Haste and Casting Time question

[Ridwane]

Hello everyone !

I've read my tooltip, even if i have a decent amount of haste, it still say 1.3 casting time ? Can twink get below or is it normalized that way ?

Thanks.

 

[Phrontistery]

just tested cause i was curious and with 27% haste at level 20 i was 1.2sec casts on shaman

(22% scaled in bg my casts were still 1.2sec)

i doubt you can't get 1.2 at 19, but yeah this was testing on a 20 so idk

 

[Vandiir]

There's no lower bound to cast times being reduced by haste (I remembered several expansions ago when 10s twinks were able to get their 1.5 sec cast times down to 0.7 seconds, back when 10s secondary scalings+boatloads of secondaries from enchants was a thing), though to get any noticeable cast time reductions requires a nontrivial (or even non-obtainable) amount of haste.

Also for many classes, GCDs (global cooldowns, including GCDs from instant cast abilities) are also affected by haste, though the hasted GCD (unlike cast times) would go no lower than 1.0 seconds (assuming non-hasted GCD=1.5 seconds for a class, one would need 50% haste to reach have 1.0 sec GCDs).

Basically, here's the math:
Given you have cast time C (in seconds), and haste H (H being in decimal form, like 24% haste would be H=.24), then your hasted cast time, C_H, would be
C_H=C/(1+H)

It thus follows that if you instead want to find the value haste H such that C_H=t (where t is your desired cast time), then t=C/(1+H) implies
1+H=C/t => H=(C/t)-1=(C-t)/t.

For example, if C=1.5 sec and H=20%=.2, then C_H=1.5/(1.2)=15/12=5/4=1.25 sec (and tooltips may round that value to the nearest .1 digit, which can be either 1.2 or 1.3).

Or to reduce 1.50 sec to 1.30 sec, we get H=(1.5/1.3)-1~.153846, or about 15.39% haste.
Or from 1.50 to 1.26, H=(1.5/1.26)-1~.19047619, or ~19.05% haste.
Or from 1.50 to 1.24, H=(1.5/1.24)-1~.2096774, or ~20.97% haste.
Or from 1.50 to 1.20, H=(1.5/1.2)-1=.25, or 25% haste.
Or from 1.50 to 1.10, H=(1.5/1.1)-1~.36363636..., or ~36.36% or 36.37% haste.

And so on and so forth.

 

[Goesid]

There's no lower bound to cast times being reduced by haste (I remembered several expansions ago when 10s twinks were able to get their 1.5 sec cast times down to 0.7 seconds, back when 10s secondary scalings+boatloads of secondaries from enchants was a thing), though to get any noticeable cast time reductions requires a nontrivial (or even non-obtainable) amount of haste.

Also for many classes, GCDs (global cooldowns, including GCDs from instant cast abilities) are also affected by haste, though the hasted GCD (unlike cast times) would go no lower than 1.0 seconds (assuming non-hasted GCD=1.5 seconds for a class, one would need 50% haste to reach have 1.0 sec GCDs).

Basically, here's the math:
Given you have cast time C (in seconds), and haste H (H being in decimal form, like 24% haste would be H=.24), then your hasted cast time, C_H, would be
C_H=C/(1+H)

It thus follows that if you instead want to find the value haste H such that C_H=t (where t is your desired cast time), then t=C/(1+H) implies
1+H=C/t => H=(C/t)-1=(C-t)/t.

For example, if C=1.5 sec and H=20%=.2, then C_H=1.5/(1.2)=15/12=5/4=1.25 sec (and tooltips may round that value to the nearest .1 digit, which can be either 1.2 or 1.3).

Or to reduce 1.50 sec to 1.30 sec, we get H=(1.5/1.3)-1~.153846, or about 15.39% haste.
Or from 1.50 to 1.26, H=(1.5/1.26)-1~.19047619, or ~19.05% haste.
Or from 1.50 to 1.24, H=(1.5/1.24)-1~.2096774, or ~20.97% haste.
Or from 1.50 to 1.20, H=(1.5/1.2)-1=.25, or 25% haste.
Or from 1.50 to 1.10, H=(1.5/1.1)-1~.36363636..., or ~36.36% or 36.37% haste.

And so on and so forth.

Haste is easy to test in deadmines, Think the cap is 1971%

 

[jackblood]

If an ability’s CD is reduced by Haste, and it currently reads 7.2s, is it worth it to get it at least down to a whole number? Or is that not too important?

 

[Psilocybine]

Quick test result from disc priest... spell flash heal

1.5 sec @ 0-7 haste
1.4 sec @ 8-31 haste
1.3 sec @ 32-?

I wasted a lot of enchants trying to get it lower.
[doublepost=1533328761,1533327921][/doublepost]

There's no lower bound to cast times being reduced by haste (I remembered several expansions ago when 10s twinks were able to get their 1.5 sec cast times down to 0.7 seconds, back when 10s secondary scalings+boatloads of secondaries from enchants was a thing), though to get any noticeable cast time reductions requires a nontrivial (or even non-obtainable) amount of haste.

Also for many classes, GCDs (global cooldowns, including GCDs from instant cast abilities) are also affected by haste, though the hasted GCD (unlike cast times) would go no lower than 1.0 seconds (assuming non-hasted GCD=1.5 seconds for a class, one would need 50% haste to reach have 1.0 sec GCDs).

