The series so far: general ideas and issues, possible systems, combat in the game, how to model it in tabletop, a homebrew possibility for handling the combat, and dicepools and buffers to increase the actiony-feel.
Last post I was wondering whether the dicepool system I mentioned could be combined with the combat system I'd already discussed, since I was quite pleased with both. At the time I couldn't think of any ideas, but I've now thought of something. The idea is that while player actions are constrained by a pool, it's not a dicepool in the conventional sense.
I've already discussed the idea of varying die size to represent the character's focus on a task. One possibility would be to generalise that further.
A character has a pool of five (which you can think of as dice, or action points, or whatever). The first die spent on a task allows them to roll 1d4 towards it. Each subsequent die does not stack, but instead increases the die size by two (that is, one size), so 1d6, 1d8, 1d10, to a maximum of 1d12 if full attention is given to the task. Combatwise, this should work well; enemies will be rolling middling dice, likely a d6, and so a PC can handle two enemies at once with reasonable confidence, while retaining a spare die to use for manoeuvres or to overwhelm one enemy. Alternatively, they can hurl their efforts into overcoming one enemy and take the other's attacks. I would be inclined to allow a flat 1d2 or 1d3 roll for passive defence as long as a character isn't locked in place, on the basis that even a character not actively fighting back is still moving around and potentially brawling with others, making them harder to hit. But I'm not sure.
The aggressive and defensive stances could be handled in a couple of ways. One would actually be to have separate attack and defence rolls, so characters choose where to focus their efforts every round. Another would be to spend dice to parry hits (one for one) to reduce the chance of injury, or for a bonus that applies to calculating how many hits you got but not how many you took. For example, you might roll 1d8 with three dice for aggression: a roll of 4 would give you a score of 7/4. If the enemy rolls a 5, then you inflict two hits but also take one. There may be a more elegant solution, though.
A slightly more complex idea is using additional rolls for non-standard stances.
In one model, if you fight aggressively, both combatants make a separate opposed 1d3 roll and incur additional damage based on the result. This is basically a gamble on the attacker's part, and wouldn't be worth taking as it stands, but the attacker would also gain a bonus to their main die roll.
In a second model, the attacker would roll two dice of the appropriate size; the higher is used to calculate hits on the enemy, and the lower to calculate hits incurred. This model gains particularly from focusing on combat, as you're obviously less likely to roll a low defence on a higher die, and more likely to roll a high score on one of the dice. However, it could still be worth it if you really want to drop an enemy but just don't have the dice to spare.
With 1d3, there's a 33% chance of inflicting additional damage, and a 33% chance of suffering damage. Whether this is useful depends on the damage and recovery systems. Essentially, as the PCs have to fight through many combats, a system with slow recovery or lingering injuries makes this a bad choice. On the other hand, if damage represent increasing vulnerability but can be readily recovered between combats, then eating some damage to put down an opponent quickly could be worth it. In particular, it matters whether injury is more HP-like (with predictable damage thresholds) or based on random rolls, since risking another roll on an injury table that could drop you is probably unwise. That being said, as this is a party game, it could still be worth it if you can be revived easily after the fight.
A disadvantage of the merging dice model is that spending more dice only affects the ceiling, not the floor. You can still readily roll a 1 on 1d12, and will do so fairly often. While I'm okay with that in the tumultuous world of combat, where you're (effectively) rolling against a variable difficulty (in the shape of the enemy roll), I'm less happy with it for other activities. Characters in Time Faffers are supposed to be skilled at acrobatic and athletic feats, and I don't want them to be constantly failing to climb trees or run along walls. What to do?
Well, I could actually simply have them succeed at things by spending dice, perhaps setting variable costs for tasks. Want to leap across to a ledge? One die. Want to run from a ledge along a wall, rebound into a window opposite just as gravity takes hold of you, and swing from the curtain rail into a chandelier in one smooth motion? That's going to be at least three dice. This system would probably scrap Momentum entirely, although I might retain die rolls for challenges like evading sawblades or timing jumps through fire. The advantage here is that you can easily incorporate moves into combat by simply spending the dice. A disadvantage is that it transforms non-combat elements into pure narrative, and the problem here is that unlike POPSOT, a tabletop game doesn't have the advantage of visuals to convey either complex room layouts, or how cool you look. Basically I'm worried that the platforming would just boil down to "you use five dice each round to do some stunts which were seriously cool, honestly, and then get into another fight".
