Tuesday, July 26, 2011

Actual Fight Mechanics

OK, so the basic non-fight component of a conflict is this: take a swing, beat a 4, if successful, the other guy is taken out.

(I like “taken out” as a euphemism because it underscores that all ways of taking someone out of a fight are roughly equal – unconsciousness, death, getting tossed overboard and the like all fall into the same bucket, and it leaves the exact color and fiction flexible. It also leaves a hook in place later to allow players to offer their own taken out outcomes if you want to avoid death in interesting ways, but that’s a matter to think about later.)

Once we’re in an actual fight, we don’t want things to be quite so quick as all that, but we still want to respect the 4+ success rule. We could go for a numeric system (hit points or the like) but let’s think of this in terms of statuses – we have this idea that a good enough roll can result in being taken out, what else might happen as a result of a roll?

Suppose that there are two other results – inconvenienced and harmed. Inconvenienced means that the other side has gotten some transitory advantage - they’ve knocked you back, rung your bell, seized the high ground or whatever. Harm is more palpable – it’s a disarm, an injury or some other major setback. And, of course, the third result (taken out) has already been recovered.

So, at their baseline, let’s map these as follows:

4 – Inconvenience
7 – Harm
13 – Taken Out

Now, that maps to our difficulties, but it raises some immediate questions. Does it mean that you need at least a 3d pool to be able to win a fight? And what happens when two people of high skill go at each other? Do they both just die? Obviously, we need to address these issues.

Now, the first is pretty straightforward, and we’ll do something that has been done in many other systems and just have damage “roll up”. That is to say, if you inconvenience someone who is already inconvenienced, they are now harmed. If you inconvenience someone who is already harmed and inconvenienced, then they’re taken out. Pretty simple. It allows high skills to get decisive results while allowing unskilled combatants to have sloppy, ugly fights that end badly.

Still brutal, though, especially since there’s no idea of defense. Skill won’t keep you standing any longer, and that’s problematic.

The fix for this is tied into how I view the statuses. Note that it would be normal to put a checkbox next to each status an fill them in over the course of a fight, but that’s more static than I like. I actually don’t want them to be static, I want them to come and go – not just inconveniences (which are already often tenuous in games like Fate) but harm and maybe even taken out. This means that, at a high level, I want people to be able to improve their status as they play, so that’s another axis of action. Sometimes it will be a dull axis (shrugging off an injury) and sometimes it’ll be flashy (getting out of a tight corner) but the bottom line is that status can fluctuate over the course of a fight.

(Saying that, my gut is suggesting we need a 4th status, just so there’s more room for things to slide. That may be true, but I’ll sideline that concern for now. If I figure out a good mechanic for this, then one good test for it will be expanding the status list).

So how should we implement this? My first thought is to make it something you can spend extra dice on (that is, for those who don’t recall, dice that weren’t needed to hit the target number). This has an interesting upshot because it provides a double incentive to go for inconveniences and harm rather than KO’s, because you’re more likely to have extra dice left, at least in theory. Of course, the fact that hitting three “4s” may be easier than hitting one 13 may also play into that. This also provides an interesting tool for NPC behavior, since the target number an NPC aims for speaks directly to their tactics, and can be a solid part of an NPC writeup.

Anyway, at the simplest you could just say that 1 extra die can reduce things by one “step”, so a hurt can become an inconvenience for 1 die, or go away entirely for 2 dice. That’s a good baseline, but it might be too easy. This is something to test, but I’d absolutely want to fiddle with different costs, including a higher base (say, 2 dice per step), or a sliding scale (1 for inconvenience, 2 for harm, 3 for taken out if apt – or perhaps the reverse!) but the idea is solid. It just leaves two real questions – how it interacts with a taken out result, and how it sequences. Those are pretty fiddly bits, so they’re best left for tomorrow.


  1. Hmm. I thought about how to use your ideas for a conflict system too and had the idea of using the remaining dice to cause damage. So, hit the "4" and cause an inconvenience, with any remaining dice you can cause worse results on your enemy. If you defend, you have to beat a 4 too - all remaining dice can be used to defeat remaining dice of the attacker.

    That would ensure the typical Musketeer flavour: Highly skilled fighters have long duels with each other, but goons are only dangerous when they outnumber the hero 3-1 or more.

  2. I'd suggest "handicapped" instead of "harmed" since it suggests a broader set of conditions. Harmed implies wounds, while handicapped could be something like getting one's swordarm caught-up in ropes on the ship. The other side has a large advantage, but you're not out of the fight yet. It also makes it clear that the handicap is not the wound itself, but the problems it causes, so losing overcoming the status doesn't imply super healing powers, but having a chance to do something like grab a crutch to help take the weight off your injured leg. The next time you're at "handicapped", it could be your crutch being destroyed or taken from you - no new injuries, just the old one becoming relevant again.

