Friday, March 8, 2013

How many dark lances do I need?

Ooo, I rolled a 6!


This is a question I often ask. In the aftermath of Arcanacon, where I ran a list that I knew full well had too little darklight (and still didn't score well on comp... go figure), I've decided to run some probhammer on the question.

(I do still intend to post reports of my remaining battles... I'm getting there!)

Firstly, we need to know what the probability of a dark lance doing some damage is, and how many we then need to assign to have a respectable chance of taking a vehicle out of action. Thanks to Hull Points, the math is a little more complex, but also skewed in our favour. Hurrah!



I find it useful to compute a 'Kill50', or how many weapons I need to fire to have a 50% chance of achieving a kill. In the simple example above (ie no cover, everyone has 3 HP) it works out at:

K50 AV10: 3.284 darklight shots
K50 AV11: 4.224 darklight shots
K50 AV12: 5.912 darklight shots

Of course, we can't fire 4.224 dark lances at a Rhino, so some rounding is required for a sensible value. I think it's also worth looking at the chance that, in allocating a rough K50 number of shots, other damage may be inflicted;


Shots% Kill% Damage
AV 10344.95%82.85%
AV 11447.04%80.25%
AV 12650.82%77.86%

And by damage, I mean achieve a roll on the damage table, whatever the outcome (stunned, dead, whatever). If a crucial vehicle is stunned and can't affect you next turn, I call that a win.

So it seems that by shooting enough weapons for an even chance at a kill, we give ourselves roughly an 80% chance of damaging a target vehicle and therefore inhibiting its utility against us. This gives us a clear guideline on the number of weapons we need to be pointing at targets, depending on their armour.

Of course, to extrapolate from that to how many lances we need in our army will depend on the number and variety of vehicles our enemies bring to the table; and also, how many of them we are prepared to accept functioning in any given turn.

From what I've seen in Imperial Guard armies recently, from 4 AV12+ chassis plus one flyer at 1200 points to double that at 2000 is quite average. If we go by the old maxim of one dark lance per 100 points, we have just enough firepower to suppress half of the Imperial Guard's ground forces at 1200, and not even that at 2000.

It's worth mentioning though that all of those Chimera chassis will go down in half as many shots when caught on the side. Thus, the key to dealing with superior armoured forces is to shoot them in their hurty bits whenever possible!

And, since the prevalence of vehicles becomes significantly larger much quicker in higher points value games, I'd suggest scaling your dark lance or equivalent numbers appropriately. 1:100 may be appropriate at 1200 points, but I feel it needs to double at 2000; at a minimum 24. For what its worth, this gives a trend like:



And of course, the elephant in the room is Flyers. If you thought the numbers above were bad, the K50 on an AV12 flyer like the Stormraven is about 25. The Necron flyers are much better though, with a K50 in the vicinity of 20.5, and by the time we get down to AV10 flyers like our very own Razorwing, it's only about 15.5. So if you can, deal with flyers some other way and save your limited dark lances to degrading the enemy's ground-based materiel. My personal preference is to get into close combat, where they can no longer hurt me. Failing that, getting behind or otherwise out of their arc can often negate them for a turn - bad luck if you're on an objective.

So how many dark lances do you need?

More.

You need more.

On a more serious note, we also need more data. Survey your regular opponents (surreptitiously, of course ;) ). How many vehicles are they taking per point? If we could put together global estimates of what we are likely to face in a take-all-comers scenario, then we could start to make an optimal darklight distribution pattern.

In the mean time, probhammer says: use the tables above to work out how many dark lances you will need to suppress half of your opponents vehicles, and take at least that many.

Hopefully, in a follow-up post I'll address the ratio of dark lances/blasters/heat lances/other weird-arse things in the codex. In the mean time, may your dark lances be as effective on the tabletop as they are in (our) fluff!

3 comments:

  1. Great post! Really helpful. Did you take into account that a double immobilise result destroys a 3HP vehicle?

    I made a java program a few months ago to brute force (repeating the simulation until I had the value stable to 3 decimal places) calculate the effectiveness of dark lances and heat lances a while ago (discussed here).

    My value for 3 dark lances against AV10 was 45.7% (wreck or explode). The 0.75 difference is probably from the double immobilised result causing a wreck, as all immobilise results after the first cause 2HP to be removed.

    Personally I think its important that you can achieve your first turn goal in numbers of vehicles destroyed, and have a steady flow of late game lances (blaster in warrior squads, reserve reavers with blasters, etc). The first turn goal of the 10 dark lances that I start on the board is to destroy at least one transport on the first turn (so that my splinter fire has a viable target). Baring in mind this is taking into account that I might be going second and lose some lances.

    ReplyDelete
  2. Thanks Mush!

    I must admit that I didn't take the double immobilised result into account. I didn't think it would make a big difference - 0.75 percentiles sounds about right. I should probably do it again including that for the sake of rigour, although I haven't worked out the best way of doing so. When I have more energy I will, for now it is a low-ball estimate :)

    I agree on destroying at least one transport in the first turn. We have so many excellent anti-personnel weapons, we can't have those juicy infantry targets hiding in their metal boxes! For that reason I'm moving away from using dark lance on the Razorwing, as they fool me into thinking I have more AT firepower than I actually have. Not on table = not effective.

    And how many dark lances to take to ensure that some of them are still on the table after your opponent has a turn... that is another whole kettle of fish! Conclusion: more :)

    ReplyDelete
  3. Yeah, I like to cheat when the maths gets to complicated and just make my computer do it by brute force. Programming makes you lazy. :D

    ReplyDelete