Friday, July 22, 2011

40k-metrics: Frequently Asked Questions

Since I first started talking about my metric system, I’ve gotten a steady stream of people asking how to calculate various aspects.  I thought I explained that in the first post, but apparently not clearly enough.  So rather than answer questions forever, I thought I should make a quick and dirty guide to doing a breakdown. 

I’m honestly not going to teach you how to mathhammer, if you can’t calculate wounds on your own, you’re hopeless.  But I will go over the various assumptions I make and how to arrive at the same (or close enough) numbers that I use.

It's all about the numbers
First, to calculate BS4 shooting hits, I do not multiply by 2 and divide by 3.  For a plethora of reasons, I just multiply by 66%.  In fact, I do all my multiplications by the percentages rather than ratios because not only are they easier to do in the long run, but they give a consistency that you can’t get with ratios without doing a lot more work.  So a 2/3rd chance = 66%, 1/3rd chance = 33%.  Feel free to do ratios instead, but if you wonder why our numbers differ slightly, that’s why.

I assume 3 hits for small blasts, 5 hits for large blasts, and 5 hits for template weapons.  No matter what numbers I picked someone would disagree, so deal with it.

Armor saves, like shooting and close combat, I use percentages.  So when calculating how many marines pass their armor saves, I use 66% rather than 2/3rd

Twin-linked and Preferred Enemy are other tricky topics for some folks.  I guess I will teach you how to mathhammer after all, since this was a big issue.  Figure out your initial hits and write that number down.  Take the amount of misses, and multiple those by the chance to hit and add the second group of hits to the original hits.  The same thing works for things that allow re-rolls to wound, like poison or lightning claws.

Remember for vehicles even though the statistic is “Dead Rhinos Per Game” it is actually measuring penetrating hits per game.  I explained this multiple times but people are still messing it up.

Close combat and vehicles.  Cap any single unit at 15.00 per game to keep the numbers sane, otherwise a unit like Nobz would have more penetrations than entire shooty armies.  Plus, it just makes more sense that way.  Also, be very conservative to which units you give CC hits against.  A unit of tactical marines should not be given hits against vehicles just because they have a powerfist and krak grenades.  I would only give the hits to a unit that is designed and relied upon to be a source of anti-mech for the list/army.  If the list doesn’t rely on that unit’s CC ability to kill mech, don’t include it.

I think that was all the FAQs, but if you can think of any more, post them in the comments and I’ll edit this post to include them.  Have a good weekend.

edit: all close combat anti-vehicle swings assume the vehicle moved at combat speed, and thus a 4+ to hit.


  1. Thanks, I was working these through and actually figured out the two problems I was having by accident while I was tired.

    The 2/3 vs .66 issue isn't going to produce that much of a difference in the final totals to affect the broad stroke inferences you've discovered. However, I do believe you also did a 4+ to hit for CC vs. moving vehicles, correct?

    How do you figure "dedicated CC vehicle killers"? Black Templar THDC Terminator squads all pack S9 powerfists (vs vehicles), so if a transport pulls up near them, I will walk over and whack it. Especially since I have a chainfist in the squad. Which raises their "DRPG" stat a lot higher than the shooting version.

  2. All that is nice and dandy.

    Now take an army or two you did already and show it again, this time with all the equations in plain view.

    I still think this system falls short of its potential.

    5 turn statistics are complete fail as possibility of entire army taking no losses entire game is null. That means that they need to be backpedaled to one turn statistics.

    Also mixing shooting and CC bemuddles things totally. As I see it there should be three pairs of one-turn statistics - dead marines, dead rhinos and dead landraiders - from shooting and from CC which gives six numbers:

    This allows for:
    1. Better source for comparisons as you can treat the numbers for shooting and CC separately
    2. No stupid caps resulting in arbitral numbers "cos it sounds ridikulus".
    3. No out of cosmos figures for an army that without losses both shoots _and_ CC enemy vehicles for five turns (sic!)


  3. It seems odd that you use 'Per Game' when talking about Vehicles while against Infantry you only count one turn. Is there a particular reason for that, such as being able to count things like Combi-Weapons and Hunter-Killer Missiles more effectively?

  4. No system is perfect. I like to say first that I do like getting the per turn stats and breaking out the CC vehicle hits from ranged.

    Lehcyfer, if you can't duplicate his numbers after this you should reread paragraph 2 sentence 1. Your hopeless.