Another alternative would be to have dice merge for combat, but work differently for other tasks. They could work like an actual dicepool, with set numbers of successes, as discussed last time. They could simply add +1 to a single die roll, which allows a character to automatically succeed at simple feats when focusing on them (1d6+4 means you'd pass anything with a difficulty of 5 or less).
A third option would leave players reliant on Momentum to accomplish difficult tasks, or to guarantee success. We could keep the merging dice system, so there's always the chance of a low roll. However, you gain Momentum for beating a target, which can be used to bump up later rolls. You need a 2+ to jump the gap, a 3+ to run along the wall, a 4+ to spring from the wall to the opposite window, and a 5+ to grab the curtain rail in midair and swing from it onto another ledge. This would need tying into some kind of speed system, as otherwise there's no reason for players not to simply spend five dice per round accomplishing one stage of the procedure. It would need balancing with whatever gets decided: the target numbers if the whole thing must be accomplished in one round (about one die per task) are very different from if you split it across two. I'd also want to consider whether players vary in their skill at different kinds of action, in which case they might gain a small bonus, or a boost to die size.
Finally, I could switch things about and allow players to spend dice as necessary. If the first roll fails, you can burn another die to try again (either as a reroll or an addition); otherwise you can move on to the next part of your task that turn. It's only if you repeatedly flubbed that you'd actually fail a task. In this model, difficulties and dice would mostly determine how long it took to complete a task, rather than whether or not you did - which for many aspects of play (just about anything unopposed) would be fine.
So while I've played around with variants, an actual dicepool is also a possibility. You could decide how many dice to dedicate to a task, and then roll them cumulatively. I originally thought that you'd want d4s or something small as the base die, in order to limit the range and make low pools beating high pools a real possibility. The larger the dice, the smaller the probability of one die beating several, right? Luckily, I did the maths, because apparently not. I suppose to some extent it makes intuitive sense, because it's more likely you'll get a freakishly high number on a small set of dice... yeah, I'm losing it. But them's the numbers what I got.
(positive result in your favour, any degree)
- 1d4 vs 1d4 = 6/16 = 0.375
- 1d4 vs 2d4 = 6/64 = 0.094
- 1d4 vs 3d4 = 1/256 = 0.004
- 2d4 vs 2d4 = 105/256 = 0.410
- 2d4 vs 3d4 = 121/1024 = 0.118
- 1d8 vs 1d8 = 28/64 = 0.438
- 1d8 vs 2d8 = 56/512 = 0.109
- 1d8 vs 3d8 = 70/4096 = 0.017
- 2d8 vs 2d8 = 1876/4096 = 0.458
- 2d8 vs 3d8 = 5560/32768 = 0.170
So actually, for a system like this, I might want to use d10s or something. A minor disadvantage of larger dice, of course, is that you're increasing the gap between averages, and so setting a difficulty that'll on average require three dice (16ish) puts it out of reach of a single die - but then maybe that's okay. Especially if we're using Momentum.
Using an additive dicepool has the distinct advantage of giving a built-in floor to rolls. More dice will always reduce the risk, and you can guarantee a certain minimum level of performance. This allows a trade-off: do you try to accomplish more in a round and risk failing some of it, or make certain you'll succeed at a small number of things?
Hang on, why am I not just using a normal target-based dicepool, you may wonder? Well, partly sheer laziness at considering yet another system, but also partly that I'm not convinced it brings anything new to the table. It still has the floorlessness problem seen in merging dice. I don't immediately see it offers any benefits in terms of fighting styles. Possibly it would make setting difficulties more straightforward, since intrinsically you can't get N successes without rolling at least N dice.
So, still not sure what's the best option here, but at least I've got a few possibilities that might work out.