Basically, here's the math:
Given you have cast time C (in seconds), and haste H (H being in decimal form, like 24% haste would be H=.24), then your hasted cast time, C_H, would be
C_H=C/(1+H)

It thus follows that if you instead want to find the value haste H such that C_H=t (where t is your desired cast time), then t=C/(1+H) implies
1+H=C/t => H=(C/t)-1=(C-t)/t.

For example, if C=1.5 sec and H=20%=.2, then C_H=1.5/(1.2)=15/12=5/4=1.25 sec (and tooltips may round that value to the nearest .1 digit, which can be either 1.2 or 1.3).

Or to reduce 1.50 sec to 1.30 sec, we get H=(1.5/1.3)-1~.153846, or about 15.39% haste.
Or from 1.50 to 1.26, H=(1.5/1.26)-1~.19047619, or ~19.05% haste.
Or from 1.50 to 1.24, H=(1.5/1.24)-1~.2096774, or ~20.97% haste.
Or from 1.50 to 1.20, H=(1.5/1.2)-1=.25, or 25% haste.
Or from 1.50 to 1.10, H=(1.5/1.1)-1~.36363636..., or ~36.36% or 36.37% haste.

And so on and so forth.

Pet peeve of the interwebz.....a long post that seems legit but fails testing.

"to reduce 1.50 sec to 1.30 sec, we get H=(1.5/1.3)-1~.153846, or about 15.39% haste." ...the actual percentage is 11.27%.

Even if things seem true..


research for yourself. Good rule to live by

 

[Baiktos]

For shaman healer, 11.44% (36 haste) is ideal because it's the lowest obtainable value that still gives 1.3s heal (with 1.8s LB)
1.7s lightning bolt is around 18-19% if i remember correctly. Hardly worth it imo because it won't drop heal down to 1.2

 

[Swoops]

Haste and cast time scales continuously. The "breakpoint" where your cast time tooltip changes is pretty irrelevant. It's just a spot where it rounds differently.

Haste breakpoints don't really exist anymore (outside maybe a few minor exceptions at higher levels like warriors fitting globals into battle cry).

 

[fahr]

For shaman healer, 11.44% (36 haste) is ideal because it's the lowest obtainable value that still gives 1.3s heal (with 1.8s LB)
1.7s lightning bolt is around 18-19% if i remember correctly. Hardly worth it imo because it won't drop heal down to 1.2

And for resto druid man? To get to 1.3 sec heal?

 

[necroaqua]

Quick test result from disc priest... spell flash heal

1.5 sec @ 0-7 haste
1.4 sec @ 8-31 haste
1.3 sec @ 32-?

I wasted a lot of enchants trying to get it lower.
[doublepost=1533328761,1533327921][/doublepost]
Pet peeve of the interwebz.....a long post that seems legit but fails testing.

"to reduce 1.50 sec to 1.30 sec, we get H=(1.5/1.3)-1~.153846, or about 15.39% haste." ...the actual percentage is 11.27%.

Even if things seem true..


research for yourself. Good rule to live by

As Swoops said, the cast times are still going down - the tooltips are just showing rounded to the second.

As to the math, I'd guess the correct math is cast time = base cast - (haste % * base cast). Or haste % needed = (base cast - desired cast time) / base cast time.

So (1.5 - 1.3)/1.5 = 13.3% haste needed to get down to exactly 1.3s cast time. Guessing the rest is tooltip rounding.


Without breakpoints, haste is just nice for faster heals/globals/resource regen. Its hard to measure the value of a heal that got out .1 second faster and saved someone where it wouldn't have without the haste. In PvE, a lot of those small things add up over time (.1 second less on a cast over the course of a 10 minute fight means a significant damage uptick). In PvP, a lot of the times things happen much quicker and theres a lot more breaks in the action that kill any snowballing advantage like that. Something like crit might be a flashier and easier to see when it happens.

Side note, I believe haste does use breakpoints for resource regen like rogue energy or hunter focus, as you don't have fractions of energy/focus. I believe those are a simple 10 energy/s * (1+haste %), but I haven't actually looked. May be worth shooting for an extra haste item or 2 if it hits you that 1 extra energy per second. Or perhaps they've normallized it and instead do a time based system, so the haste doesn't actually make it 11 energy per second, but rather that 10 energy now comes over .9 seconds or something...

 

[Psilocybine]

As Swoops said, the cast times are still going down - the tooltips are just showing rounded to the second.

As to the math, I'd guess the correct math is cast time = base cast - (haste % * base cast). Or haste % needed = (base cast - desired cast time) / base cast time.

So (1.5 - 1.3)/1.5 = 13.3% haste needed to get down to exactly 1.3s cast time. Guessing the rest is tooltip rounding.


Without breakpoints, haste is just nice for faster heals/globals/resource regen. Its hard to measure the value of a heal that got out .1 second faster and saved someone where it wouldn't have without the haste. In PvE, a lot of those small things add up over time (.1 second less on a cast over the course of a 10 minute fight means a significant damage uptick). In PvP, a lot of the times things happen much quicker and theres a lot more breaks in the action that kill any snowballing advantage like that. Something like crit might be a flashier and easier to see when it happens.

Side note, I believe haste does use breakpoints for resource regen like rogue energy or hunter focus, as you don't have fractions of energy/focus. I believe those are a simple 10 energy/s * (1+haste %), but I haven't actually looked. May be worth shooting for an extra haste item or 2 if it hits you that 1 extra energy per second. Or perhaps they've normallized it and instead do a time based system, so the haste doesn't actually make it 11 energy per second, but rather that 10 energy now comes over .9 seconds or something...

See previous post...

 

[Swoops]

See previous post...

Vandiir's post is accurate.