Undead: The cost of wounds; stat discussion

Terribly formatted data

Hello Forum goers! This is my first post, and from lurking the boards for a while (not that a lot happens here most of the time), I am unsure if anyone enjoys this part of the hobby as much as I do, but I figured I’d try to start a discussion anyway. I didn’t see much other discussion on it that wasn’t wildly out of date (IE kow 2nd ed)

As can be seen in the linked google doc, aptly named for its general lack of readability, I was looking at what you pay in the undead army (because its the one I play) to deal a wound to your opponent. Obviously the killing power of a unit is not all you are paying for, but I found what you get for the more expensive wounds to be surprising in many cases.

For anyone looking at the document, I am aware that my rounding on the average wounds is wildly inconsistent.

A couple of things to note about the method here.

  1. I used whichever unit size had the best points per attack, as it will naturally have the best points per wound. In the case of a troop being less than 1 point difference to a horde, I used the horde.

  2. I did the math using any combat upgrades available to the unit, but no magic items or external auras. So skeletons and revenants have two handers, vampire lord has blood frenzy, etc. but zombie units are not benefiting from vicious for example.

  3. Mages got left out because obviously they are not costed for their damage and it would be fruitless.

  4. TC was assumed to be active for average wound calculation, and there was an assumed defense of 5+. Obviously things will change if defense is low enough to waste some of the CS or if TC is turned off.

Things that I noticed the most was the premium that gets paid for movement, and that upgrading to a regiment is almost entirely a defensive buff. I am in no way suggesting that movement isn’t worth a significant amount of points, just look at the price increases to mounts they decided on for 3rd, but it does mean in a lot of cases, taking the slower option is more efficient if they are just going to front line your forces anyway.

This shows itself best with soul reaver cav vs inf. You pay 10 more points, downgrade 1 CS to 1 TC, and double your footprint, for 2 additional movement. The offensive power is the same. This makes inf a significantly better use of points once the fighting starts. Although both of them are by far the most well costed wounds in the list. Which means if you are bringing a hammer that isn’t a bunch of great weapon wielding Vampires, you should plan to surge it into a flank, because that is the only thing they cannot do better than the other hammers.

Speaking of paying for movement, and good surgeable hammers, the zombie trolls vs wights surprised me the most for this whole thing. I see wights touted as the other reliable hammer in the undead army, and mv 7 with fly is for sure something to be valued, but as a flank surgeable hammer you get way more bang for your buck with the trolls, and that’s before adding the vicious aura. I had always been annoyed with the zombie troll unit because the regiment version is just a worse mummy troop in literally every single possible way. (which is never what I expect to see with KOW units honestly), and so didn’t look too hard at the hitting on 4s horde version over the hitting on 3’s wights. But the wounds aren’t massively different, as seen by what you pay for them.

The other thing I noticed is that combat characters cannot just run off and do combat. You pay way too much vs the offensive power they give. Which is probably why vampires never take the blood rage option, better to stick around for support when your wounds aren’t that point efficient. If only duelist worked vs heros and not just individuals… I’m pretty sure avoiding hero hammer was a design goal for Mantic, so kudos guys!

A comparison I’d also like to mention is vamp on pegasus vs Lykanis. I used to always pass up the Lykanis as the Pegasus was my 50MM fast flanker of choice, but with the new slayer aura virtually equalizing their attacks, and the vamp paying for the opportunity to pay for spells even if you dont take any (this irks me with the regular vampire lord too, paying for the opportunity to pay feels bad), The lykanis is really only losing 1 movespeed and flight, for a significant cost reduction.

I am aware this is twice I’ve mentioned “only” giving up flight for significantly cheaper combat power, but maybe, like me, you all hadn’t thought too hard about how much you are paying for it when it may not be necessary and could let you fit in something that is.

I’m really not sure how to feel about the revenants. I never liked any of the skeleton units from a mechanical standpoint personally, and this edition the math seems to support my gut feeling. The revenant points per wound is solid enough, but only on giant footprints, and for inf you have to sacrifice the 5+ def that make them attractive to begin with, and with Cav its only for the round you have TC. Both revenants and skeletons don’t seem to offer enough as shambling hammers, but otherwise seem worse than zombies and mummies as anvils. It almost feels like something is there for the revenants though, I just cant place what.

Shooting as both expected and unexpected is horrendous for undead. I was not surprised to find the archers to be the worst way to spend points on damage, but with how many people swear on bringing 2 catapults for consistency, it sure is a consistent way to pay a lot for wounds.

And finally on a related/not related note, since I hardly think a whole other thread is a worthwhile place to cry about it, I sure miss my zombies legions having 40 attacks. It still didn’t make their points per wound very good, but I am a fan of the bucket of dice approach.

