The Witch King only needs 10 Will: Part 2 of How to Build Your Witch King

How many Will and Fate does this guy need?

The Witch King is one of the most powerful and ubiquitous models in the MESBG competitive scene. He’s the centrepiece of almost all Angmar and Mordor builds, and is the particular star of the Witch King/Suladân lists that have been taking the world by storm for the last year or so.

Last week we looked at the three main variants on this model, and the different lists into which each might fit. As part of that, we concluded that a variant on horse with the Crown of Morgul is the most generally efficient, and fits best into most lists. That conclusion is reflected in competitive results, with most lists that include him giving him these equipment options.

However, I think most players are building him wrong (or rather, inefficiently). To be clear, that’s not referring to the horse and Crown; those are excellent purchases and should almost always be taken. It's also not referring to the 3 Might, because that's obligatory. Instead, what I’m referring to is the Will and Fate purchased. Most players seem hesitant to leave home without at least 14 Will and 3 Fate, with many players going higher still (on the Will; I haven’t seen anyone cheating by taking more than 3 Fate).

In contrast, I think that 10 Will is actually totally fine, and the optimal number might even be less than that if it was legal. This conclusion is based off intuition and analysis, but also data, and it’s that data that we’ll be focussing on today.

The data: 26 tournament games with the Witch King

The Witch King neutralises Elessar to win me a big Queensland tournament

As the header says, I’ve taken the Witch King to 26 tournament games. That includes 4 games with a 90-point budget Witch King (3/10/1 on foot without the Crown), 11 with a 140-point standard Witch King (3/12/2 on horse with Crown) and 11 with my current favourite build (3/10/2 on horse with Crown). The 11 games with the 3/12/2 were variants of a Witch King/Gûlavhar Angmar list, while the other 15 were different versions of Witch King/Suladân.

I’ve also used the Witch King in far too many practice games to count, at a variety of levels and in a variety of lists. For several reasons, I’m not incorporating these into my data today. Most significantly, I don’t have a handy written record of them all, like I do for my tournament games (being a blogger does have some advantages). That makes them much harder to analyse, and means I’d almost inevitably forget a bunch. And secondly, these games tend to support the same conclusions as the tournament data, just with more haziness.

The Witch King shuts down Helm with plenty of Will to spare in a practice game

That’s the data itself, now let’s move onto the analysis.

First analysis: how much Fate do you need?

This analysis is pretty simple: how many times did I lose my Witch King across the course of the 26 games (in which I had 2 Fate in 22 games, and 1 Fate in 4 games)?

As it turns out, I only lost the Witch King once across all of those games. In my fourth game at the Queensland GT, I had charged my Witch King into a random Mahud Warrior. I was then countercharged by a Camel, dismounted by the impact hits, and then absolutely skewered in a single turn of combat. It was pretty embarrassing for the Witch King, although he had already won me the game by killing two Mahud Kings in a single turn. I think that I was Wounded twice and saved one with Fate, so it’s possible that a third Fate roll would have kept my Witch King alive. I don't actually have a record of this, however, so I may have saved neither with Fate (in which case a third Fate would have been useless). 

The Witch King having a great game, right up until the moment that he doesn't

I also had one game where the Witch King was Wounded by a Ranger, and saved it with Fate. That was actually the only other time I have spent Fate on my Witch King across 26 games.

These numbers lets us estimate the value of purchasing an additional point of Fate by estimating how many times it would have saved the Witch King’s life. Doing so gives us a figure of 0.5 Witch Kings saved over 26 games, for an average of 0.018 Witch Kings saved per game. Multiplying that by the cost of the Witch King is a bit of a fudge (it doesn’t take into account the fact that the Witch King would have only had a limited number of additional turns of game time if he had survived, nor the VPs that were awarded for killing him), but if we do we get a figure of 2.32 points worth of value per game.

Overall, that’s not great value. It implies that for every 5 points you spend on Fate for the Witch King, you’d have been doubly better off if you’d taken something else instead. I’d be wary of taking this too literally, but it at least supports the idea that 2 Fate is probably fine and you don’t need a third.

