Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

Wehrkind wrote:(near-death star?)

Hubert Farnsworth approves.

Wehrkind wrote:
How did you approach the issue? I was thinking making unit size a function of attacks, str, ranks and cost with some sort of mitigating factor based on vulnerability to certain spells. So to put it another way, Size is a function of expected CR per point and Steadfast expectation minus vulnerability.


Troubles I would see off the bat are:
1) Combat res vs what? Is this going to vary on opponent? If so, does a proper death star plan require you load for bear assuming you will mostly be fighting dogs? Or do you figure out optimal sizes vs specific armies, and then pick the size that best fits the armies you expect to face?
2) I don't know how to incorporate base size
3) How do you put in steadfast to the equation
4) How do you account for vulnerability? I was thinking the expected points loss per casting of the big killer spells, with each having some factor attached to it. If you then changed the factor up for a spell being more common (we have 20 Life mage players in our league etc.) or down for less.
5) Just thinking of how to build that equation is making my brain hurt.

Once it all worked out. It was a pretty simple realization that there was a metric to measure a unit by.

It ended up being a numerical analysis. The trick is: You simply measure 100 of the testing units attacks against a Ws3 T3 4+as foe. This was their offensive output. Then you measured the idea of 100 attacks coming in from a Ws 3 str3 foe. This was their defensive capability. The final number was the understanding that you had to add the number of ranks, banner, and outnumber to the equation. It was a pretty elegant system where you could tell a unit was worthwhile if the number of wounds caused minus the number of wounds suffered plus your ranks and other bonsues was greater than 0. The higher the number, the better the unit. The lower the number, the worse the unit. (a LOT of units were mired in negatives). Goblins rocked at a ~-5 and Chaos a chaos warrior rocked at a ~+29

Label the units from lowest points cost to highest points cost, and the disparity became obvious immediately and abuseably. Then the exercise came down to whether or not it worked on the flanks. Most units with a unit in the rear, or in their flanks were auto-breaks. Only the units that were at a +15 before penalties managed to stick around (in fact +15 was the BE point.)

This system still works, but steadfast throws a whole new wrinkle in the system that his making me re-think the whole exercise.

Unit killer spells haven't really been factored, unfortunately.

Ahh that's a good system! It would be pretty easy then to to make a similar spreadsheet for say WS5 Str4 opponents or whatever to help balance against the stronger enemies, where AS benefits give way to other defenses.

How does your system account for unit sizes being better/worse? So whether 50 gobbos was more efficient than 40 or 10? I have a feeling it is in there somewhere, but I can't tell what part suggests more optimum sizes.

Thanks Wehrkind!

Once you spreadsheet it out, you realize pretty quickly that you don't need to tabulate multiple combat situations, because it all scales equally, (until you get to t7) in 7th ed, and now, it all scaless 100% equally.

The percentage in difference between attacking ws5, t5 as 2+ with Goblins and with chosen warriors was the same as attacking ws2 t3 no armor save.

There were a couple of other considerations: 1 the system broke after you capped out at +3 rank bonuses. So you never needed to really go after "optimim unit size".
In fact, 20 was just about always perfect until the HE's came around. Units died so easily that you basically had to do a CostBenefitAnalysis between whether or not you thought the +3 was worth it. Unless you were unbreakable, or stubborn.. it very rarely was. So you're right, the consideration for an "optimal unit" is in there, it's just not solvable.

For large bases, 1 attack base made the model overcosted 100% of the time. It made 2 attack base models undercosted 100% of the time. Had marauders been on normal sized bases, (as originally intended) they would have been about perfect for cost. Both Marauders, and Orcs were brutally overcosted for their base-size. Especially since you lost a full 12% effectiveness from the unit. You still lose that effectiveness, but it's not as impactful since far more attacks are coming out of the unit.

Unit to unit, optimal sizing seems to be a function of having exactly 1 model left at the end of turn 7 after 4 turns (8 phases) of losing combat. Which is to say: Somethign that is going to be very matchup dependent.

Wehrkind wrote:

Ragnar4 wrote:
Troubles I would see off the bat are:
2) I don't know how to incorporate base size
3) How do you put in steadfast to the equation
4) How do you account for vulnerability? I was thinking the expected points loss per casting of the big killer spells, with each having some factor attached to it. If you then changed the factor up for a spell being more common (we have 20 Life mage players in our league etc.) or down for less.

Steadfast is a bitch to deal with. It depends on the size of your opponent, which makes it tough to look at a single unit with pre-set conditions.
Base size can give you more attacks, but the ability to reform in combat means that attacks thrown out will vary.
How about monstrous infantry that throws out more attacks form the 2nd rank and stomps? The old method of make 100 swings and take 100 hits doesn't really tell you how it will perform on the table, or if one unit is any better than another in 8th.
How do you deal with units with 1st round advantages (choppas, flails, charging lances), when most combats go more than a round?

-Matt

Hmmm that's interesting! So the strength of attacks coming at you didn't really matter, even though it would negate your armor?

What I am thinking of is say the difference between Marauders in Light armor vs Marauders with MoT for a 6+ ward. At Str4 the Marauders in Light armor should feel the 16% change in survivability twice: once for wounding on 3's and once for losing their save, while the Tizz marauders only feel it once as their ward doesn't change. It seems to me that should matter, as it would make cheap models with little armor more cost efficient vs strong things than say WoC. Gobbos' ability to die violently peaks out at Str5 while WoC don't plateau until Str7 (8 if they have shields), though after 6 they stop getting easier to wound.

See, I SHOULD be studying for my micro-econ final now, but it is all stupid math, so this is more interesting, and sorta math

LOL. 90% of this crap is rooted in Micro Econ. IT's where I got the idea for most of my unit to unit analysis.

Funny how that stuff begs to be used for something other than econ itself... never seems to pan out there without some rather excessive assumptions. (I am working towards a PhD to teach, and the way the concepts are approached, or skipped to go straight to the math, makes me insane.)

We probably should start a thread on this and knock ideas for it back and forth. It -should- be solvable on a "from the front/flank/rear" standpoint at least, with optimal (or sufficient) zones of effectiveness for a given model/stat block vs a given enemy. If one could come up with curves for various troop types vs various enemies one could then build units with specific uses in mind, or understand better if a unit size is just not good for much anything. Plus the math would be interesting, and I could use the help with it