For anyone that made it this far in this long, disjointed, ramble of a first post, thank you. I’ve kind of lost the opportunity to nerd out about KOW irl, and really wanted to drop my thoughts somewhere. Please feel free to add your own musings, tear my thoughts apart, or just have a good rest of your day.

6 Likes

Welcome to the forum and great post. Thanks for your analysis on the points per damage of the Undead army in Kings of War. Like you, I also find the data aspect of this hobby particularly fascinating.

Your approach towards analyzing the effectiveness of units based on average damage is useful, but I think an additional metric is even more useful: the percentage chance of breaking a unit. Although damage scales linearly, a slight increase in output can drastically alter the chances of routing a unit. This is the bricklayer vs gladiator approach, and with a piece trading game like KoW, gladiators matter a lot because they set up future beneficial trades.

I also agree that a unit’s combat effectiveness is not solely defined by its damage output or resilience. Factors that facilitate engagement, such as surge, flying units, or terrain mitigation, significantly enhance a unit’s combat potential. Conversely, larger base sizes can hinder a unit’s maneuverability and ease of engagement, even if the statline is otherwise good (looking at you Orcs).

I built out a python calculator to simulate combat between all the units for an elo table, and much of it aligns with exactly what you’re saying. Not necessarily an intuitive way to think about it but useful in list building.

2 Likes

There was a thought in the back of my mind about how much damage is the breakpoint any particular unit role needs for maximum efficiency. After all, 8 wounds for 300 points vs 7 for 100 doesnt matter if 8 is what breaks the enemy unit.

I would definitely be interested in what the defense/nerve matrix looks like across the armies. List building to make sure you hit certain wound thresholds could carry some weight over more point efficient wound maximization.

I did this just looking at the faction I’m most familiar with. Looking through the defenses of all possible enemies sounds like quite the undertaking :sweat_smile:

1 Like

I have a metric I scripted last year called shots to 6–how many no piercing shots a unit could take until it was at 6 nerve remaining, which is about 50/50 to rout with inspiring. I’ll see if I can calculate that for the various factions. What would be the best way to present that? Seeing 30 units with that data in one faction isn’t useful because you’re not building a list from 30 types of units. Let me know what you think and I can calculate it.

Off and on I put together a blog post to look at the data behind KoW: https://dataanddice.com/abyssal-dwarf-army-summary/ latest quick article looking at one faction and cost effectiveness. I’m always looking for ideas for what would be of interest to the data folks out there.

1 Like

Well that blog is just a fascinating read. You’re a war general after my own heart.

The % to kill def 4 and def 5 is the chance they rout those def and nerve values in 1 turn correct?

I otherwise cannot say I fully understand what the ELO and Cost effectiveness is measuring, or how the simulations would offer more data outside of eventually just landing at the averages anyway.

Not sure regarding the shots to 6 stats. The odds a non piercing unit is being taken for its ability to rout units seems non-existent, so i dont know how much that data would apply to… maybe I am misunderstanding what you are proposing for that though.

Edit: After reading more blog posts, I have figured out what the ELO rating is measuring. Still not 100% sure what simulations are adding over just using averages, but maybe i just gotta keep reading lol

1 Like

Elo and cost effectiveness are tricky to describe in short form. Basically for Elo, here’s the overview: https://dataanddice.com/updated-combat-simulations-for-cok-2024-crushing-strider-and-brutal-still-matter-a-lot/

For cost effectiveness, I ran a dual method analysis: https://dataanddice.com/decoding-unit-efficiency-in-kings-of-war-a-dual-method-analysis/

On simulations vs average: it’s hard to calculate averages with all the variables. And moreover, the spread of potential damage has a bigger impact on whether a unit can rout an enemy than on the average damage. Eg a unit that does 1 damage on average may have a 1% chance of routing a typical opponent, while one that does 3 damage has a 40% chance. Is the second unit 3 times as valuable in combat or 40 times? Simulations with Elo try to capture that nuance more than just the average damage.

For the “shots to 6” metric: it’s not actually considering shooting units off. It’s a proxy way of comparing, say, a Def 5 15/17 vs a Def 4 20/22. It gives you one metric, which is how many hits a unit can take before it has a 50-50 chance of failing a nerve test. I find it to be a useful resilience metric.

Interesting read/discussion.

Touched on briefly above, you’ll find plenty of armies geared to be able reliably take out the de5 15/17 type unit, so often the metrics are done on that basis - which potentially is a different build style to those aimed at taking out the de3 20/22 junk horde spam lists.

