getting away from what FO really is. My opinions are based on Bethesda's words "We're not going to get away from what we do best and make a top-down Balder's Gate type of game."

Gotta make nuka-cola out of lemons I suppose.

 

Well, Pete Hines explained that bit.

 

Then let's hope we get ISO!! Perhaps rendered in 3d to appease the newcomers.

 

Sure, Reklar - to note two things:

1. I think we're on the same side of the conversation, maybe.

2. You're not using the word "hitbox" correctly. It generally refers to the invisible boundary designated by the devs, within which any shot is actually a hit. Expanding the hitbox, then, has nothing to do with changing visual feedback; rather, it means your chances of scoring a hit get better. Make sense? Never mind the problems I suggested above, though: since this fundamentally relies on player (vs. character) reflexes, I still think it's a terrible idea.

 

Actually, yes. If your low-PE character has terrible vision during combat or your low-AG character experiences laggy controls during combat - both of which have been suggested by some of you as a way to represent the range of SPECIAL attributes during combat - why should these things be different outside of combat when the representational mode (real-time, first-person perspective) is otherwise identical?

I guess the game manual could helpfully explain, "The SPECIAL character system encourages character flexibility and multiple approaches to in-game problems. We knew you'd be irritated to play a nearsighted, clumsy oaf throughout your first-person sojourn in the wasteland, however, so we ensured that any clumsy, nearsighted character will exhibit those negative qualities only during combat. Have fun!"

 

Thanks for replying! I'm pretty sure we are as well, I simply wanted clarification on what you were suggesting, probably mostly because of terminology rather than semantics.

Ahh, that's the key bit of information I was missing, so yes, it makes sense now. I always thought a 'hitbox' was that little box surrounding the crosshairs, so when you suggested increasing that in size I thought it would be looney. You are correct about the player reflexes and I agree the concept is flawed. Let us hope this doesn't become something we have to worry about with Fallout 3. :shock:

-Reklar
(a Fallout/RPG fan)

 

This brings me to yet another reason why SPECIAL won't work in real-time. Yippee.

There's the basic principle that character skills, in a role-playing game, should be much more important than player reflexes. Fine - particularly if you happen to agree. But there's also an inherent problem, which we haven't explicitly mentioned, with a system dependent on both player reflexes and character skill (like any real-time implementation of SPECIAL, whether in first-person or not): player reflexes improve at different rates, and sometimes not at all.

Revisiting the changes suggested by Quigs and others, to implement a low-PE, low-AG character in real-time FPV, let's remember that those changes would all interfere with the interface between players and characters - by imposing latency between player reflexes and character reflexes, for example, or by limiting the visual input the player receives through the character's perspective. A necessary consequence, however, is that a low-PE, low-AG character which challenges player reflexes will be easier to play for players with better reflexes.

Take me as an example. My first choice is RPGs, but I'm a pretty good FPS player; I always launch FPS games on the harder setting, I'm in an active UT2K4 clan, etc., etc. For me, the crazy targeting reticle which represented low skill in Deus Ex was an irritant, but I still got headshots fairly consistently even in the second level of the game. Friends of mine, on the other hand, found it absolutely maddening, and had a fundamentally different experience from mine even though we built exactly the same character. The difference was entirely in our mousing reflexes, which then relocated the character's lack of skill outside the actual game; I was able to overcome the character's low skill level much more effectively than my friends, and I therefore had a much easier time in the lower levels of the game.

The point is that any system relying even partially on player reflexes is dramatically imbalancing, simply because players come to the table with an incredibly broad range of reflexes - based on innate ability, overall health, previous gaming experience, or even basic patience. As I suggested in a previous post, one way of mitigating this problem would be to simplify combat to such a degree that player reflexes are barely required in any case - like the total snoozefest of Morrowind's combat system - but that's a bonehead solution which alienates both the RPGers and the twitch crowd.

 

It might work if everyone moves slowly enough for reflexes/steadiness to not be a problem, and also if a target is highlighted with glowing red outlines when it's targetted by the player, and a percentage for the chance to hit is shown. But then it would probably not be a point to it being first person in the first place.

 

That would be my conclusion based upon the discussion so far, but perhaps someone has an idea how RT can be implemented with SPECIAL. I have to admit though, I cannot currently think of any means by which it could be done, considering that RT is inherently suited to action-y style games and player reflexes. I can understand some peoples' point of view to the contrary, but it still does not sound practical or even desirable to design such a system.

-Reklar
(a Fallout/RPG fan)

 

However: if we're sticking to Quigs' suggestion of forcing some interface latency for low-AG characters, which was one of my examples, then lowering the speed of the game would undermine that system by partly re-equalizing low-AG and high-AG characters. So you design an imbalanced system which reacts differently to players of various skill levels with the mouse and therefore does a poor job of actually representing the character range offered by SPECIAL - then decide, in order to mitigate that dependence on player reflexes, to slow down the overall game speed so that the system can't represent the character range offered by SPECIAL. As I pointed out awhile back, at normal game speeds the combat would be heavily dependent on player reflexes, while at slow speeds the significant delta between low-AG and high-AG characters would be unrepresentable.