 

Second analysis: how much Will do you need?

This is the much bigger and more controversial topic, so let’s dive right in.

The first question here is how often I ran out of Will and lost my Witch King due to the Will of Evil special rule. The answer to that is pretty simple: never. That’s actually true of both tournament and practice games; I’ve never lost my Witch King to running out of Will, nor even to being killed by a magic power that he couldn’t Resist because he was down to 1 Will. People assume that it’s a serious threat, but I’ve always found it relatively easy to avoid.

In most games (like this one), I ended up with 2-4 Will remaining

Of course, that’s only half the story: there’s also a question of how often I had to stop casting or fighting in order to avoid running out of Will (and dying). If my Witch King spent the second half of every game cowering in a corner, then it wouldn’t be much of a boast that he didn’t technically run out of Will.

Out of the 26 tournament games, there has been one in which I had to stop casting or fighting earlier than I would have liked. There have also been two more in which I might have benefited from an additional Will point in the final turn, but didn’t actually have to stop casting or fighting.

Of these three games, I lost two and won one. The game I actually had to stop casting or fighting was against Nilesh at Clash, in which his Elves slowly ground my troops out from within his fortress. I nearly ran out against Vaughan at Dagor Daggorath– in which his Uruk-hai also hid in a defensive position and ground me out, while my Witch King magically duelled Saruman– and against Andrew Coleman at Clash, as my Witch King hurtled around the board to Compel forward and kill the Crébain holding the Relic.

The Witch King really wanted to have an extra Will in this game

Of the three games, two of them involved enemies holing up in defensive positions and playing long, slow games (in that the casualties came slowly, not that either player played slowly). And all three involved big, high-importance spells being thrown by the Witch King (a critical Black Dart to kill a banner against Nilesh, various Your Staff is Broken attempts against Vaughan’s Saruman, and Compelling forward the Crébain against Andrew). In all three games, I would have liked an additional Will point in the final turn.

And the other 23 games? In none of them was Will really an issue at all. There were games in which I technically finished on a single Will remaining, but they were because I’d thrown 4-6 Will at my final spell because I knew it was the final turn. In all 25 of these games I could have gone at least a turn or two longer and been able to continue using my Witch King as desired.

I ended several games at Clash with over half my Will remaining!

One more complication here is that I probably could have used more Will in some of these games if I had had it. If I’d had 16 Will on my Witch King in all of these games, then I could have thrown 2 Will instead of 1 at a few Transfixes, or 3 Will instead of 2 at a few Black Darts. It’s likely that some percentage of these casts that currently failed or were resisted would have instead gone off. However, it’s also likely that most of those Will points would have been wasted, either because the spell went off anyway or because the extra die wouldn’t have helped. This also is only likely to be a substantial factor in the games where I ended up close to the wire on Will, which was a very small percentage of the 26 overall (around 11%). And finally, this is also only relevant for the low priority spells: I was already throwing lots of dice at the spells I really cared about, so where it impacted would have been the ones I was willing to deprioritise.

In summary, having additional Will on my Witch King would have had the following effects across 26 games:

-         A couple of turns of additional casting or combat in one game;

-         A more reliable final cast in two games (one of which went off anyway);

-         Some small number of additional low-priority spells going off at some point in the 26 games (hard to estimate, but likely only 2-3 additional spells across all the games).

In exchange, I would have had 15-30 points less in my army, depending on which builds we’re comparing. That translates into 3-6 less models, depending on whether they’re Orc Trackers, Black Numénóreans, or Angmar Orcs.

The 600-point version of this list really doesn't have the numbers to be dropping models for Will

I think it’s quite clear that in the 25 games in which I did not run out of (or have exactly the right amount of) Will, I would have preferred the 3-6 extra models I did have. That’s also true of both the Isengard and Dunland games where I nearly ran out as well, as in neither game was the last spell especially impactful.

