Some Assault unit mathhammer

I got bored tonight because my sister claimed the TV and kept watching Jamie Oliver, so I decided to do some mathhammer on all the assault units that are available to put in a raider, figured I'd share

a few things stand out:
1) Grots are awesome
2) Sslyth really get their value from their shooting, it makes them very comparable to grots
3) Incubi are comparable to Sslyth and Grots, albeit a bit more tanky and less killy against non-TEQ (which you shouldnt really be fighting anyways )
4) Sslyth and Grots are quite good
5) Wyches are almost perfectly equal to Bloodbrides, apart from the point lost when they die
6) Harlequins are quite good, if you have a method to avoid the overwatch
7) Did I mention that I REALLY LIKE SSLYTH AND GROTS?

Notes:
All stats are given on a per-point basis (ignoring the price of a raider)
Potential/Point = Attacks/Point (i.e. max damage)
S4 Tanked/Point = S4 Shots Tanked/Point(see below)
I assumed charging (which is to be expected when using a raider!), aswell as rampage for the grots
I assumed S4 AP- shooting for tanking potential, if you take melee mandrakes suddenly suck and wyches are still awesome tarpit units (wych is what everyone seems to use them for anyways), although still slightly inferior to wraithguard
I probably made some calculation error somewhere, today was a long day
GEQ = WS3 T3 5+, MEQ = WS4 T4 3+, TEQ = WS4 T4 2+/5++


EDIT:
Will probably do Scorpions and Eldar units with exarchs tomorrow, too much work to work everything out seperately today

Interesting that mandrakes score high and require no transport - also they have shrouded and are failrly durable

Yes, especially if you remember that they also have a potent shooting attack, I find them very comparable to wracks

EDIT: My personal favorite here is the harlequins, extreme damage and super fragile ( seems more like a Dark Eldar unit there GW!!!), I had expected them to perform worse, but a storm of rending attacks as well as FC is pretty potent

I'm finding the data to be oddly presented since all values are a 'per point' ranking, yet basically all the units are using different point builds.

Why not just do units of 1 model of each type so it's a bit easier to see how each unit choice actually stacks up against each other in a theoretical value basis? For instance, I'm intrigued by the Sslyth numbers, but then I look and see they're the most expensive unit in the mix - meaning that I couldn't opt out of taking Grots and take Sslyth and expect similar showings. Also, just as a random quibble, a 5 man Grot squad is pretty unusual, as is a 10 man Incubi squad

I'll also admit I would have liked to see a shooting v. assault damage breakdown - since you indicated you included shooting it's a little uncertain where the actual benefit is coming from. For example - should I even bother assaulting with the Mandrakes? Or will I get 90% of the effect simply with shooting? Et al.

Also, as far as 'tanking' goes did you calculate in multiple wound models or no? Because a Sslyth, say, could take three wounds, armor save one, FNP another, and suffer a third and still not be particularly negatively effected as far as damage output goes.

Thor665 wrote:

Why not just do units of 1 model of each type so it's a bit easier to see how each unit choice actually stacks up against each other in a theoretical value basis?

It's because you have to find the mean value of a unit rather than its individual models. Judging the unit un-upgraded can be misleading because sometimes the loadout can make a unit more efficient than different units, and when you try to find the most optimal loadout for a unit, you have to compare it to every other loadout in the same unit to "prove" it's better than the rest. If we try to prove the best loadout for every single unit, it starts to look a lot like a Sierpinski triangle.

It would be very difficult to find the best loadouts for the entire codex. You'd need to divide the work and have people check it.

Which I would love to help with...

You can largely ignore the point values&amount of models per squad, since all the squads are homogenous, so 10 incubi deal double the damage and cost double the points of 5 incubi, therefore the amount of damage dealt per point invested is actually the same for both squad sizes, they were more meant as an indicator for me as to how many points it would cost to fill a raider with them. (mind you, 10 sslyth is also pretty weird, they have Grot LD)

As for the shooting, I did not actually factor it in, just stating that if one were to factor it in mandrakes are actually quite good(a little more point efficient than wracks, in fact) and sslyth are slightly superior to grots when shooting before assault, IF taking the same amount of points of them (so 7 Sslyth are slightly superior to 5 grotesques etc.), seemed more reasonable to me as it makes for actual comparison, considering points are what most (all) players use to limit their army selection. Ill try to add the shooting to it aswell though (but first I gotta get some sleep, 11 hours of college takes its toll )

As for the multi wound models, I simply stated the amount of S4 shooting hits required to fully clean them up, as acounting for all the possible variables in a reduction in squad damage output is impossible (you cannot express all variables mathematically, such as for example different squads firing leading to several models being wounded yet none killed)

Also, I took what I recall to be the most point-efficient loadouts (for example a wych is more point efficient than a wych with special weapon, and grot liquifier guns arent helping melee in any way)

TL;DR
Comparison holds only for equal points invested in a unit, shooting is NOT considered, sslyth > grots

Okay, I get it now, I was reading it wrong.

So, functionally, what we're really looking at is four charts.
The bottom right chart is functionally meaningless unless you're wondering the relative unit cost to fill a Raider.
The top left chart is basically a chart of what is tougher to kill...and you say per point, but the odd squad sizes have to be part of that considering that your end metric was 'wiped' therefore making it kind of useless, again, unless you're running those squad sizes, and also utilized AP- S4 weaponry.
The other two charts track a, functional, per model point efficiency for inflicting wounds against three model types. So, in a way, anything that scores well on all three metrics there is amongst the most point efficient assault models available to us.

So looking at the top 4 in each;
Guardians, Grots, Sslyth, Mandrakes, and Bloodbrides rock vs. GEQ.
Wraithblades (sword), Grots, Harlies, and Banshees vs. MEQ.
Grots, Wraithblades (axe), Incubi, and Harlies vs. TEQ.

Yeah, I'd like to see shooting as a possible metric track.
Feel very good about my Grot favoritism in TAC builds.

What choice did you make for Rampage attacks generated when crunching the numbers?

Yeah, those units appeared to me as the top units aswell, so I figured I'd do shooting for them, Adding Sslyth because I think they are good replacements for (if not better than) grots: If you include pre-charge shooting, they are more efficient than grots for their point cost, also, their sheer volume of attacks makes them pretty good against terminators!

For the grot rampage, since I figured no one would ever use huge squads of grots anyways, I took the average of bonus attacks, so it is worth mentioning that grots aren't so awesome if you don't get rampage!

But yeah, in the end I might agree with you in that grots are in fact better generalists than sslyth (especially considering the fact that grots can ID!), sslyth have an edge on the charge when you can include the shooting, but you end up being in combat for more than a turn usually! also, you probably wouldn't be assaulting TEQ anyways, because they hurt

Grots also have the benefit of PfP - which means as they game goes on Grots get better in a number of ways depending on the chart they're feeding off of. They also have the option for a character upgrade, which is also quite useful if you're using them for escorts (which is what you'd be doing with Sslyth also)