[bcbrown]

I ran a couple experiments with three variables:
* Swarmcaller versus Swiftblade of Zek/Swiftwind
* Bow of the Huntsman(20/30) versus Rain Caller (20/45)
* Bought arrows (1 dmg) versus Summoned arrows (3 dmg)

I didn't realize until after the experiment that my two bows had the same damage with different delay. I knew Huntsman had a better ratio but completely forgot they had same damage. Oops.

I killed one hill giant with each setup, and I didn't do all the possible permutations. For one fight I killed Shiel instead, so for that fight I'm not including the average damage. This data is all incredibly noisy, since each fight is only 40-50 hits. I was hoping it would hint at a similar situation to melee damage, where there's a modal value that's the most common value to hit for, with long tails above and below that value that are affected by str, attack, offense/defense skill differential, level differential, whatever else.

Here's the data. The first row is against Shiel, the rest hill giants. The three numeric columns are modal hit, average hit, and the calculated MH bonus + bow damage + arrow damage.

bought arrows   sboz    gleeds      32      31
bought arrows   swarm   gleeds      49  48  48
summoned        sboz    gleeds      34  37  33
summoned        sboz    raincaller  34  40  33
summoned        swarm   gleeds      51  55  50
summoned        swarm   raincaller  51  53  50

The fact that both bows have the same damage value is too bad, but it does look like the MH bonus + bow damage + arrow damage calculation does predict the modal value pretty well. I'm guessing (MH bonus + bow damage + arrow damage) / delay is gonna work about as well as a rule of thumb as (2x damage + MH bonus) / delay does for mainhand melee.