Frag or Krak?

I'm not keen on weapons with options. I always end up worrying that I've made the wrong choice. Case in point: the grenade launcher. Should I fire frag or krak grenades? One gets me more goes with a worse grenade, the other gets me a single shot that might actually do something.

Trying to solve this sort of problem on the fly is not something I feel the need to add to my game. I already have enough to think about – trying to optimise my grenade launchers during the shooting phase is just stressful. So why not eliminate that stress in advance by math-hammering the problem?

The question here is this: am I better off firing a single, pretty effective grenade, or a variable number of less effective grenades? Both have issues – if I roll up a single frag, I should have gone krak; with a single krak, I’m going to miss half the time. But both also have advantages – krak is an S6 weapon with d3 damage; frag has the potential to deliver six wounds, and volume of fire (albeit on a small scale here) tends towards consistency, which is always nice.

For reference, a frag grenade is S3, AP0 and does 1 damage, but gets d6 shots. A krak grenade is S6, AP-1, does d3 damage and fires a single shot.

DAMAGING HITS

The table below shows a standard comparison of toughness and save (toughness down the side, save along the top). Where they meet, the number in the box represents the number of frags you'd need to roll up to achieve the same chance of causing damage as a krak grenade. If the box says 2, for example, then 2 frags should achieve the same as a single krak (or will do slightly better). At this point, we’re only concerned with causing damaging hits, so the krak grenade’s d3 damage isn’t in play.

A green box suggests that frag is the better option; yellow, that frag and krak are balanced; red suggests going with krak; and grey indicates that even 6 frags will be worse than a krak (statistically speaking). My thinking here goes like this:

Against a theoretical T1 7+ save target, both types of grenade will wound on 2s with no save possible. So even a single frag will match a krak, and anything more than that is a bonus favouring the multiple-shot frag grenade. As the toughness or save increases, a krak grenade becomes more likely to cause damage than a frag: against T3, 4+ save, a krak wounds on 2s, and the target saves on 5s; a frag wounds on 4s, and the target saves on 4s. Factoring in the hit roll (at BS4+), a krak has a 0.28 chance to damage; a frag has a 0.125 chance. But three frags have a 0.375 chance of damaging. And there’s a 2/3 chance that you’ll generate 3 or more frag shots, meaning that, overall, a frag round is more likely to cause damage, despite its weaker stat line.

Following that logic, if you only need two frags to match the krak, go frag – there’s a 5/6 chance of doing the same or better than krak. At three frags, we have a 2/3 chance of doing better with frag, so go frag again. Four frags is the critical point. If four frags will do slightly better than a krak (yellow boxes), it's a toss up - half the time, frag will be worse; half the time it'll be better. So you might as well go with your personal preference. But if four frags exactly match a krak (red boxes), then go krak - now you have a 50% chance of doing worse with frag (rolling 1-3 shots), and only a 33% chance of doing better (rolling 5-6 shots). And once you need five or six frags, stick with the krak (unless you feel lucky, I guess). Curiously, frag comes out ahead in almost every case. Krak is only the better choice against a 2+ save (if the target is toughness 4 or higher), and against toughness 6 (if the save is 4+ or better) - toughness 6 is krak's comparative sweet-spot, where it's still wounding on 4s but frag has dropped to 6s. The thing is, whilst krak is better than a single frag, it’s still not amazing. This means that frags can quickly make up the ground once we get into multiple shots. So in most cases, if we're just looking to cause single wounds, frag is the way to go.

TOTAL DAMAGE

However, all of the above is (intentionally) ignoring the krak grenade’s d3 damage. On this second table, I've doubled the damage output of the krak grenade to represent its better damage characteristic - I've averaged it to 2 to make life easier, but it does mean that this is something of an abstraction. In some cases (such as single-wound targets), this extra damage is wasted, but sometimes it matters (looking at you, marines).

Broadly speaking, this doubles the number of frags required (with rounding accounting for the appearance of any odd numbers here). Suddenly, the weighting shifts heavily in favour of the krak grenade. Now, frag is only the better choice against toughness 1 (unless the save is 2+), and toughness 2 (if the save is 6+ or worse), and that's a ridiculously limited choice of targets - in practical terms, if we want to maximise damage, we should always choose krak.

(You’ll notice that some boxes are now grey – these represent situations where we’d need more than 6 frags to match a krak. As such, the krag grenade always outperforms frag in these cases. As we might expect, this is primarily an exaggeration of the previous table, with T6 and 2+ saves once again the krak sweet-spot.)

There’s some slight fuzziness here. A krak grenade can, at best, still only kill a single enemy model. A frag grenade could potentially kill three 2-wound models, but it’s massively unlikely (somewhat less likely than a one-in-a-million chance, quite literally). Flukes can occur, and dice can roll hot, so the best that this table can do is suggest the option that is more likely to produce the most damage over multiple repetitions.

CONCLUSION

This one is fairly clear cut. If you're going after one wound models and want to spread damage around, use frag; if you’re targeting models with multiple wounds, krak is more likely to cause more damage.

I know that sounds obvious, but the bit that surprised me was how universally that applies, especially with frag. In my head, frag is for low toughness, lightly-armoured targets. But the numbers suggest that (with a few limited scenarios aside) frag is the better option for single wounds, even against toughness 7 or higher. Or to look at it another way: the factor that primarily determines the grenade you should use isn't toughness, or save - it's the number of wounds on the target.

Of course, as toughness and save increase, the likelihood that a model has only one wound decreases, so frag may remain a rare choice against those targets. But it’s interesting to think that, if you really need to remove the last wound from an enemy tank to win the game, you’re probably better off opting for frag.

Thanks for sticking it out with my increasingly niche ponderings. I’d hoped that we might all be browsing through the new codex by now, but that’s not really worked out for anyone. Fingers (and claws) crossed for an imminent arrival…