Exorcist Damage Speculation

[Beams]

So, after the quick discussion we had about the Exorcist having a D6 shot, Str 8, AP -3, D6 damage profile, and whether or not that would be appropriate, I couldn't stop thinking about whether or not it was a valid profile.

 

So I wrote a program that would compare it to the Leman Russ Battle Tank, fulling kitted with a Lascannon and a Heavy Bolter, when both fighting a dreadnought. Why a dreadnought? Because it was an easy thing to compare against.

An exorcist firing D6 shots dealing d6 wants to fight a single target, or at least a squad of terminators. A Leman Russ Battle Tank wants the same thing, since they don't have a huge amount of shots either.

 

Before I go any farther, the Leman Russ was run 10,000 times.

The Exorcist, 20,000.

Really, I should run it 100,000 to get a more accurate response, but since it takes about an hour for 10,000 repetitions, I figured this is a good eye-ball.

 

 

(The Exorcist is in the Red, and the Leman Russ in the Blue.

The Exorcist is based on the first 10,000 shots)

 

Observations:

 

The Leman Russ dealt 0 Damage 19.5% of the time, but dealt over 8 damage a approximately 6% of the time. The full list of damage was posted in the news and rumors forum. That being said, the Leman Russ was fairly reliable, dealing 3 or more damage 49.5% of the time.

 

The Exorcist, again assuming a D6 Str8 Ap -3 D6 Damage profile, however, missed a startling 26.79% of the time. Which makes total sense, considering it has less shots, even if the Dreadnought is less likely to save against them. That being said, the Exorcist also had a 22.06% of dealing 8 or more wounds to the dreadnought, with it capping out at a .02% chance of dealing 25 (!!!) damage. Sure, that only happened 3 times in the 20,000, but it did happen. The Exorcist dealt 12 or more damage 7.83% of the time, which means that it has a (slightly better) than 1/20 chance to deal 12 damage.

 

The Exorcist also dealt 6 Damage more than any other damage value.

 

If the Exorcist had a -4 instead of a -3, we'd see a slight decrease in total misses and a small increase in damage, but nothing major, since the bulk of the variation is in the D6 damage.

 

 

Conclusion:

While I don't think that the Exorcist will be D6 damage, it would give us the wildly unpredictable weapon that we all know and love. However, this would also make it one of the most potent anti-Armor platforms in the game, with an outside chance of destroying anything in it's path.

 

[Servant of Dante]

I'd really like to see the exorcist be 1d6 shots at 1d6 damage. Preferably with -4 AP. It would keep the numbers largely the same from what we have now, and keep the randomness :P which I kinda like. I think it would also help preserve the fluff of firing a random number of missiles, since this way you can see that each missile does 1d6 damage, whereas if shots are increased and damage is 1 it is the same as a very large blast weapon would be, imo

[Beams]

I'm expecting it to be D3 damage, which will push it closer to the leman Russ. (Which, in 7th would have been 180 pts vs the Exorcists 125)

 

The exorcist will still be unpredictable, and with a higher AP will be able to deal more damage against heavier targets, making the Exorcist missiles worth it in comparison to other options.

[Servant of Dante]

D3 is probably more likely. I'm expecting

 

Heavy D6 S8 AP-4 Dam.D3

 

Would be most likely

[Eddie Orlock]

I was curious why you elected for a numerical method when an analytic method was available. Naturally I just can't leave well enough alone so I produced this graph via said analytic methods.

Seems to be a fairly decent curve fit to your work. I assumed d6 shots, d6 damage per wounding hit, hit on 3+, wound on 3+, saves are defeated. With these parameters the speculative missile system has an almost 30% chance to 'one-shot' an 8 wound target, a more likely result than missing entirely at around 20%.

Six wounds appears oddly probably until you remember that it's the one wound sum that appears in all hit sets and that about a third of the time the volley only results in 1 wounding hit meaning that that mode dominates where it is applicable.

[Eddie Orlock]

As a follow up, I think I'll hang on to this spreadsheet, as it seems useful in calculating general explosives effectiveness in the dark future.

[Servant of Dante]

I'm definately more interested in statistical probability than quasi-random virtual dice rolling experiments :P thanks Eddie :D

 

Interesting analysis though Beams

[Beams]

Thanks.

 

The only real difference between our curves is I figured a 6+ save was left, which gave the exorcist a slightly higher chance to whiff.

As for my modeling, I'm a programmer not a mathematician, and once you have one program that runs dice, which is one of the first programs you learn to do, it's super fire and forget vs. figuring out the formula for the likelihood of each different damage value.

And while there isn't really a difference between them, it also allows me to say that "out of 20,000 tries, we never got past 25 damage" instead of "Well, there's an improbably low chance that you might get 26 or higher."

Also, I can than throw individual units against each other in a literal battle. I had a program that would run 7th edition characters in a battle versus each other for 1,000+ trials. Did that to see if I should take a power maul or power sword, and to factor in all that different variables without a major spreadsheet. (You want the power maul if your getting into a challenge) Eventually I want to have the program run entire armies, with a variety of strategies, so that I can see what kind of odd things you can do. Like how many Grots does it take to defeat a Gladius Strike Force, and stuff like that, all while taking into account different maps with different terrain and cover. Basically, like a Bravest Warriors for nerds.

Also, sure it's "random dice" but at that many trials the luck factor is pretty much removed.

[Servant of Dante]

Since the experiment basically just approaches the statistical value, I prefer to just have that. And I like exact answers (and I kinda enjoy the statistics of dice rolling, formulas and all).

 

The points you made still stand though.

