Tuesday, 24 June 2014

Attacking and Defending

There's a stylistic divide between combat styles that I usually like to see represented - the difference between a light, nimble, agile warrior, and the heavy, slow, strong warrior. Since I'm currently thinking about a game called Thunder Hunters where you get to hunt dinosaurs, it's a topic that's been on my mind.

This is me talking through my current design problem, so it's going to get a little long and technical.

In my PowerFrame RPG, characters have Attack, Avoid, Damage, and Toughness ratings. Attack opposes Avoid to score a hit, and Damage opposes Toughness to figure out the severity of the hit. In PowerFrame, the attacker and defender make opposed rolls, so there's a total of four checks for a successful attack. For Thunder Hunters, I'd like to try and cut that down a bit.

I'm also currently working with a d6 dice pool system. Stats would be rated from 1 to 6, which represents the number of dice you roll. Your opponent's stat is the target number you need for success.
For example, with an Attack of 4 against Agility of 5, you'd roll 4d6 and count one Success for each die that rolls 5 or more.

Two Rolls

To capture the fast vs tough divide, my initial thought was to have four stats that determine basic combat performance:

  • Attack, from a skill you can develop.
  • Damage, probably defined more by the weapon than the wielder.
  • Agile, representing mobility.
  • Robust, representing resistance to damage.

That's pretty much the same setup as PowerFrame, but the system works a bit differently. The defensive Agile/Robust are static target numbers, so that halves the required number of rolls.

  • The attacker rolls Attack, counting Successes against their opponent's Agile.
  • If they score at least one Success, the Attacker adds the Successes to their Damage and rolls that many dice against their opponent's Robust.
  • Successes against Robust allow you to create advantages or wear down your opponent. Enough Successes allow you to inflict a Kill result.


  • It's possible to model things that are hard to hit and accurate but weak and fragile, and at the other extreme model things that are easy to hit and dodge but strong and tough. It's also possible to model things that are other mixtures (accurate and damaging, but slow and vulnerable), or even good or bad in all categories.
  • It's easy to use different stats to deal with different attack types (for example, you dodge both a strike or a grapple with Agile, but use different stats to resist their effects).
  • This spread of stats provides more dials that can be fiddled with in combat, potentially resulting in a richer simulation.


  • You still have to roll two dice pools.
  • Two rolls to resolve an attack might not be consistent with the rest of the system, reducing things outside of combat to a simple single roll.
  • With only 1 Success needed to result in a damaging attack, the chances of dodging outright quickly become very slim.
  • Perhaps in the end, making two rolls is just a longer way to arrive at a singular probability?

One Roll

I am also experimenting with a different approach. The only two stats are Attack and Defence, with Successes on a single Attack roll resulting in combat advantage.

Weapons, armour, and other equipment would let you modify your Attack pool or reduce that of your opponent.


  • You only need to make one roll, so resolving an attack is fast.
  • It's consistent with other single-roll non-combat skill usage.


  • You need different resistance stats to deal with different types of attack (wound, grapple etc). Since you need these stats in the two-roll system anyway, I'm not sure if it's a disadvantage. However, something irks me about having a bunch of monolithic defence stats rather than being able to see the separate components (Strike Defence and Grapple Defence would both contain a measure of agility).
  • You lose the fast/accurate/strong/tough divisions. There's only dangerous/tenacious. While I think this could work very well for a game where the expected opposition is similar in capability to the PCs, I think in a game where people hunt dinosaurs it may not offer enough scope to describe the capabilities of different creatures.

At first blush it seems like a reasonable approach to have small Theropods be low Danger and low Tenacity, and larger ones such as Tyrannosaurus be high Danger and high Tenacity. It seems fine to be able to kill smaller raptors more easily and struggle to overcome a T-Rex. However, something was bugging me.

Then I thought a bit about traps. Imagine a spiked log swinging through the jungle. It scatters a pack of raptors; if it hits them it'll kill them, but they are small and agile and have time to get out of its way. Now imagine it swinging towards a T-Rex. It probably doesn't have time to get out of the way, but the wounds inflicted may not be fatal.

Ideally, this is how I view combat: against small, fast opponents there should be a low chance of hitting for a lot of damage. Against larger, slower opponents there should be a high chance of hitting for a little damage.

Where the single roll falls down is that against small enemies a dangerous attack has a high chance of doing a lot of damage, and against larger enemies it has a low chance of doing a little damage. The accuracy probabilities are out of whack. If we consider it in relation to the log trap example, it'll probably work almost as well against the T-Rex, but it will obliterate the pack of raptors.

In the end, using one roll won't replicate the probability of the two-roll method if we care about the effects of accuracy and damage, agility and toughness. That said, in many cases even a high Agility won't allow you to evade an attack altogether. If you're either going to get 1 Attack Success + 4 Damage against Robust 3 (average 3 Wounds), or 3 Attack Successes + 4 Damage against Robust 5 (average 2 Wounds), it's almost worth reducing it to one roll for simplicity's sake anyway.


I'm still mulling over the pros and cons, looking for options I might have missed.

It might be possible to apply damage and toughness as set numbers on top of the initial Attack roll, or based on it somehow. This would cut back on one of the rolls, but I'd have to make sure the numbers work and it's not too dull.

I might be able to define defensive stats by combining two other stats: Against strikes, use Agile + Robust. Against grapples, use Agile + Might. It would require me to rethink the way I'm setting up the stats, but that's not necessarily a bad thing.

There's this weird thing where I imagine an attack from a raptor and a T-Rex. The raptor is smaller and faster, but the T-Rex has much greater reach and a huge attack zone. Even disregarding the lethality of the hit, I can't see it being any easier to dodge the bite of a T-Rex than to avoid a man-sized raptor. So, it seems like Attack rating may actually go up with size, rather than down.

That also plays into my current Action Point system - I want to make being chased by a T-Rex daunting, by having it cost more to dodge its attacks. This sort of feeds into the feeling that I could reduce things to a single roll. If I reduce accuracy with size, it would be easier to dance around a T-Rex for days. If I reduce its attack rating, I'd have to come up with some other mechanism to drain AP.

Things are all up in the air at the moment, and I really need to get some stuff down on paper and see what happens. I think I'm probably going to work on the two roll system, but I'll keep an eye on it and see if I could somehow collapse the necessary stats to use the one roll method.