    I think it would be interesting to see some of the vehicle statistics broken down into 1 turn stats. So I did. However if you make the same assumptions (no losses) for each army then you still get stats that can be compared.

    Your 3 points.
    1. I do like breaking down shooting and CC by turn sometimes.
    2. "stupid caps" are used in statistics all the time, you have to throw away outlying numbers. A 10 man nob squad with claws can make 40 str 8 attacks on the charge. Those stats "lie". It's still much easier to reach out and kill a rhino with ranged firepower. If that's what your looking for get your death star unit and forget stats.
    3. Again, we are comparing stats not simulating games. If you compare both armies at full strength for 5 turns it is a good comparison. Even if you had everything broken down by 1 turn what then? Are you going to give a handicap to the army that goes 2'nd for casualties?

    The purpose is not to simulate a game, it's to try to gauge each armies offense. The whole premise is defense relates to points. This system tries to give you the ability to measure the offense of the army.

  5. The caps are unnecessary when you split the CC and shooting - they do not "lie" then - the nobs really can do this damage, though most of it is overkill except when they can split attacks against several targets.

    Stats that combine shooting and CC at the same time are misleading. No army ever is able to shoot _and_ CC with all its units for 5 turns, even if we forget the no losses absurd.

  6. @Devjon Right, the 5 turn thing for vehicles initially was to take into account combi-weapons into account, but also to show bigger differences. If you look at a DRPG of 50 and a one of 60, that seems like a big difference worth noting, right?

    If you just did it by a single turn, it would be 10 and 10.2. Not really as informative that way in my opinion.

  7. 10/10.2 per turn is 50/51 per game. 50/60 per game is 10/12 per turn which is visible difference all right for me.

    On the other hand the per turn statistic gives me a workable estimate for my chances of pulling off what I want to do during the next turn while gaming.

  8. Once the game starts, the whole thing is somewhat irrelevant, but thanks for correcting my math.

  9. I don't have a problem with the way the stats fell out for now. Remember this is a first cut from a limited number of extreme examples (4-0 vs 0-4 at one tournament).

    If you don't like the "per game" part vs vehicles then see what you get with per turn figures. The way the numbers are rigged though you have to mathhammer kill "units" of MEQ per turn with each unit or penetrate individual AV11 vehicles per game. Since MEQ tends to come in groups of units that should be destroyed and moved on from, while you kill vehicles individually, I don't see a big problem there.

    @lehcyfer: I'm figuring it the way you did for my convenience, which is how I noticed that DRPG for THDC (Tank Hunter Dual Cyclone) Black Templar Terminator squads do kill 8.89 Rhinos per game by shooting, but they get 25 penetrating hits in CC per game.

    The question would be, is this a problem? Should I be worrying about a squad with 48" weapons getting in CC? Should I be calculating it even?

  10. I tried the six numbers, per turn approach on three kan wall armies from Mercer blog (the first one is Mercer's army, the other two are my tweaks) to see how statistics change with modifications.

    You can find the discussion here:

    The numbers below are for Mercer's Kan wall, my tweak with big shootas and my tweak with rokkits:

    DMS/CC ___ DRS/CC ___ DLRS/CC
    22,47/52,38 _ 4,24/23,07 _ 0,11/5,61
    23,29/51,65 _ 4,83/22,41 _ 0,10/5,61
    23,51/51,13 _ 4,72/22,35 _ 0,16/5,61

    The resulting differences are not very impressive, but nonetheless visible.

    You can analyze the spreadsheet under the link below to see the results and equations for separate units.

  11. Oh, I included glancing to death in destroying vehicles.

    @algesan: "Should I be worrying about a squad with 48" weapons getting in CC? Should I be calculating it even?"

    That's what buggers me and that's why I propose breaking the statistics into separate shooting and combat. Analyzing your army in the spreadsheet you can exclude the results you are not going to use.

  12. "The question would be, is this a problem? Should I be worrying about a squad with 48" weapons getting in CC? Should I be calculating it even?"

    I certainly wouldn't. Save the points on those chain fists. ;)

  13. i shouldn't have deleted the last few sentences.

    I've punched out vehicles often enough with those squads. Often in a countercharge situation where I could hit two things. More often I've shot a squad and then dove in to finish them off so they wouldn't be bothering my guys next turn.