Finally, would I have preferred to have a couple of extra turns of combat or casting in the game against Nilesh, rather than the extra few models? I’m not sure, but probably. The extra magic could have shut down Déorwine, or the combat could have gotten me the couple of kills I needed to Break Nilesh. On the other hand, with less models to spare my battleline would have likely killed a few less models itself, and may also have quartered at the end (putting victory solidly out of reach). I think, overall, that I would have preferred an additional 4 Will over the extra few models here, but it’s very close. So that tells us that in 1/26 games I would have preferred to have additional Will rather than the extra models.

It would have been really nice to keep fighting and casting with the Witch King this game

An alternative way of looking at this question is how much average value I would have gotten out of having the extra Will in each game. As a fairly gross oversimplification, I think that having an additional 4-6 Will would have translated into an additional 3-4 turns where an enemy combat hero was Transfixed (based on the table above and how many extra casts I would have had in total). Being generous, that might convert into about 7 additional casualties on my side, or 1 kill for every 4 games. That roughly implies an average value for the extra 4-6 Will of around 2 points per game, which is less than we got from a single additional Fate point!

That gives us 2 potential metrics for evaluating the benefit of 4-6 additional Will:

-         It would have given me more value than its equivalent points in warriors in 1/26 games; or

-         It would have given me an average of around 2 points per game in value.

Overall, that’s not a great innings for the extra Will. It seems to have averaged a very poor return on investment, and been the inferior choice in the vast majority of games. That first stat is also somewhat misleading; I think it would have been narrowly better in that single game, while clearly being much worse in the many games where I finished with heaps of Will floating around. In fact, from this analysis it actually appears that 10 Will may be too many, and that the most efficient number may be somewhere around 8. Of course, you can’t take less than 10 Will, but the stats here certainly don’t speak in favour of taking more than the minimum allowed.

Caveats and counter-arguments

'Nice argument, but I am no man'

There are a few possible caveats to this. Firstly, 26 games is still not that big of a dataset. It’s supported by the many, many practice games in which I have used the Witch King (from which I can remember a small handful of games in which Will got tight), but the stats I’m actually using are capped at 26.

You could even justifiably reduce that figure if you thought that the tournament to which I didn’t take the Crown or the 11 Angmar games (in which I took 12 Will) weren’t comparable enough to be included. I’d push back on at least the Angmar point, but even if you did exclude them it doesn’t really change any of the conclusions: more Will being the better choice in 1/11 games still isn’t a great showing, and the average value would still hover around 2 points per game.

If anything, the Witch King's Will points were in even hotter demand in my Angmar lists

A more reasonable concern is that I’m basing these conclusions off counterfactuals, i.e. what would have happened if things were different. Counterfactuals are inherently hard and messy, because they didn’t happen. Perhaps if I’d had 16 Will in all of those games I would have used the Witch King more aggressively in combat, or thrown more Sap Wills early, or even fitted in extra turns because of my reduced model count. These scenarios are certainly possible. Equally, maybe my lack of models would have stopped me taking advantage of opportunities, or seen me Break a turn earlier, or denied me the model count on a critical objective. These scenarios are very possible.

It is also possible that in a much more magic-heavy meta than the Australian one, additional Will could be more viable. That’s very possible, as one of the only games where I did come close to needing extra Will was against Saruman (out of 9 tournament games against spellcasters, or 6 if you exclude lists with token magic like Thranduil, Elrond or a Shaman). I’d still note that I ended the vast majority of my tournament games with way more Will than necessary, and also that in most lists the Witch King is not the primary target for enemy magic (that’s generally Suladân or Gûlavhar, who are worse at Resisting, more fragile, and generally scarier to boot). But potentially if your meta was very magic-heavy and the Witch King was likely to be copping the brunt of that, then I’d see the logic for a few more Will.

Saruman kills my troops in combat after the Witch King 

And finally, this discussion is specific to the Witch King on foot or on horseback, because I don’t have any tournament games with him on a Fell Beast. If you wanted to take the Fell Beast variant, then I think 14 or so Will would be a much more justifiable pick. You’re likely to be spending more time in combat, you’re a more tempting target for magic, and it’s a bigger deal if you had to stop casting or fighting. I still think 14 is probably fine, but I don’t have a dataset of tournament games to draw conclusions from here.

