Ranger Geo Parsing (many weapons)

[Snaggles]

As a slippery slope argument against optimizing dps, a ranger might as well get a Shovel of the Harvest over a Meljeldin. It’s less than a 1% worse ratio weapon considering the damage bonus and you can summon food with it.

Being able to live on burnt bread and Fuzzlecutter Formula 5000 would be end-game bag management.

[Cecily]

Shovels are objectively less cool than claymores.

[bcbrown]

Snaggles' parses inspired me to do some more parsing of my own, and I finally got the time to do so this past week. I had two aims: 1) measure the gap between the DPS a raid-geared 60 ranger can do and that of an EC-geared level 59 ranger without any AoB items, and 2) look into whether it's worth getting Swiftwind and a good mainhand for a ranger who already has a good EC 2h weapon.

I buffed with Storm Strength, SoN, Call of Earth, Spikecoat, and Skin Like Nature. For the initial set of Shady fights I used Greater Wolf Form, but for the Basher Nanrum parses and the follow-up Shady parses I skipped Greater Wolf Form. For all 1h weapons the offhand was Swiftwind (40% haste), while the 2-handers had 34% from the seahorse belt. No proc damage is included in these calculations.

First, Shady:

A Shady Swashbuckler
weapon         time  damage    dps
bfg trueshot    281   11483    40.86
bfg             513   16841    32.83
spade 2         491   14285    29.09
sboz            477   12903    27.05
spade          1009   26915    26.67
spade 3         990   25673    25.93
rev whip        661   17082    25.84
infestation 2   755   18871    24.99
fellspine       573   14304    24.96
infestation     284    6742    23.74
swarmcaller 2   610   13945    22.86
swarmcaller     768   16530    21.52

Interestingly, the only weapons that could kill Shady were BFG and Swarmcaller, the highest and lowest DPS. While first parsing Infestation I realized I was about 15 points from capping Piercing skill, so I returned after capping it and reparsed. I also initially had a data-entry error with the Herbalist's Spade that showed it with similar DPS to the swarmcaller, which is why I returned for a couple more parses with both the spade and the swarmcaller.

One other nice side effect of re-parsing the same weapon is showing how even after 500-1000 seconds there's still a lot of variation in the final DPS numbers. So even though I'm showing four significant digits all these results should be treated as being +/- 2-3 DPS.

Shady apparently has as much AC as some of the harder raid targets, so this test was meant to be a proxy for raid dps. I was specifically curious if the low delay of the rev whip would help it break away from fellspine or infestation, but although it was the highest of the three the differences are all within the margin of error.

For group/xp fighting I picked Basher Nanrum as a target. This was mostly because he was conveniently close to the tunnel, but he's also a static level, and at 15k hp it was possible to run longer parses than the other bashers in Grobb.

Basher Nanrum
weapon         time  damage    dps
disc bfg        96     7695    80.16
bfg             400   22468    56.17
sboz            487   23053    47.34
spade           350   15291    43.69
fellspine       455   18302    40.22
rev whip        422   16852    39.93
swarmcaller     467   15489    33.17

The rankings were mostly the same as against Shady, but with higher numbers. The parse lengths are shorter than against Shady, partially because it's more annoying to reset the fight, partially because doing all these parses and analyzing them takes a long time and I was starting to flag on the project.

Regarding the second aim, trying to see whether someone with a good 2-hander should try to get a Swiftwind, I'd conclude that Swiftwind will indeed do more dps, but not until you get a sboz or better to pair with it, and it'll only be slightly better than an herbalist's spade. It's also worth reiterating that this isn't with equivalent haste: the 2-handers had 34% while the 1-handers got 40% plus the atk boost from Swiftwind.

Regarding the second aim, comparing dps to Snaggles, I'd like to normalize haste, because I'm less interested in the benefit from having the sky cloak (50% spell haste is obviously and straightforwardly better than 0 spell haste). I don't find that sort of conversion to be especially intuitive, but I think the right way to convert from 40% haste to 91% haste is to multiply by 1.91/1.4 or 1.36, and 1.91/1.34 or 1.43 for 34% haste. I'm not going to worry about how different Basher Nanrum is from geonids from an AC perspective; it should be fairly comparable I think.

Snaggles got 52 DPS with Swarmcaller, 70-90 DPS with the other weapons. Normalized to 91% haste I got 45 with Swarmcaller, 64 with sboz/swiftwind, and 76 with non-Trueshot bfg/swiftwind. So a clear gap, but pretty cool to see it's not too big.

Lastly, I'd like to thank Goregasmic for lending me the Herbalist's Spade.