We talk about D&D a lot around here, and last week we were talking about combat length. We have different measures of combat length: theoretical combat length, actual combat length, desired combat length, and all those loose numbers were coming up. We started charting theoretical fun per round of combat, and our loose consensus was that fun takes a nosedive after round 4 or 5, when PCs have used up many of their resources (and thus their tactical choices) but there might still be two or three rounds of combat left.
Playing two or three rounds of combat (or even one) where the funometer hovers above the zero mark is not what we want, or what we want to provide to players. Now, we do a lot to gameplay in an effort to make fun cascade naturally from the rules like water over Multnomah Falls, but now I'm thinking of a brute-force solution.
Every combat ends after four rounds. Why? Because after four rounds, the outcome is known - not necessarily by the players at the table, but by the PCs and their enemies. Battle is joined, everyone fights furiously for 24 seconds, and then the smoke clears and everyone can see the writing on the wall. Maybe in character, someone doesn't and chooses to keep fighting, but the fight is already won.
When a four-round combat ends, many of the combatants will still be standing, unlike in a standard combat when it ends only if and when one side is reduced to nothing. It's like when two nations go to war: Only when the goal is death does one keep marching until the other side is nothing but rubble and bloodstains - usually, the conflict is a means to an end. Clausewitz described war as a continuation of politics by other means, and we can apply that to other conflicts as well. After four rounds, one side sees the writing on the wall and capitulates to the other.
Of course, many PC groups will go ahead and execute any of their enemies who decide they can't win. That's the kind of reputation that causes enemies to stop surrendering. Much better to take prisoners, ransom valuable creatures back to their lords, send others to prison, or extract bonuses from them somehow and let them go. Maybe you can even make some allies.
If both sides are still standing, who won? It's hard to just decide that one group or the other came out victorious. How the next few rounds would go are determined by many factors, including hit points, available tactics (encounter and daily powers), current tactical position (are the archers safe from melee? are reinforcements about to arrive? is every member of one group or the other flanked?). Keep in mind that this is all theoretical (i.e., I haven't tried it yet), but here are some ideas.
Percent of Group Hit Points Remaining. At the end of the fourth round, sum up current hit points and total hit points for both sides, and compare the fractions. Whichever group has a greater fraction wins. PCs have a decided advantage here because of their many healing powers... but they have a sizable (and planned) advantage in most combats anyway, so that might just be as planned. This is pretty simple, and you'll want to bend it as you go: the dragon doesn't care that the petty humans are healing, it believes it can still outlast them.
Available Combat Resources. For every limited power still available to either side, add 5% to that side's hit point fraction. Don't count PC daily attack powers or utility powers, or they'll always have too many to be defeated - PCs have to use them or they don't help. On the enemy side, any unexpended encounter power counts, as does any unexpended recharge power or any recharge power that recharges on 50% of the time or better. (For the monsters, non-attack powers count.) If a monster has some power that isn't limited but has a huge impact under certain conditions and those conditions are in play (the vampire is grabbing a PC, so it could bite that PC for big damage and hp regain), count it for this purpose.
Current Tactical Situation. Who has the tactical advantage when the four rounds are up? This is less specific, but add 5% to either side for each major tactical advantage. Has the party created a zone that gives only enemies difficult terrain? +5%. Is the fight in a dense jungle, and all the enemies have forestwalk? +5%. And climb speeds? +5%. Artillery has secure positions? Wizard has a wall of fire up? Half the party is invisible? Everybody on one side is dazed or prone? Half the other side is blind? Someone on either side is dominated? Et cetera.
Add it all up and see who comes out on top. If the values are too close, within 5% or 10% of each other, you might want to run the combat for another round. But on the other hand, you just did all this math... so maybe you don't.
Alternative Simpler Method. You could also just count the hits each side gets. Each time one side hits the other side with an attack, +1. When one side drops someone on the other side (even if it's a PC and it gets back up), +1. Other major adjantage (dropping a fellow down a pit for a couple rounds, wall of fire, someone is petrified), +1. At the end, count up the hits and see who wins. I resist this because it minimizes the impact of damage rolls... though if dropping someone is worth +2 instead of +1, maybe is doesn't. Hmmmm....
Why is the Combat Short? Because I said so. For the purpose of this Gedankenexperiment, it's because after four rounds both sides can predict the outcome. I chose four because, though five seems like it would still be fun, four puts a little time pressure on both sides. You only have four rounds to make the difference, go! If you're the sort who is uncomfortable coming to this conclusion with your players, you can try this with in-game reasons. After four rounds, the city guard will come to break up your fight in the marketplace. After four rounds, there will be only just enough time for one side to dash through the closing portal out of the Abyss. After four rounds, the sleeping gas will knock everyone out, and whoever is more weakened will sleep longest. You're creative. You can think of something.
There are some very reasonable comments on the unnecessary complexity of the above system for determining which side is at the advantage. I agree. I was also up against the deadline of an impending meeting when I wrote it, so I had little time to explore alternatives. Let's do that now, shall we?
Best is if, when the timer dings, you can look at the battlefield and it is obvious to everyone which side is in charge. When that's so, great! That side wins.
If that's not possible, here's a (much) simpler alternative to the above mathtangle. Count each PC that is bloodied. Count a PC twice if it is at or below 0 hit points, or if it is otherwise removed from the fight (however temporary); petrification and domination are good examples. The enemies get that many points. Count the number of enemies in the same fashion. The PCs get that many points. Team with more points has the advantage. Add bonus points for special circumstances, such as a secret resource one team has.
Consequences. One result of this idea is one I really like, and the reason I might use this over reducing hit points or upping damage as suggested below. PCs and opponents have a reason to stop fighting and give up / let the other side live. If you keep fighting, you're going to take more damage and be worse off for the next fight (in PC terms, you're going to expend more surges before the next encounter). Stop early with a victory, and you can hit more encounters in a day. I think this is more likely than the rapid burnout suggested below.
The other consequence that comes to mind is less fortunate. We've all had those amazing combat scenes where two or three characters drag on at 3 hit points and no other resources just long enough to save their comrades and barely drop the beholder. Those would never happen again, which is unthinkable. That's why anyone playing with this system should also include fights where no one is willing to back down - the PCs know that surrender means death or is otherwise unacceptable, or the monsters are berserk, mindless killers, or just too proud to stop fighting. And in some difficult fights, the PCs can keep fighting to the point where they are hanging by a thread - which is the fun, nail-biting part - and bring their opponents to break off the fight to preserve their lives or resources, or just to get away from the crazy itinerant warriors who won't stop fighting like a reasonable person.
Which makes me think the PCs should be given the option - when they're on the losing side. You can always say, "It looks like they have a significant advantage over you. You might do well to surrender." If the players know you're playing with 4-round combats, they'll understand what that means and be able to decide whether take it to the limit or play it a little safer.