SWAT 4

Rants, Rambles, and Raves… On Tactical FPS games

From a person who has been gaming since he was in diapers, using a Nintendo, and all the way into the future.

More realistic ballistics:

In old times, weapons were little more then you hit target, you do damage. And each hit from a given weapon doing so much damage, like 10% of life bar per bullet from SMG, 20% from rifle, Which is a piece of crap to start with, even by modern standards. More modern implementations might take into account mass, velocity, and hit modifiers.

Objects are generally boiled down to objects that can be penetrated and objects that can’t. In Raven Shield they take it a step even worse, an object that can be penetrated is usually limited to a destructible one. I.e. a bullet may penetrate a pawn (rainbow, tango, hostage), a window, or a door. But it won’t go through a carboard box, a cereal box, or wood fence, because these are all “solid bullet proof objects”.

When a bullet hits something, the mass of the bullet and the velocity it hit at should be taken into account along with the type of material hit and the density of it and the bullet. For example compare trying to penetrate a strong tungsten alloy and a tin can, which takes more punch?

SWAT4 essentially operates on the concept that if an object may be penetrated, the bullet has M mass moving at V velocity used to calculate it’s momentum. While the object being hit requires a certain amount of momentum to penetrate; if I looked deep enough into the games SDK, I wouldn’t be surprised if “momentum” was just a synonym for velocity to the developers… When the momentum is exceeded, the the damage caused by the bullet on it’s way through is modified by a modifier range or factor of how much damage it can still do.

What should really happen, in my opinion?

There should be several types of impact material, e.g. liquid, wood, concrete, metal, plastic. The idea is pretty simple, let’s say you take a gun and fire into a pool of water. For an object to occupy a space within that water, it has to displace the same volume of water out of it’s way, to move through it, it has to keep displacing water out of the way. Shooting through block of wood that is as thick as the water (e.g. same volume of) works about the same way, but the density is different, taking more force to displace the wood out of the bullets way in order to ‘penetrate’ the wood.

For the sake of performance it would make sense to have several known (common) densities with known properties and tune accordingly. It would still fall short of accuracy with the real world but make it easier to apply ballistics to all objects within the game.

The biggest problem with ballistics in modern FPS games, they continue to use the “can penetrate” and “can’t penetrate” stupidly…

Proper damage models:

Shooting a the representation of a human in a tactical FPS should have similar effects to the real world. Most often, you get a set of hit boxes based on superficial anatomy. A head, a torso, arms, legs, etc. And if your lucky a modifier that adjusts damage done based on the hit box, e.g. head shots fatal, leg shots minor damage or something like that.

I think they should take it a step further and apply more things based on the hit boxes. For example, why not take a leg? Break it into 4 hit boxes, the foot, the crus, the knee, and the thigh. If the target gets hit in any of the leg hit boxes, they should die! Bleed to death or fall into shock awhile later maybe… but not die out right as in games. Getting hit in the crus or the thigh should effect ones ability to walk, getting hit in the knee or the foot should probably force the target to crawl instead of walk.

The torso should be split into h8it boxes based on internals, the “general torso”, the heart, lungs, maybe the kidneys, e.t.c. A human might be able to survive several shots to the chest but I would think a double tap to the heart would have more (quickly) lethal effects, even if not as much as going for a T-Shot, I would think it more likely to disable the enemy in the sense of more games, and kill the target more readily.

Compared to games like Raven Shield (last patch) where it doesn’t really matter where you shoot a tango, just hit them 2ce and they are as good as dead, pardoning latency and the FN Five-Seven. And SWAT4 where the only side effect is it pays to follow the Mozambique drill — just to do enough damage quickly to put the target down. While in real life, the idea is if the double tap to centre mass doesn’t stop the target, taking down their nervous system will put’em down for good.

Weapons Attachments and kit restrictions:

I think AAO:SF has the best in terms of realism but makes for a too oft abused configuration, M4 with aimpoint + suppressor + M203. Raven Shield limits you to one attachment but that sucks, especially since it makes weapons like the AUG, SA80, and G36K more powerful (built in scopes) — further pissing off people who know about the Tavor Assault Rifles!

I think the proper solution would be to allow 2 attachments for all primaries, such as an barrel and an receiver attachment. So using the M4A1 carbine as an example. One could have an M4 with a scope and a suppressor or with a M203 and scope but not a bipod and a suppressor or an M203 and a suppressor, e.t.c. That would at least keep it fair.

Load outs, I really think a merger of the “package” and “individual” item ideas would work. For example, so many slots, like something like this:

Player Kit:
— Primary Weapon = None | SMG | AR | SG | SR | LMG | LAW
——— Front Attachment: None | Suppressor | Grenade Launcher | Bipod
——— Rear Attachment: Default sights | CQB scope | Magnifying scope | Night scope | Laser aiming module
——— Ammunition: FMJ/Slug | JHP/Buckshot
— Secondary Weapon = None | Pistol | TASER | Special
——— Pistol Attachment: None | Suppressor | Laser aiming module | Tactical light
——— Ammunition: FMJ | JHP | Cartridge

— Specialist item = None | Binoculars | Radio | Laser designator | Night vision goggles | Spy gizmo | Gas mask

— Demolitions item 1 = None | Breaching shotgun | Door breaching charges | Plastic explosive | Satchel charge | LAW ammunition | Breaching hammer | Breaching bar

— Protective Gear = None | Light | Medium | Heavy

Where light = Level IIIA, medium = level III, heavy = level IV on the NIJ standards.