    Now, what I have done is drop the DRPG(CC) stat from consideration for them. That way if I have 60+ in DRPG, the CC capability is a bonus to be aware of. Just like the fact that if 5 terminators with Preferred Enemy (AACNMTO) get to charge and land all their CC attacks, they wipe out 9+ MEQs.

  14. 15 is way too many tank kills for a CC unit per game a saner 'max' figure would be somewhere between 5-10 though my gut tells me 5 or six would be about right.

    I have deliberately done a certain 1750 list for ToS, I think competitively It's a bit wank, however ToS is hardly a competitive mecca so I have chosen to be a little more liberal with the list than I would usually allow myself.

    Would you mind running the stats on it and a spam list I would usually run and maybe give some analysis based on the numbers and your observations on lists thus far ?

    I am quite interested to see how much I may have 'handicapped' myself in the name of indulging my artsy side ....

    great work with this btw, I feel you are really onto something with this.

  15. It's 15 penetrating hits, not literally dead tanks.

    And I'm not doing people's list breakdowns for them anymore, sorry. I haven't the time.

  16. Fair enough I appreciate blogging etc can suck time, I'll live ;o)

    15 penes is more or less in line with 5 dead tanks , so yeah as 'Max' figure it seems fair to be set around there.

  17. I love this logic. Here are some of the thoughts that i had in regards to your main obstacles (being movement, range, and overkill).

    Movement- this one is tricky and will likely require some statistical modeling to get it right. But units should be compensated for their ability to move long distances. This would most likely serve as some kind of multiplier on the units metrics. Just an idea.

    Range- This should also be compensated for somehow. Again, statistics would be your friend on this. I think it is pretty obvious that a 6" melta attack is usually not as good as a 48" lascannon and that is lost in this system (I know you have covered this exhaustively), I think a multiplier system would be ideal for this as well.

    Overkill- You touched on it with your comment about Nobz penetrating Rhinos but it really applies more simply to shooting. I squad of 5 Fire Dragons can penetrate 3 rhinos per turn but really in 5 turns they really only have the potential to kill 5 rhinos (not 15). That excess would be considered overkill. I think that this would probably be best dealt with by capping that number somehow (maybe 1.5 times potential number targeted). This observations makes Long Fangs jump off the page compared to other heavy weapons squads.

  18. Actually, Holanes does mention something that I'm curious about. Why do you only have a cap on Close Combat attacks against vehicles? Is it because units only get enough attacks in Close Combat to be worried?

    Also, what about Flamers? Say, a regular Flamer compared to a Plasma Gun or a Meltagun or something similar, you are going to get higher numbers for the Plasma Gun and Meltagun than you are for the Flamer. But what about lists that take Flamers for Horde Control? It seems like the ability to kill Hordes is similar to the ability to kill Land Raiders: you don't compensate for weakness in other areas by being very good at it, but you can't afford not to be good at it.

    But since you're using MEq, a Plasma Gun will get 1.11 kills by itself (in the hands of a Marine) but a Flamer will only get .83 kills. So if you want to kill Infantry you will take the Plasma Gun. But against weaker Infantry (such as Orks and quite a bit of Tyranids) the Flamer is actually the better option (getting 2.5 kills rather than 1.11). Does the presence of bolters (or equivalent weapon) grant enough Horde Control to take Plasma Guns instead or is Horde Control unnecessary?

  19. My initial thoughts would be...

    Hordes exist, but they aren't nearly as prevailing as MEQ lists at tournaments. You have to bring enough to deal with them without gimping yourself against marines or vehicles. That much everyone knows.

    Frankly, you DO see more Flamers on the table top in actual tournaments than Plasma. So what gives? My guess is that people count on weight of fire to kill MEQ more than taking dangerous AP2.

    But look at the top NOVA lists. Typically they run one or none units with flamers. That says something about flamers and success. Do you really think that a 5 man Grey Hunter unit with a single Flamer is going to turn the tide against horde Orks or Nids? Ofcourse not.

    My conclusion is that those flamer inclusions came down more to points efficiency. Flamers might kill fewer MEQ than Plasma, but they usually cost less. Depending on the unit, a Flamer might kill more MEQ per point than a Plasma. Food for thought.

  20. Mm'kay, that does make sense. Thank you for answering my questions.

  21. So how is this worthwhile if "[o]nce the game starts... it's somewhat irrelevant"