[Ficinus]

It takes a lot longer (and is incredibly boring/tedious) to do the calculations for each level of damage output.  The 10,000 trials are a good enough representation.  Do you really need to know that the odds of doing 36 damage hitting on 3+ and wounding on 3+ with no save is something like 2¹²/(6⁷3¹²) which is about 4096/148769467776 which is a .0000000275325311% chance of occurring?  Can't you just know it will never happen?

[Servant of Dante]

I mean, I don't need to. Your method is certainly easier. I just like exact answer, and I find the calculations involved satisfying, and it's interesting to know how often is "never"

 

I'm sorry for how I worded my initial comment.

[Ficinus]

It was a bit interesting to see how never "never" is in this case.  The odds of doing 36 damage appear to be close to one in fifty million.  I sort of doubt, over the entire course of eighth edition, people will shoot an Exorcist fifty million times. Assuming eighth edition lasts a full decade, then an Exorcist launcher would need to be shot 13,691 times per day (counting the two leap days over the next decade) to get close to fifty million.  The most interesting part is, though, that even if we were to shoot an Exorcist fifty million times, there's actually only a 63.21% chance of getting one or more 36 damage volleys, which is a pretty good shot but not great.  So, I think we can say that it is incredibly unlikely that it will happen in any individual's experience and that it's incredibly unlikely in all individuals' experiences.  Also, this all assumes that the Exorcist is ignoring its target's save.  If it doesn't, that would drive the odds down even more.  It's also assuming the Exorcist is shooting at a target where it needs to roll damage.

 

Calculating the odds of getting six damage, though, is a much more daunting process that I do not want to calculate.  Calculating 36 damage is easy (have to roll six shots, all have to hit, all have to wound, all have to roll a six for damage), the number of ways to roll six damage is crazy (can be done on any number of shots, can be done with any non-zero number of hits, can be done with any non-zero number of wounds, can be done with a multitude of different damage rolls).  There are some shortcuts (calculate the odds of a given number of hits for all numbers of shots, then calculate the odd of a given number of wounds for a all possible numbers of hits, then calculate the odds of each potential damage roll combination for each number of wounds), but it's still a tedious procedure unless you have a program to do it all for you.  Like you, I'd be interested in seeing the breakdown of all the odds, but there's no way I'll ever be bored enough to do the work by hand.

[Servant of Dante]

Yeah, it'd never happen (probably). This is interesting in any case. I'm looking forward to seeing all the weapon stars for 8th edition in full when it's released.

[Drider]

What does the curve look like if we use the a command point to reroll the number of shots fired? Also what would be the optimal maximum number to reroll, It's obvious you would want to reroll a 1 or a 2 but would you want to reroll a 3? It's under the 3.5 average but would you be hurting yourself if you rerolled it?

[Servant of Dante]

Statistically if it's under the average it's going to be good to re-roll it. There are 3 results on the die higher than 3, and 2 that are lower.

[Ficinus]

I would say that it's probably going to be a bad idea to re-roll the number of shots with a command point.  Even if you get a one, sure, you're likely to get more shots, but you wasted one of your command points on a random number of shots.  Maybe if you're targeting something on its last legs and you want to be sure you'll finish it off it might make sense, but re-rolling that one and getting a two or even a three, while more shots is probably not going to be worth burning the command point (though, who knows at this point).  Also given that you'll be rolling to hit and to wound, it's only so likely that those extra shots will be translated into more damage.

 

This is all just speculation, of course, on command point re-roll tactics.

[Beams]

I would avoid rerolling a 3, out of the idea that whenever I get greedy karma punishes me. There's a pretty good chance (50%) that if you reroll a 3 you'll get a 3 or lower, which is too high of a chance to waste a command point.

 

Where only rerolling a 1 or 2 would give you a 33% chance of rolling a 1 or 2, and a 66% chance of rolling higher.

 

I'm curious, and so I might run some programs after work, about manipulating or hypothetical exorcists.

 

Like, if we need something absolutely dead, can we spend a command point to kill it?

Basically, figuring that we have a single command point, a basic if then tree occuring whenever one of these situations shows up, and we haven't yet

 

Roll a 1 or 2 on shots fired,

 

Reroll if no hits

 

Reroll if no wounds

 

And if there is damage of 1 or 2 reroll.

 

And see how much damage we can really can out of a single exorcist. I don't expect it to do too much, but to make it more reliable and to whiff less.

 

Also thinking the exorcist missile might just be damage 2.

[Servant of Dante]

Yeah with a single reroll at the cost of a command point I'd almost never want to reroll a 3, but I was just saying that if you have a reroll, statistically rerolling a 3 on a d6 will improve your roll

[AliciaDominica]

I'll just leave this here as an AOS mechanic that I could see them adjusting slightly for the exorcist...http://i.imgur.com/mKUAwHH.png

[Servant of Dante]

Ooooh that would be interesting! I'd have fun with that!

[Beams]

That sounds awesome... It would also be decent to have a command point on the ready for that!

[Knight of the Raven]

I can totally see Games Workshop giving the Adepta Sororitas the only weapon that jams in the entire 8th edition.

 

(Though with how gracious they appeared to have been with orks, I can also see them making all ork ranged weapons jam too.)

[AliciaDominica]

To be fair, it would only jam if you choose to overload it. Which is fairly fluffy for an incredibly archaic and hard to maintain design. There is a reason Space Marines prefer Las Canons to Exorcist organ guns...

[robofish7591]

That could be a pretty fluffy rule, since exorcists are supposed to be finicky machines anyway.  Plus it would make it feel a bit different from battle cannons, now that they are just d6 shots.

[micahwc]

Aren't exorcist machine spirits supposed to be especially mean and vindictive? I'd love a rule for a jam, and then the machine spirit fired all the missiles on the next turn out of spite.