Overall, I think the data is surprisingly clearcut. Having additional Will just doesn’t come up that often, and you’ll generally be better off without it.

Why it works (without data)

I won’t go into too much detail here, because I’ve already covered this topic in many tournament writeups. But in short, most games require less from your Witch King than you might expect.

Firstly, games are shorter than most people think, especially at tournaments. Time is a really relevant factor, especially when you’re facing opponents who aren’t used to all the tricks available to Mordor or Angmar. If your opponent has to spend an extra fifteen minutes on each of their turns thinking about what you can do with your magic or your Spectres, or what will happen if they fail that Terror check, then you may end up with much shorter games than you’d expect. If the game only gets 4 proper turns of combat, then 10 Will is way more than you really need. Hopefully you’ll get more than that, but it depends a lot on the tournament structure and opponent experience.

With all the tricks of Angmar and AoL to consider, this game did not get many turns!

Secondly, many games are practically over long before the dice are put down. I generally find that most games are decided in the first three turns of combat, and only exceptionally close ones are much impacted by the following turns. If you can get a critical edge in those early turns then whether you run out of juice in the lategame doesn’t really matter. In many of those 26 games, it wouldn’t have mattered if I had run out of Will because the key enemy pieces that the Witch King had been focussing on were dead by Turn 3, so I didn’t really have good targets to keep casting at.

Thirdly, as discussed in the last article, combat and casting are often incompatible. The Witch King is phenomenally versatile because he can switch from one to the other as required, not because you can expect him to do both at once. He can, in a pinch, but if he starts a turn on the frontlines because he charged last turn then he can’t reliably cast this turn (because the enemy can charge him first and tag him). And, as mentioned in that last article, he’s a F5 army leader with 1 Wound; you don’t want him on the frontlines every turn! So budgeting for fighting and casting every turn is unnecessary, because you’ll rarely do both.

Spend enough time on the frontlines and eventually your Witch King will get killed by a random camel

And finally, you don’t always need to be casting. If your opponent’s key pieces are out of combat or shut down this turn, and there aren’t any juicy unattended banners wandering round, then generally the Witch King won’t get much benefit from casting. In my game against Damian at Clash, my Witch King ended up not casting for three turns because the Mouth kept Transfixing Aragorn and there weren’t any other good targets around. I had the option to be throwing Black Darts at random warriors I suppose, but ultimately I was better off just holding onto the Will in case things went wrong later. They didn’t, and I finished up the game with heaps of spare Will.

These factors (plus the Morgul Crown, which makes 1-die casts quite reliable for low-priority spells) all come together to heavily mitigate the need for more Will. The thing I want to emphasise here is that it’s not a matter of being able to get by with less Will, but rather that more Will wouldn’t help in most circumstances. There are always exceptions, but it turns out that those exceptions occur about once every 26 games.

Wrap up: the maths is messy, but the answer is not

I’m a fairly stingy person in real life, who needs a lot of convincing that it’s okay to buy that coffee or replace my painfully-unfashionable coat. Those tendencies have followed me through to MESBG, and have definitely influenced my decisions on how to equip my Witch Kings.

Thankfully, in MESBG I appear to be more correct than in real life (although I’m still holding onto my ugly coat). Witch Kings really can get away with the bare minimum of Will and Fate and be just fine, at least for 25/26 games.

That 26th game was admittedly a Big Sad

If you’re a dedicated Witch King player, I’d really recommend you try out the 10 Will option. I’d love to hear whether it turns out to be an issue for you, and in what percentage of your games. More data would be much appreciated, particularly to test how much playstyle makes a difference here.

Next week we’ll be back to wrap up this article, with a data-fuelled-deep-dive on the Morgul Blade. How useful is it, and how often would it have helped me out if I’d had it across those same 26 tournament games? Find out next time!

Until then, may your Witch King always have just the right amount of Will and Fate!