    If I understand the rules you're proposing, one potential issue is that a three-die character can beat a two-die character all of the time by playing defensively. The weaker character can drag things out (14.8 turns average), but eventually they'll lose.

    I'm running simulations comparing different strategies for the players and I'll post the results once I have a comprehensive set of them. The general rule I'm following is that reserved dice reduce the harm caused by the opponent, and any leftovers reduce the current state, to a minimum of unharmed. Both players roll and move simultaneously. One modification that I'd have to test is that harm prevention is 1:1, but repairing existing harm costs 2:1. This would encourage the three-die player to play more defensively all the time, instead of just buying off harm the following turn, since the two-die player can't take them out in a single hit.

  3. @Belchion Yes, that totally works, and I mostly just steered away from it because I'm feeling like trying something weird. if it all falls apart, that's totally part of plan B. :)

    @Mark yeah, I'm chewing on the scenario, but I'm undecided if it's a bug or a feature yet. On the bug side, you're right, it's very predictable - it would take a run of VERY bad luck for the 3d guy to lose. On the feature side, I really d want to underscore the idea that each die step really is that big a jump (especially at the low end) and this supports that.

    At some point, I would be well served putting a number on how predictably I want the 3d guy to beat the 2d guy. 95%? 75%? I'm not 100% sure at the moment.

    (also, some of this is also going to be informed by how the rules interact with multiple combatants. Even if I'm ok with the 3d guy trouncing the 2d guy the vast majority of the time, it's a very interesting question how he should do against two 2d guys)

  4. Here's some fairly rough data from running a bunch of simulations: http://alfedenzia.com/misc/robfight1.txt

    3-reckless vs. 2-defensive means that player 1 had three dice, and was attacking full-out each round, while player 2 had two dice, and would reserve a number of dice sufficient to heal completely, if possible. Any left-over dice are used to attack. In this case, player 1 wins 89.6% of the time, loses 7.6% of the time, with mutual death happening 2.7% of the time, and stalemate (defined as combat exceeding 20 rounds) 0.1% of the time.

    mixeddefense reserves enough dice (if possible) to prevent the opponent from probably killing them outright, but makes no effort to heal, while mixedaggresive plays defensively unless there's a chance to win in the next move, no matter how remote. More sophisticated strategies are, of course, possible and likely, but in many cases, playing cautiously wins the day.

  5. Do you envisage one check box for each status (and marking off the next highest check box), or requiring the opponent to make the roll to mark the check-box?

    Although I'd personally shift the levels up one with your system. That is:

    4 - success
    7 - inconvenienced
    13 - harmed
    19 - taken out

    With a success being the minimum requirement to start accumulating points from unused dice, but without necessarily inconveniencing the opponent.

    Then again I'm probably thinking of using the spare dice to accumulate "advantage" in a fight, which can be "cashed in" during the resolution phase to increase damage to gain a result. Of course, your opponent's advantage negates yours.

    This would represent things like parries drawing your opponent more and more off-line/off-balance, until you make an opening and use it, or forcing your opponent back towards the edge of the cliff.

    And rating advantage gives a good indication of how the fight is going at any moment. You have your natural back and forth, without any death spirals and skill use becomes positive (constructive), in that you are attempting to produce a result.

    And similarly an opponent could cash in their accumulated advantage in order to reduce the effect of your strike if you roll well. Representing a wild last minute parry/dodge that puts them off-balance, but saves them from taking a serious hit.


  6. @rev Advantage actually works _magnificently_, and I've designed a few Fudge systems along similar lines, but they fall into the one great trap of combat systems - they're fantastic for one-on-one duels, but quickly get very kludgey for general melees. And, to be totally honest, I may be about to run into the same problem again with this approach - I'm tryign to steer as close to it as I can without actually colliding. Won't know til I'm done if I pull it off.

  7. @mark Thank you, this is definitely feeding into thoughts for the next post?

  8. Of course you can always treat an entire battle as a duel by using advantage to simulate the strategic result rather than the tactical. Also means you could apply advantage to maneuvers such as bugging out or protecting the princess, in situations where smashing your foe to the ground isn't the actual objective of the fight.

    [I admit the main reason I find this idea interesting is that it makes a combat an objective that you are building toward, rather than reducing the opponent. It also means that appropriate non-weapon skills can also be used to assist the battle directly.]


Note: Only a member of this blog may post a comment.