Movement/terrain mitigation are obviously key as will determine whether you actually get the charge off and how effective that can be.

The Rare Monster designer in the Legendary rules is an interesting guide to how/what rules & abilities effect price and how things are costed

I used the word “average” and what I really meant was “odds” and that’s on me. We have all of the numbers to determine the % chance a unit breaks another unit before being broken by that unit, and while something like 1.25 million :wink: die rolls is fairly likely to settle around the same %, it is in theory unnecessary. (but admittedly much cooler than boring old math)

The only real variable, is “who struck first?” and while the simulation did add something to account for this, its probably the most complicated unknown in reality, with things like terrain, surge, nimble, and wild charge making “highest movement usually goes first” only applicable to a handful of unit pairings.

As far as “shots to 6” it sounds like we are looking for the odds to be dealt wounds at each Def value, and the odds that each wound dealt breaks them. A cherry picked sampling of various def/nerv units regardless of army is probably the least cumbersome way to do it. Or even just setting the stats without actually tying it to any particular units.

Welcome to the thread!

The potential difference in optimal offensive units between de5 15/17 and de3 20/22 is why I initially took a look at points paid per expected wound. It is definitely a flawed metric in many ways for list building.

If you have any thoughts on how to list out the optimal offensive loadout to “reliably” take out a variety of def stat spreads, we can then determine which units get you there for the least investment, with the smallest footprint, or with the most movement. Which is definitely something I’d be interested in, since it would even help address the movement and terrain variables.

I dont think I’ve seen the rare monster designer, I will have to look for that. Thank you.

1 Like

You’re absolutely right about the simple math behind calculating combat outcomes in Kings of War. That said, while the individual calculations are simple, combining all of those odds into one ends up with a tricky mess. What we’re dealing with here is a classic example of a situation that’s perfectly suited for a Monte Carlo simulation approach.

In essence, while the underlying mathematics might be straightforward, the brilliance of the Monte Carlo simulation lies in its ability to make sense of the chaos and complexity of Kings of War without having a massive equation to test for each situation.

For the “shots to 6” metric, here’s how it works: This metric helps us figure out how many hits a unit needs to take before it reaches a point where there’s a 50-50 chance of it breaking if it’s inspired. Let’s compare two units:

  1. Unit A (Def 5, 15/17):
  • With a Defense of 5, this unit will be wounded on 1/3 of the hits it takes.
  • To reach a Nerve of 6 (from 17), it needs to accumulate 11 wounds.
  • Since only 1/3 of hits result in wounds, it takes 3 times as many hits to achieve those wounds. So, 11 wounds x 3 hits per wound = 33 hits to reach a Nerve where there’s a 50-50 chance of breaking.
  1. Unit B (Def 3, 20/22):
  • With a Defense of 3, this unit will be wounded on 2/3 of the hits (easier to wound than Unit A).
  • It needs 16 wounds to reduce its Nerve to 6 (from 22).
  • Here, since 2/3 of hits result in wounds, it takes 1.5 times as many hits to achieve those wounds. So, 16 wounds x 1.5 hits per wound = 24 hits to reach that critical Nerve point.

For comparison, a Def 4, 20/22 unit would need 32 hits to break (because it would be wounded on half the hits), which is functionally equivalent to a Def 5, 15/17.

What this tells us in terms of strategy and list building is quite interesting. For instance, from an attacker’s perspective, a Def 5, 15/17 unit is just as resilient as a Def 4, 20/22 unit because both require the same number of hits to reach the ‘break’ threshold. This insight can be crucial when you’re planning what units make a good ‘anvil’ or deciding how to allocate your attacks in the game. I use it as a shorthand to evaluate the staying power of various units. In general, it’s fine to ignore crushing/piercing because they only tend to matter if they would exceed the 2 to wound–so trash armies waste crushing, but everything else is functionally a wash.

I then use a shots to 6 vs points ratio to get to what you’re getting at from a listbuilding perspective.

2 Likes

This is a good example of why I initially thought about points paid per attack as well, even if they don’t necessarily translate to high average wounds. A unit being likely to break after 24 hits (example 2) means anything with less than 24 attacks cannot hit that threshold no matter how well they roll.

So although the odds of translating those attacks to hits, and then to wounds, is lower, there is an outside chance that a higher number of total attacks unit breaks a unit. While the elite unit had no real chance at all, due to the unlikelihood of the nerve check on the maximum hits they could even achieve. Another reason I miss my 40 attack zombie legion, it was always low, but they always had a chance!

I am not sure how well that translates into the data, but maybe this is the type of thing you are talking about that makes the simulations make our lives a little easier.

2 Likes