dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
740
share rss forum feed


Pollux7777

join:2010-02-16
Saint Paul, MN

1 edit

[Theory] I need someone better at math than me...

Disclaimer. Math was never my strong suit, so the answer to my question may be painfully obvious. But I need some help with simcrafting between two trinkets.

I play a fury warrior. Polluxx on ravencrest if you're interested enough to look me up. I'm contemplating upgrading one of my trinkets (lei shens final orders) with a rep reward from operation shield wall (helm breaker medallion).

Lei shens trinket has haste on it, which isn't a great stat for a fury warrior. The shieldwall one has strength, which is much better. So that's a no-brainer.

But it's the proc on lei shens that throws a wrinkle into the equation. According to wowhead the trinket has a 15% chance on direct white/melee/ranged/harmful-spell landing to increase strength by 2866 for 20 seconds, with a 45 sec ICD.

The shield wall trinket is a static +1152 strength. Still with me?

I'm dual wielding single handlers with relatively fast swing speeds, so I'm hitting a lot. So basically once that 45 sec internal cool down is up it procs again pretty quickly.

So the question then is which is better overall? 1152 strength ALL the time, or almost three times that amount some of the time?

I don't know how to form an equation to figure that out.

EDIT: I almost forgot. The shieldwall trinket also has an on use crit bonus, so I'm guessing when you factor that in its probably better overall, but I'm still interested in the math.
--
Polluxx - LvL 90 Human Warrior - (US) Ravencrest

Cptbeatstix

join:2011-12-21
Carrollton, TX
Reviews:
·Verizon FiOS
It's hard to judge based on just plain math. What you are asking is more of having to copy your exact character and hope for the random 15% proc to proc a lot. All in all, I think the static str increase over will be more helpful in the long run seeing as crit procs are your friend. Best thing to do is to sit at the dummy with both trinkets. Do 5 runs with one trinket and average the dps. Repeat for trinket 2. Should give you a clear window on which one is better. Look at it this way, even if its 20 more dps, a dps increase is a dps increase.

fenix_jn

join:2006-12-28
Miami, FL
Reviews:
·T-Mobile US
·Comcast
·Sprint Mobile Br..
·AT&T U-Verse
reply to Pollux7777
downloadT1.html.zip 200,107 bytesdownloadT2.html.zip 200,417 bytes
Well simulcraft @ 10K iterations tells me that with your current trinket you'd output 86560.5 dps while with the new trinket (skullrender medallion) your output would be 85316.7 dps. I have attached the simulcraft result pages: T1 uses your current data while T2 uses the new trinket.


Goldheart

join:2002-06-09
Ballston Spa, NY
kudos:1
reply to Pollux7777
Just guesstimating I think the shieldwall would likely be better.


Pollux7777

join:2010-02-16
Saint Paul, MN

1 edit
reply to fenix_jn
Wow, thanks fenix. I was not expecting those results.

I'm on my iPad right now so I'm not able to open those files. Just to clarify, in T2 you used the helm breaker medallion, right. I'm assuming when you wrote skull render medallion that was just a typo.

Or is there an item called skull render medallion?

EDIT: never mind I just looked it up. Skull render is just the horde version of the trinket.
--
Polluxx - LvL 90 Human Warrior - (US) Ravencrest

fenix_jn

join:2006-12-28
Miami, FL
Yeah, my bad. I just looked you up again and noted you were alliance. But yes, the data is the same (redid it with the proper trinket produces an output of 85240 dps so about 100 dps loss which is likely crit chance/procs)


Venom1

join:2011-10-21
Augusta, GA
reply to Pollux7777
i'll buy the shield wall one and trade you, crit is lowest priority for ret after spirit /sigh


puppy

join:2010-01-28
reply to Pollux7777
So basically you want to compare 2866 str for 20 seconds out of 45 seconds, or 1152 strength all the time?

Assuming 100% uptime (trinket procs as soon as the cooldown is done, which at 15% proc rate is a very optimistic assumption even with your fast swings), you would get an average of (20 / 45) * 2866 str = 1273 strength.

This is not triple of 1152, just barely a bit more. While I like proc trinkets (woo icon, snazzy effects), for this little difference I would take the 1152 str all the time. 100% uptime is not possible, and often you can waste a big portion of a proc if you are getting out of fire.

If this was an on-use trinket then you have a closer argument since you can time it for burn phases/bloodlust/etc.


Pollux7777

join:2010-02-16
Saint Paul, MN
I get your logic puppy. Although wouldn't the actual equation be 20/65? Since its 20 secs, then a 45 sec cool down, for a combined total of 65 seconds?

Correct me if I'm wrong there.
--
Polluxx - LvL 90 Human Warrior - (US) Ravencrest


Caelharrad

join:2012-04-13
Fenton, MO

1 edit
Pollux, you're closer to correct - there's a 45-second internal cooldown, then you have to wait for the trinket to proc, then you have 20 seconds of the effect. (I'm pretty sure that the ICD doesn't start ticking until the proc effect is finished; if the ICD starts ticking as soon as it procs, then puppy's math is correct.) So, you could use 20/65, but that assumes the trinket procs INSTANTLY once the 45 second ICD is finished. The actual equation would be 20/(65+x), where "x" is how many seconds it takes for the trinket to proc once the ICD is up.
...
If you do the math with x=0 (instant proc), it comes out to 881.8 strength. If x = 5 seconds, you're looking at 818.9 strength. Your actual number is probably somewhere within that range. In any case, the Shieldwall trinket is likely to be better by a good margin.


Pollux7777

join:2010-02-16
Saint Paul, MN
Good post Cael. I think the best thing at this point is to take cptbeatstix advice and run some samples with each. I can always return the item if its not working out. Thanks for all the replies fellas.
--
Polluxx - LvL 90 Human Warrior - (US) Ravencrest

cymraeg
Thread Killer
Premium
join:2011-06-07
Dodge, NE
most of a warriors damage is during the CS window and execute phase, much rather have constant strength, than a maybe proc during those times, plus the added on use crit from the shield trinket is a nice boost during the execute phase with banner and reck.
--
Hen Wlad Fy Nhadau!
"What happened in here?" " Seems to have been a bit of a squabble." "They're all dead!" " More of a tif then."


puppy

join:2010-01-28
reply to Pollux7777
I haven't played WOW since Panda came out, but I seem to remember ICDs start as soon as they proc (counting down while the proc is happening), so it would be 20/45. That's how I always calculated it in the previous expacs. If it's 20/65 then that's even worse :P


puppy

join:2010-01-28
reply to Pollux7777
As for cptbeatstix's advice, just be aware that 5 samples is too vulnerable to randomness. Maybe you didn't hit buttons fast enough on one try, or a string of crits in another. That's why people use those simulators, where they do 5-10K trials, to even out the randomness.

Cptbeatstix

join:2011-12-21
Carrollton, TX
Reviews:
·Verizon FiOS
reply to Pollux7777
You can never really judge off just 10 dummies. I've had times where my Lei Shi trinkets constantly proc'd during a 7 minutes LFR fight. Other time it proc'd maybe 2 times the entire fight. Best advice, Test it out on LFR on one week and then the other on a second week of LFR. You might want to consider trying arms next patch, with the rework of Taste for blood, it may improve the overall dps since it will be a guaranteed 2 overpower per MS (five when max stacked). I do like the way its designed now though. 500% increase to Heroic strike. One shot macro + banner = 400k+ hits. /cast chubby.