I really like the 2nd edition of Warhammer Fantasy Roleplay. For its intended function and within its intended setting, it's one of the friendliest, smoothest-running systems I've seen. Its links to WFB remain obvious and sufficiently useful for inter-game borrowing, but it also stands alone as a proper roleplaying system of its own. It's not perfect, but it's flexible enough that even a mediocre GM can improvise through most complications and uncertainties.
The biggest trouble I've had with it has been excessive combat. I realise we're talking about a setting designed to maintain perpetual war between every group that can pick up a stick and call it a war banner, but that's to serve the miniatures game. The roleplaying game just uses that huge conflict as background flavour, something the player characters seldom enter directly themselves. Most of the official early '80s and '90s WFRP adventures (especially the excellent Enemy Within campaign that we're currently playing through again) were explicitly written in a style based more on Call of Cthulhu than Dungeons & Dragons. And their character levelling system (the 'careers') fits with this too; it's very unlikely any PC will ever reach the point that they can stand toe-to-toe with a major demon, dragon, or similarly powerful monster.
Yet WFRP2 remains burdened with a more D&D-like set of combat rules, inherited and adapted from the Warhammer minis combat rules. And combat in WFB (any edition) can be tedious as fuck when you've got a huge block of basic troops flailing stupidly at another huge block of basic troops, each side gradually wearing the other down with millions of dice rolls over several rounds. WFRP seldom has such great numbers, but its combat rules allow for more detail and complexity, so a 5-vs.-5 skirmish can last for even more rounds, and see just as many dice rolled, dragging the game on for hours, all to reveal that... a group of scholars and merchants and peasants are not very good at hitting stuff!
From a story-telling perspective, that's lame: A huge time investment to do very little and not advance an investigation plot at all. And that's annoying enough when the plot-as-written calls for a big fight scene against a group of, say, Chaos cultists. Then everyone's neatly boxed in to a fairly featureless warehouse for 2 hours, all to reach the final point of saying, "One side or the other won, and the following people got injured doing it." Unless you specifically want a combat game over a more plot-driven game, that's pretty inefficient and even boring.
And it's worse again when the combat wasn't even planned. It makes sense that the thief character would try a bit of burglary while in town. It makes sense that she'll fail from time to time. And it makes sense that sometimes that failure will be so terrible, no matter how lenient the GM wants to be, that it only makes sense that the town guard will be fully alerted, drawing a huge fight to the players. Sometimes there are believable ways around this, but the towns of the Empire are usually small and compact and densely populated; raising the alarm is easy to do (in a setting where invasion is frequent), and sneaking around is not. So now we're running an impromtu cops & robbers scene, where the only interesting outcomes are death, capture or escape. Three outcomes. Shouldn't be hard, should be a minor blip in the story-telling process. And instead, the rules as written turn it into yet another half hour or more burned away rolling dice.
So this is the long, waffling background to my new house rules. The combat in WFRP2 definitely isn't the most painfully slow I've experienced by a long margin (Pathfinder has, at times, been far worse, and the various Star Trek RPGs always seem to get shitty combat rules, as I've noted before), and sometimes it's even fun and interesting to play a decent WFRP fight scene. If you want to run a weapon-centric game, then it's actually quite a good option. But if you want to rush through most combat, then I propose the following (and bear in mind, this in the first version of this, pre-playtest). The main idea is to reduce combat to a single skill check, over a single round, same as any other skill check.
When a fight scene occurs, at the GM's option, each character in the fight rolls one attack roll (WS or BS, as appropriate).
If a character rolls a failure, note the number of degrees of failure. Otherwise, that character scores no significant hits.
If a character rolls a success, note the number of degrees of success. For each degree of success, that character has scored one hit against an opponent. Assign hits to specific targets. Assume all characters are free to move around as normal; character position only matters if the GM rules that someone definitely can't be hit (out of range, no matter how the attacker moves around).
(If you're using hit locations, the first hit is on the usual location, with the attack roll's dice swapped around. Subsequent hits land in locations equal to the original location value, plus or minus (attacker's choice) 10 x the number of the hit beyond the first. The attacker may never choose hit locations greater than 100 or less than 01. For example, you score 3 hits with a roll of 23, so hit 1 (n=0) lands on location 32 (23 reversed). Hit 2 (n=1) could be either location 22 or 42 (32 +- 10n). Hit 3 (n=2) could be either on 12 or 52 (32 +- 10n).)
For each hit, roll the attacker's damage, per the standard weapon damage rules, and subtract from that per the standard armour and toughness bonus rules. Defending characters with parry and dodge may use these as normal too, once each per attacker. (I've assumed that once per attacker is a reasonable substitute for once per round). The guiding principle is advancing the plot, adding more chances for the player characters to do interesting things, not roll dice.
If any target is killed and there are still hits assigned to it, re-assign these to other targets.
If either side is completely killed, they lose by default.
Once all individual hits have been resolved, for each side add up all its characters' degrees of success, and subtract from that all their degrees of failure. The side with the higher total degrees "wins" the fight scene, however the GM chooses to intrepret that.
The default assumption should be that the losing side flees. This is what most people would do in real life.
It's up to the GM to decide what sort of fleeing would be appropriate
and possible, though if the player characters are the ones fleeing, then
they should have some say in where they hope to escape to. Consider movement speed, encumbrance, stealth, terrain, dumb luck, and any other relevant factors.
If there is a compelling reason that the losing side might want to hold on to the bitter end (special training, Orc culture, defending their homes, etc.), then the closest thing the group has to a leader character may make a Command roll to try to keep everyone in place. With a successful Command check, that side still loses, but not by fleeing (see capture/kill, below). If the winning side includes the Terrifying talent, then the losing side automatically loses all Command rolls to resist fleeing.
(It's weird that WFB has this mechanism more or less built in, as Leadership checks, and yet WFRP's rules just assume that everyone's default choice is to fight to the death.)
If a losing group is incapable of fleeing (trapped, rooted to the ground, zombie mindlessness, geas, etc.), then they still lose, but not by fleeing (see capture/kill, below).
If a group loses, but doesn't flee (for whatever reason), then the winning group has to decide whether to capture or kill the losing group. Capture keeps more story threads open (especially if the players are the losing side!), but certain characters may be immune to capture (ghosts and such, which must instead be busted; characters with suicide pills, etc.). For other characters, capture may mean pretty much the same as kill, unless the captors take special precautions (I'm thinking fish out of water?). Resolve this decision on a case by case basis, i.e. roleplay it out. Consider each side's motives, morals, resources, and context.
(I think far too few roleplaying combat systems have substantial capture rules, even though that's a very common result in real skirmishes. It basically means the Altdorf city watch, by the letter of the roleplaying rules, will almost always kill suspects, and almost never actually arrest anyone.)
The one thing I know I haven't accounted for yet is magic. Maybe some sort of channeling test, similar to an attack roll, to see by degrees of success how many castings may be attempted?
This is still a lot of detail, a lot more bookkeeping than most other skill rolls in the system, but I think cutting out the multiple rounds and piles of pointless rolls will speed things up a lot, without altering the end result to the detriment of a good game. You'll also notice that I intentionally haven't removed anything at all from the original character and gear stats, so that when the GM does want to run a more detailed combat (maybe a boss fight against the supreme villain NPC, or a player-vs.-player duel to decide the fate of the party), then it's easy enough to revert to the official combat rules. Nothing is lost, and the GM gains the ability to fast forward through the more tedious slogfests.
I'll revise this once we've given it a play test. In the mean time, feel free to add suggestions for getting this to run even smoother and easier.