And allot like 3 or 4 slots for grenade “packages”, say 3 slots holding a “two pack” of grenades.

— Grenade slot 1 = None | 2 x Fragmentation | 2 x Flashbang | 2 x Smoke | 2 x CS Gas
— Grenade slot 2 = None | 2 x Fragmentation | 2 x Flashbang | 2 x Smoke | 2 x CS Gas
— Grenade slot 3 = None | 2 x Fragmentation | 2 x Flashbang | 2 x Smoke | 2 x CS Gas

Which in my opinion would be the most flexible and much faster then having to equip each slot individually.

Breaching methods:

Ballistic breaching with shotguns, any shutgun capable of destroying the door and frangible breaching shot as well — in effect making hinge and knob side breaching possible. Explosive breaching with any suitable explosive, as well as dedicated door breaching charges, using shaped charges to blast through doors and windows, etc. Mechanical breaching with bars, sledge hammers, and even fire axes that just happen to be sitting handy.

The ability to break down doors, blow up walls, come crashing through windows, is just a requirement. You can’t have a tactical game unless you can conduct a full breach. Not having to use the “door” is also more realistic, why send everyone through one door… When we can sneak a team into the next room, frame charge through the damn wall. And employ snipers to pick off X-Rays through the windows — after coordinating kill zones ahead of time that is.

Vehicles:

One of the big problems with games is the “vehicle” selling point. In games like BF2 or Halo, they often become the gods of the battlefield, you have to use the vehicles to win.

By nature of things, vehicles make good transport and are very useful. But have to be controlled to avoid them becoming all powerful monsters of destruction in adversarial games. Having stuff like personal motorcycles and dune buggies or HMMWVs for transporting troops would be nice.

What I would like to see is AI controlled support vehicles that can be called in by team leaders using a “radio” item. So for example, a squad leader could call in an attack helicopter or a gun ship and have it circle in a radius overhead, attacking every enemy in sight.

To make that even sweeter, in some maps the enemy could have AAA and shoot the thing down !!!

That would open up so many possibilities, such as vaulting a few DPVs into the target area, send a group/groups to secure the objective, while the other members of the patrol set up a parameter, guards the cars for egress, and calls in a few helo’s for backup. The possibilities for hit and fade runs are infinite when that is combined with a map editor…

Game Types:

Round based Cooperative, following “mission” and various special coop types, e.g. search and destroy, recon, hostage rescue. Round based and respawn based team death matches in “objectiveless” team with the most points wins, capture the flag, and search and destroy games. Heck, you could probably convert coop levels over for use as player Vs. player missions by using some form of squad based spawn points for the defending team.

Mission Planning:

It’s imperative to be able to play the mission ahead of time and adjust in combat. Raven Shields F4 map sucks horribly but at least you can draw on it. A real planning map that can accommodative multiple floors would be perfect.

Team leaders need to be set, teams organized, mission objectives reviewed, and assault teams synchronized on their objectives.

Human Capabilities:

People can walk, run, crawl, roll, jump, e.t.c so why why not do like wise in game? One of the most annoying thing in Tactical FPS games is when you get killed in one of those, “I could have just … like in real life” moments. Walking should be quick but not to quick, watching a SWAT team on the moves a perfect example. Running should be quick and really jumble your aim around. The ability to crouch and crawl is a requirement IMHO, and on top of that. When going prone while running or dashing, one should go into a dive.

If your walking and drop to the ground, odds are you want to get on the deck eh? But if your running like a bat out of hell, odds are you might not want to stop where ya are and lay down!!! I think to avoid the bunny hopper problem, jumping should be limited to a very short height and while running, just enough to be useful, yet realistic for someone in ~15 plus kilograms of combat gear.

That would allow one to still move quick, jump quite short heights, and avoid the usual crap of not being able to jump at all, because nubs and pansies abuse it so often online.

One thing I would really like, is a special “physical” action button. One that changes what it does based on the situation, whne close to people, go for a rifle butt/pistol whip/CQC attack, when close to objects, try to push, throw, kick, or otherwise employ them. Such as kicking a chair across the room or flipping a table over — be versatile.

Team Organization:

It should be kept simple, AAO is a perfect example. Platoon leader, Squad/Section Leaders, Squad members. I think it’s useful to occasional have specialists but when it goes to the extent of say, BF2 or other class based configurations. I think you really loose all of the flexibility gained by having a tactical FPS modeled on anything beyond the general infantry…

Kit selection should define specialization options, not soldier “class”.

