Onuist Posted April 14, 2015 Posted April 14, 2015 (edited) I tested spell speed this evening. Fast 2.23 seconds consumed Frames to Effect: 14 Frames to Completion: 35 Frames of Recovery: 32 AVERAGE 4.3 seconds consumed Frames to Effect: 53 Frames to Completion: 79 Frames of Recovery: 50 SLOW 7.93 seconds consumed Frames to Effect: 128 Frames to Completion: 132 Frames of Recovery: 106 I used naked Wood Elves with 10 Dexterity as a baseline. Recorded casting in FRAPS at 30 FPS and checked speed frame by frame in After Effects. Dissected single spells and verified with queues of 5 spells. Speed seems to vary some within categories, as fireball and burst of summer flame gave different results. There is a slight delay between when the system declares a spell has been cast and when it realizes that spell's effect, probably based on the effect's animation. I tested the same spell with 10 Dex and 20 Dex to find that the next spell occurs 30% faster, so Dex expedites casting time and recovery time, but had little effect on realization time. For comparison, a hunting bow is 3.0 seconds consumed, 45 frames to completion, 45 frames of recovery. Fast spells are ~25% faster than hunting bows, deal ~100% more base damage and can reach a 1.4x damage multiplier (talent, bracers), so a good fast spell is akin to a hunting bow with a 2.4x damage multiplier and swift aim. Below is my raw data pasted from notepad. --------- FAST --------- (DRUID summer flame) 10 frames to be told cast 13 frames to see damage 35 frames to end casting animation 32 frames to start new casting animation //13 to damage [0.43 seconds] 35 to end [1.17 seconds] 32 to new action [1.07 seconds] 67 frames total [2.23 seconds] (DRUID summer flame [20 Dexterity]) 7 frames to be told cast 12 frames to see damage 29 frames to end casting animation 23 frames to start new casting animation //12 to damage [0.4 seconds] 29 to end [0.97 seconds] 23 to new action [0.77 seconds] 52 frames total [1.73 seconds] (WIZARD fireball) 14 frames to be told cast 17 frames to see damage 36 frames to end casting animation 34 frames to start new casting animation //17 to damage [0.56 seconds] 36 to end casting animation [1.2 seconds] 34 to new action [1.13 seconds] 70 frames total [2.33 seconds] --------- AVERAGE --------- (DRUID blizzard) 50 frames to be told cast 53 frames to see damage 79 frames to end casting animation 50 frames to start new casting animation //53 to damage [1.77 seconds] 79 to end [2.63 seconds] 50 to new action [1.67 seconds] 129 frames total [4.3 seconds] --------- SLOW --------- (DRUID returning storm) 128 frames to be told cast [4.27 seconds] 132 frames to end casting animation [4.4] 106 frames to new action [3.53] 238 frames total [7.93 seconds] *Changed a few frames after verifying with a paused queue. Fast -1 frame, Average +3 frames. Edited April 14, 2015 by Onuist 4
Daemonjax Posted April 14, 2015 Posted April 14, 2015 Good stuff. Yeah, there's animation frame delays, which make things more difficult. You might get more accurate results by queuing the casts up using the shift key and getting the average total frames between the recovery bar becoming empty.
Onuist Posted April 14, 2015 Author Posted April 14, 2015 (edited) Good advice. Thanks. 67 frames seems more accurate for Fast [2.23 seconds]. 335 frames transpire between unpause and the beginning of my 6th burst of summer flame. I'll double check my numbers tomorrow. Hopefully I can still edit my post then. On the bright side, 67 is only 1.5% lower than 68, so reliability seems high so far. Edited April 14, 2015 by Onuist
Onuist Posted April 14, 2015 Author Posted April 14, 2015 (edited) More weapon comparisons. Assuming an optimized druid and rogue at level 12. Uses 20 Might and 10 Dex, because Might is easy for me to calculate and has huge effects on static DR, while Dex is difficult for me to calculate and should benefit all attacks evenly. Talon's Reach 48.75 damage (slash, 32.5 * (1.0 + 0.3 might + 0.2 bracers)) 80 accuracy (vs. deflection, 65 base + 10 spell + 5 wood elf) 2.23 second routine Summer Flame 42.50 damage (burn, 25 * (1.0 + 0.3 might + 0.2 bracers + 0.2 talent) 80 accuracy (vs. reflex) 2.23 second routine Rogue w/ Stilettos 34.65 damage (pierce, 11 * (1.0 + 0.3 might + 0.5 sneak + 0.5 deathblows + 0.45 enchant + 0.2 reckless + 0.2 savage)) 84 accuracy (vs. deflection) 1.32 second routine Rogue w/ Hunting Bow 34.38 damage (pierce, 12.5 * (1.0 + 0.3 might + 0.5 sneak + 0.5 deathblows + 0.45 enchant)) 91 accuracy (vs. deflection, assuming wood elf + marksman) 3.0 second routine Summer flame is more accurate than the others in practice, because reflex is easier to debuff than deflection (hobbled & 2:1 attribute deflation). Hobbled is -24 reflex and comes from a level 1 druid spell [huge AoE]; blind is -28 reflex and comes from a level 1 wizard spell [large AoE]; parties built around casters can consistently open with both for an effective +52 accuracy vs. reflex but only +24 accuracy vs. deflection. Other conditions tend to be more costly and/or single target, but if the wiki is to be believed, paralyzed gives -200 reflex, which means summer flame would always crit. DPS vs. Sky Dragon: 6.83 Summer Flame 6.49 Stilettos 4.79 Talons 2.97 Hunting Bow This assumes no buffs / debuffs, but includes modes and the multipliers above. Stilettos are inflated by theory, but also benefit most from buffs and DR reduction. These numbers assume two conditions enabling sneak attack at all times, discount time melee rogues spend waiting, positioning or unconscious and apply to a singe target. Two conditions is manageable (hobbled + blind), but seconds spent positioning is never 0 for melee though is often 0 for ranged. AoE is simply a 1x factor for each additional target, although you usually only have 1 - 3 targets worth focusing in a given combat. So far my takeaway is that casters can be roughly equivalent to melee rogues at single target damage for as long as they can spare 1 spell every 2.23 seconds. Of course, if you're first reaction when fighting a dragon is to burn through level 1 spells or dive into melee, you have bigger problems than DPS minutiae. Edited April 14, 2015 by Onuist
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now