Рассуждение:
Scenario 1
Mastery - 65%
Haste - 50%
1.3 s cast
462 casts
first 5 frostbolts yield 500k + 5 x 65k icicles stored.
FB 6 and every FB after yields 165k total dmg
That leaves 457 casts less final cast (below) 456 casts.
on cast 462 release 1 FB and remaining icicles for 425k.
total damage is: 500k + 75,240,000 + 425k
76165000 over 600 s - 126942 dps.
Scenario 2
Haste 63%
Mastery 45%
1.2 s cast
500 casts
first 5 frostbolts yield 500k plus 5 x 45k icicles
FB 6 and every FB after yields 145k total dmg
That leaves 495 casts less final cast (below) 494 casts
on cast 500 release 1 FB and remaining icicles for 325k
total damage is 500k + 71,630,000 + 325k
72,455,000 over 600 s - 120758 dps.
Отсюда сделан вывод, что "чистая" мастери лучше хасты.
Но:
Just accounting for the difference in meta procs (based on my last logs from mastery forge and last logs from haste forged over same fight count) of 22% and 35% respectively:
Scenario 1 yields 22% of casts at 1 s
so that's .78 * 462 = 360 casts at 1.3 seconds or 468 s of the fight leaving 132 seconds left at 1 s casts or 132 casts. This brings the new total to 492 casts. 492 casts minus the first 5 and last 1 yields 486 casts at 165k, new total being 82,105,000 over 600 seconds or 136842 dps.
Scenario 2 yields 35% of casts at below/effectively 1 s
So that's .65 * 500 = 325 casts at 1.2 seconds or 390 s of the fight leaving 210 seconds at 1 s casts or 210 casts. This brings the new total to 600 casts. 600 casts minus the first 5 and last 1 yields 594 casts at 145k, new total being 86,955,000 or 144925 dps.
Отсюда, по логам, определили, что аптайм меты с большей хастой больше, а значит и дпс получился опять выше, при этом не учитывали что бомба бьет больше из-за доп тиков. Но т.к. мы упираемся в кап 1.0сек фб под метой. Нужно брать все тот же кап в 12к хасты.