@Surlent: Yes, I've played Far Cry - twice, actually. With that reference, you've suggested a way of handling Endurance, but I think that's the easiest physical stat to implement anyway. Far Cry didn't give me any ideas about how to implement PE and AG in a real-time combat engine using SPECIAL.

 

Uh, agility could still decide how fast people move and attack, and what base chances they get in skills based on agility. But perhaps that isn't enough of a first person shooter for your discussion (or argument)?

I did mention how it kind of defeats the purpose of making it first person if you slow down things to unrealistically slow speeds and also outline targets with red lines, but that'd be one way to give it almost the same premise as the TB system used in Fallout. Only the called shots and the successful targetting of targets behind other targets seem to suffer due to time constraints and precision.

 

It boils down to battlemath.

Obviously Strength will translate into melee damage...
Perception may decide whether you hit a bomb-tripwire
Endurance whether your first-person character can walk, run, or move at all
Agility your Armor Class, how much damage you can take

It really wouldnt be that hard.

 

Reminds me of...morrowind! ah, constant clicking, how I missed thee.

 

It really wouldn't be that SPECIAL, either.

I'm not blindly orthodox wrt SPECIAL, but I'm not sure how anyone could suggest the above scheme without recognizing that it would gut the SPECIAL system. Then again, maybe DarkSign is suggesting that fundamentally altering SPECIAL is the only way it can work in real-time - in which case we're in perfect agreement.

Part of the, ah, challenge in this thread is that folks like DarkSign either don't know anything about how SPECIAL works, or don't care about preserving it in any recognizable form. The original question raised by RolePlayer wasn't whether SPECIAL should be used in FO3 - it's whether SPECIAL is even possible in real-time.

I've raised this objection several times and nobody's answered it: I don't see how you could represent the significant delta between low-AG and high-AG characters without making high-AG characters almost unplayably fast - which would make such characters much more dependent on player skill/reflexes, ironically meaning that investing in high AG would actually make the character's skill relatively less important (i.e., a character who's supposed to be highly agile would have more of its agility displaced on player skill than a character who's less agile). As we agreed, you could somewhat mitigate this dependence on player reflexes by dramatically lowering game speed, but I think the result would be low-AG characters who were maddeningly slow.

My previous question, in a response to Snuffles, also remains unanswered: "If your low-PE character has terrible vision during combat or your low-AG character experiences laggy controls during combat - both of which have been suggested by some of you as a way to represent the range of SPECIAL attributes during combat - why should these things be different outside of combat when the representational mode (real-time, first-person perspective) is otherwise identical?"

The folks who think that real-time SPECIAL "really wouldn't be that hard" have so far displayed a remarkable lack of interest in grappling with these very real implementation problems (or showing why they aren't problems at all!).

 

I suppose I have to ask about the thread's assumptions then.

If our thead definition of real-time excludes things like NWN and KOTOR (leaving only FPS-like things like deus ex) then as is obvious, you cannot have an FPS that has no dependence on player reflexes, that doesn't mean the stats can't be an overwhelming factor, the alien and the human played very differently in AVP2.

Now, if the definition does include things like NWN, there's no reason that all of the old fallout mechanisms couldn't be maintained if someone wished to, what interests me is why anyone would bother, the special system failed to produce worthwhile combat in FO or FO2, if there is only going to be one PC and hence tactics are impossible, why not use something like a deus-ex system that gives the player something to do even if he doesn't have any tactical decisions to make?

 

Quick note: the intention was basically analyzing why it should, or should not, go into a realtime combat model. I merely hinted on the problems i saw with a realtime, first-person mode, and briefly passed by simple realtime with a (possible) isometric PoV. The discussion basically envolves the pros and cons of any realtime system, under multiple PoVs.

 

Well, you could keep player speed constant and speed everything else up for low AG characters and slow it down for high AG characters. The player could also be in charge of the slider that determines how fast everything is moving, thus they could choose to slow things down much more if they come upon an opponent that is too fast to be targetted, which may well happen for a low AG character coming across a high AG enemy.

The present system doesn't take PE into consideration when judging how far a character can see, so I don't think a first person view would have to either? Just have PE affect distance modifiers for ranged weapons, like it's always done.

For people with good perception the game could perhaps offer an option to zoom when trying to aim at a target far away. The game could have a quick zoom key which would check the distance to the target closest to the mouse pointer, and then zoom in to that distance almost immediately (or immediately if that's not too jarring). Lower PE characters might not be able to zoom the full distance, and might end up at their maximum zoom instead (which might not be enough to give a good view or a sizable target area to aim at). Or everyone could zoom equally, if PE is really not going to help actual viewing and aiming. Perhaps such a zoom assistance could even replace the usual effect PE has on the chance to hit, depending on how much of a twitch game you want it to be.

But I think this is one of the major problems, the fact that targets far away would be smaller, and the fact that targets who are mostly behind other targets would become difficult to target properly. This last thing is something of a problem even in the isometric view, but at least a player can take their time aiming for that little part that is still visble. Having targets which are being aimed at glow with red outlines (like all visible targets do in isometric view) would also be a good way of telling the player which target their character is trying to hit. In first person view only one target would glow though, so that'd help.