A perfect example, in RvS if we need more grenades, we bring more grenades, if we need a demolitions man, we bring demolitions! We don’t limit the demolitions man to a special class, we equip for the job. I think SWAT4’s split of grenades, breaching, and protective gear is the right style but the devs chose to divide it wrongly in terms of a Tactical FPS. But for simulation of Police SWAT environment, quite well since there are less, uhh… Fun toys to play with in most police departments >_>.

Graphics and Performance:

An Intel or AMD chip with performance comparable to a Pentium 4 in the 2.0-2.2 Ghz range, 768MB, and a 256MB Graphics card supporting OpenGL or DX9 decently would be an ideal minimal spec IMHO.

If you have a super gaming rig, you should get what you paid for within reason. If you only have a common gaming rig, you should at least be able to play the game smoothly with lower graphics. Games should look as *nice* as you can make them on your hardware but you shouldn’t need a $300 graphics card and double your RAM just to play.

Hell, a good make of GeForce 6800 is dirt cheap now and one of the best graphics cards of it’s generation.

Communication Functions:

Text chat, V-Commands, and Text Macros.

Text chat is the lowest common denominator of the internet, and most games require a keyboard and mouse to use properly ^_^. Pre recorded Voice Commands like that used in MOHAA, RvS, SWAT3, SWAT4, BF2, and such are great. The only problem is game developers almost never choose decent stuff and many gamers never bother to use them.

V-Commands should be a quick pop up menu and issued through keystroke OR mouse gesture and drop a HUD or on screen indicator that says WHO, WHAT, and WHERE. Like in SWAT4, you can see a marker of WHO issued the V-Command, you can hear the WHAT part, and when orders are given you can see the WHERE part marked on screen.

In game VoIP has always been a pile of shit, no matter what game that has implemented it, it has ALWAYS been shit. Striking a deal with the TeamSpeak or Ventrillo people for a dream-team setup that allows the game to auto-join certain servers/channels from in game or when just joining the a given team on the server.

I’ve always liked TeamSpeak because its channel commander feature offers a lot for serious team gaming.

The ability to record text macros in game, without having to use keybinds like in Unreal 2 games, I think is something every FPS should do. SWAT3 is one game that really got it right, text + sound bites. Although, being able to open your configuration data in a text editor and adjust things to taste is, like the hugest perk ever created.

User account and Points / Rewards system:

Having any kind of spend your life playing this game and racking up kill points, and we will reward you with unlocks. Is just a crock full of shit, it’s a marketing gimmick to make you play the game longer and an excuse to sell you expansion packs that add features that, often in my experience — should have been there in the first place. Or freely able to be made by a modding community.

Game Mods:

SWAT3 was perfect, download ZIP file, copy ZIP file to folder, flick the on/off switches for what mods you wanted to play. Launch game, and play. The mods you had on got used when hosting the game but when joining OTHER peoples games, WITHOUT RESTARTING. The game would load the required mods for joining other peoples games, as long as you had them installed.

That is the way it should be, having to restart the game to swap mods or worse, having to use a conversion utility (e.g. MW4:Mercs) are just piles of crap, unless of course. You want to appear to be a company that would rather lock people into buying expansion packs or sequles that your not obligated to make, rather then letting people extend what hopefully is a decent game, and make it better!

User Accounts and Anti-Cheat:

The games got to have some way of ID’ing users uniquely, PB generally does a decent enough job of it. While I don’t care much for PB, and in fact have never had much use for it as anti-cheat (aside from the spying part). It does do one good thing.

It helps enforce server policy.

Level Design:

Missions need multiple ways of completing it, unless the game is a “quickie” that you’ll play once and then chuck on a CD-Rack, which means there is no point in paying more then ~20 bucks for it.

Problems should be approachable from many different angles, there should be multiple paths available on most missions. Objective and enemy placement should be able to be randomized, one of the greatest things of early SWAT3 was that the missions placed some “mission” specifics in a normal set of spawn points, then placed other stuff in more random locations. No playing from memory, you know X, Y, and Z are choke points, you know hostages may be on Level 2, but only the machine knows what you will find in between ^_^.

HUD:

A customizable hud, the ability o change colors and whats on it works best.

The ability to have a small tactical map or SAI like element is ideal. Having helmet cameras and UAV feeds available and the ability to reorganize the HUD and toggle cams on/off based on position in the team — is even more fun 😉

Cross hairs should be used but be able to be disabled (I for one prefer no cross hairs on my HUD, unless I’ve got grenades). Iron sights and scopes should work on every weapon available but allow for proper shooting. In CQB, you don’t aim with the sights for every shot, learn how to use the weapon.

When it comes to ammo counts, I think the traditional rounds in the magazine counter is nice, although not really necessary once you get used to how the game tunes the rates of fire. What is important is knowing how much ammo you have left. I love how SWAT4 implements it, you know how many rounds are in your current magazine. You can see how many magazines you have on the HUD and how “relatively” full they are. Magazine based ammo tracking is much more accurate the the old NNN total rounds remaining ammo counter anyway.

I think moving some features that would really improve the value of a HUD, into a separate “PDA” screen would be useful. B ecause then one could really stretch things to realistic limits, targeting people like Robocop or a T-800 are a bit futher off for most troops in the 21st Century.