I figured out the perfect way to test attack speedups. Equip
a nonmagical weapon, and attack a wisp. :)

Here's what I got punching a wisp. Since I wasn't hurting the
target, I could let the fight run much longer than usual. I
also logged it and searched through the logfile afterwards to
confirm the number of swings (the other problem I had with
long sample times was that I'd miscount the number of swings).

Time Swings Delay %chg Delay %chg Speed
Normal 177.7 50 35.54 - -
Warsong(50) 285.8 100 28.58 -19.58% 24.35%
Anthem 162.5 50 32.50 -8.55% 9.35%
Celerity 147.8 50 29.56 -16.83% 20.23%
Beserk(50) 148.1 49 30.22 -14.96% 17.59%
Verses 136.9 50 27.38 -22.96% 29.80%

%chg Delay = 100 * [(song_delay / normal_delay) - 1]
%chg Speed = 100 * [(normal_delay / song_delay) - 1]

Looking at this, pretty much everything lines up with the %chg
speed being 25% for Warsong at level 50, 10% for Anthem, 20%
for Celerity, and 30% for Verses. Beserkers I'm pretty sure
changes as you level.

The big thing to note here is that the speedup for Verses is
*better* than Warsong for soloing.

Another curiosity is that nearly all the delays have a 0.5
second biasing. You'll remember I remarked on this in my last
set of tests, during which I speculated it was something I was
doing. With a sample this large, I'm confident enough to say
it's probably a feature of the game, and an artifact of the
weapon delay system Verant is using.

I will test enchanter speedups when I can nab an enchanter to
buff me. Considering I'm giving one of the guild enchanters a
tinkerers bag (so he can carry around aqualungs to sell), this
should be pretty easy. :) I will also try it with various FS
and rusty weapons.

--
John H. Kim
kim@...