Does anyone have some excel sheet or similar where you have made calculations on damage per point or damage per activation including rerolls, surges, build in bonuses or similar? I am looking for top 5 for each faction or something like that.

So rerolls are very difficult to factor in to the damage per point equation. If you want somewhere where you can look at the very dice, surge abilities, etc. Against various defensive bonuses, this is the place for you.

Sounds like a Paul Heaver thing.

4 hours ago, ffmonkey said:

So rerolls are very difficult to factor in to the damage per point equation. If you want somewhere where you can look at the very dice, surge abilities, etc. Against various defensive bonuses, this is the place for you.

Thanks! That is a great tool for sure, but the rerolls are so important that if they are not factored in the results will not show the right results. Ezra with Kanan hits way way harder then Ezra alone, eRangers average damage goes way up with built in rerolls ettc etc.

I agree, it’s difficult to model though. How do you judge what result is objectively better? At what point should the model reroll and what dice? You could model Ezra as yellow green with a built in DMG X 2 Sur x2.

18 minutes ago, ffmonkey said:

I agree, it’s difficult to model though. How do you judge what result is objectively better? At what point should the model reroll and what dice? You could model Ezra as yellow green with a built in DMG X 2 Sur x2.

This right here - rerolls are really tough in IA, as it depends so much on the game state.

Like say I roll the single damage side on a blue die (and for the sake of argument say I've got lots of surges on my other dice). Is that a re-roll? Well there are two better sides (2-damage) and only one worse side (single surge) so in a vacuum I'd say yeah, you're better off re-rolling it.

But what if I'm attacking a 6-health figure and I'm doing exactly 6 damage? Then I won't re-roll. But what if also my opponent has Zillo? Then, yeah, I might want to re-roll. But what if also I haven't yet activated the Emperor, who could add an extra damage on his activation? Then I won't re-roll. But what if the figure I'm attacking would have a chance to go before the Emporer activates? Well then I would re-roll. But what if......

I think if you were going to try to incorporate re-rolls into a model, you would have to be able to set parameters with a specific goal in mind. I think it would generally be more helpful to ask the question "what are the odds that this figure can do at least 6 damage", and then assume re-rolls to try and accomplish that, rather than just asking the general question "what's the most damage this figure can do". Although against some high-health figures that second questions might actually be the right one to ask.

It's all complicated, is what I'm saying

I have no clue how to include rerolls in damage calculations considering how context specific the decision to reroll or what to reroll is. Not to mention situational rerolls like Alliance Rangers. Right now I just do the calculations based on the normal attack and remind myself they have a reroll when I factor in "intangibles" when list building.

Unless you're a statistics (and excel) wizard I think that's the way to go

Edit: Speaking of Alliance Rangers, if memory serves the last time I ran the numbers they had the best damage per point cost of any non Spectre Cell enhanced group. And that didn't take in to account their awesome reroll.

Well, sure its hard, otherwise I would not have asked you all!

I have been fudge:ing it for a long time, estimating that the a reroll is worth +1 damage (and +1 range), but I felt the urge of trying to find some hard facts. It started by me wanting to find out if there is a squad or setup that can actually match the damage output and in some sense HP ratio of the Spectre cell. So I figured "Hey, lets ask the smart people!".

1 hour ago, Ram said:

Well, sure its hard, otherwise I would not have asked you all!

I have been fudge:ing it for a long time, estimating that the a reroll is worth +1 damage (and +1 range), but I felt the urge of trying to find some hard facts. It started by me wanting to find out if there is a squad or setup that can actually match the damage output and in some sense HP ratio of the Spectre cell. So I figured "Hey, lets ask the smart people!".

Unfortunately there isn't. Scum Hunter lists can perhaps match damage output when you take into account those sweet command cards. But on an basic attack basis there's nothing which is what makes Spectre so good.

Rangers are the unit that comes closest.

Time ago I started to write a simulator in python. Idea was to simulate an attack (in example Chewie Vs Vader). Repeating the simulation enough times I could collect any sort of data like: damage distribution, mitigated damage distribution, distribution of reroll effects (simulation of rerolls took into account the best option at step 3, simulating all possible outcomes face by face) and other meaningful probability distributions. The ideal wasn't to simulate the attack itself but rather whole activation, a scripted chain of command cards and abilities for specific attacker and defenders. I wanted to assess the effectiveness of a combo. At the moment I couldn't work more on the project I could simulate all the steps of an attack applying all while attacking/defending abilities at any step. Here is an example (I didn't implement charts GUI yet):

$ python.exe swia-skirmish-calculator.py -a 70 -d 25 147 -r 3 -n 5000 -s 0
+-------------------+    +--------------------------------+
| Elite Jet Trooper | VS | Darth Vader + Driven by Hatred |
+-------------------+    +--------------------------------+

[XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX] [100%]

-----------------------------------------------------------
Total damage @ range 3 (5000 runs)
-----------------------------------------------------------

PDF:
0: 30.1%
1: 28.94%
2: 28.38%
3: 11.56%
4: 1.02%

CDF:
0: 100.0%
1: 69.9%
2: 40.96%
3: 12.58%
4: 1.02%

Average: 1.2446 damage(s)

-----------------------------------------------------------
Avoidance @ range 3 (5000 runs)
-----------------------------------------------------------

PDF:
0: 2.2%
1: 10.82%
2: 34.42%
3: 34.8%
4: 15.94%
5: 1.82%

CDF:
0: 100.0%
1: 97.8%
2: 86.98%
3: 52.56%
4: 17.76%
5: 1.82%

Average: 2.5692 damage(s)

-----------------------------------------------------------
Over-surging @ range 3 (5000 runs)
-----------------------------------------------------------

PDF:
0: 89.52%
1: 10.48%

CDF:
0: 100.0%
1: 10.48%

Average: 0.1048 surge(s)

-----------------------------------------------------------
Reroll impact @ range 3 (5000 runs)
-----------------------------------------------------------

PDF:
-1: 5.36%
0: 57.9%
1: 26.32%
2: 9.38%
3: 1.04%

CDF:
-1: 100.0%
0: 94.64%
1: 36.74%
2: 10.42%
3: 1.04%

Average: 0.4284 damage(s)