Yes, but you optimize to overcome barriers to progression, not to farm faster, unless you have everything in the game on farm. Besides that, though, Karazhan is really foreign to 25-man raids and is completely different in a lot of ways(for one thing, you are never ever in a properly optimized group, whereas you frequently are in 25-mans).

EDIT: Sorry if this is kind of a derail :/

I used to take hit as a one-roll thing like melee (30% crit 10% miss means you're left with 60% hit) but reading around seemed to implement that it's 10% misses then 30% of the 90% hits can crit. I really don' t know what's true though, and it REALLY changes the difference between hit and crit. If it's a one-roll thing hit value is a lot less valueable, but still 1 hit rating is more than 1 damage, while 1 crit rating remains around the same of 0.73 spell damage.

As for non-boss mobs, I only care about bossfights, so I went and looked at all the bosses from magtheridon and on, and averaged the boss mob HP to non-boss mob HP based on few vids and bosskillers guides, on all the bosses that have any sort of info. This is very rough guestimation, but what I got was ~80% of your damage being dealt to bosses, which means you multiply the value of +hit by 0.8, bringing it for me from over 1.6 damage per hit to 1.33 damage per hit. Of course if anyone has more accurate %age of boss VS non-boss on boss fights I could use I'd tweak my numbers over ;p

As for gearing differently for different fights, while optimal, usually the main place you get to actually choose stats in is your gems, and you're not going to replace those every fight, it's just not worth the gold. So being the best you can on "average" is probably the best thing. If you're really really hardcore, you could tweak your gems for the fight you're actually working on... But really the difference isn't big enough if you max yourself out to be the "best on average".

for heroics/5-mans/solo, well if you're at the point where you're maximizing your raid DPS, you probably have more DPS than any tank can hold without full raid buffs and a heavy hitting mob supplying him with unlimited rage, plus you have more than enough DPS to blaze through any heroic of your choice... Again if you really wanna be hardcore you could have a different set of gear, but gimping your raid DPS to up your solo/5-man DPS is very not hardcore...

For trash clearing, not only you should have more than enough DPS even when you waste stats on +hit, clearing speed on trash has many factors aside from your DPS. A significant % of the time you spend waiting for people to drink, be ready, or just move around and get the mob pulled or wait for the pat to be at a good spot. So increasing your DPS by 10% is going to increase your clearing speed by a lot less than 10%. While increasing DPS on bosses by 10% makes them that much easier and less likely to wipe you, wasting a lot more time than an extra 10% trash DPS would save you. Just like everywhere you could go hardcore here and swap accordingly, but again you'll probably be wasting your time. And again optimizing for trash while gimping boss capability is nothing hardcore at all.

The solution to this is simple: carry additional gear and use the gear you deem appropriate for the situation. Swapping out a +hit item for +dam and/or +crit is helping out the raid on trash, as is putting the +hit back when you need to fight the boss.

Good example is anyone using Scryer's Bloodgem (it's a trinket, I'm on topic!) for bosses, and some type of +dam trinket on trash. That trinket is exceptional for those who lack the +hit, but useless on trash (the on use is beneficial, but there are better trinkets for trash purposes).

Where's the synergy with combustion considered? I can say with a decent amount of comfort that saving the mana gems for Combustion (or popping them every time and getting every other combustion), in addition to popping something like Scryer's Bloodgem, both at the same time Combustion is up lead to much more synergy than any of them alone.

It also hurts trinkets like Mag's Eye, as unless you have insane up-time on it, when you pop combustion you don't get the extra burst a burst trinket would give you.

Now a question i have for a math whiz out there is whether it's better to just pop your Gem every 3:15-3:20 minutes when you use combustion, or just blindly pop it every time (which could take until every 4th combustion to pair again)

If you only pop your gems every 3:15 minutes, you're better off with another trinket in my opinion.

That being said, let's look at the numbers with 33% crit:

If you pop gems every 2 minutes, affecting X spells, with X/3 being crits, and 2X/3 hits: If the trinket increases your the damage of 1 spell by A, you gain [1*A*2X/3 (from hits) + 2.1*A*X/3 (from crits)] every 2 minutes.
That makes it 4.1*A*X/3 in 2 minutes, 0.68*A*X per minute.

If you pop the trinket every 3:15 minutes, and all spells crit during combustion (best case scenario), then you'll gain [2.1*A*X] every 3:15 minutes.
That makes it 0.65*A*X per minute, assuming that all spells crit.

So, not really worth waiting for combustion.

Combustion doesn't increase the effect of your "use" trinket by enough to make it worth not using as often as the cooldown allows. If one is almost up when the other is you can wait, but even then I would think twice before deciding to not use them as soon as they're up.
Combustion crit increase over time is something very hard to calculate though.

First of all, nice thread, great to see different opinions on what trinks to use.

Im a non hit capped mage with a build including Slow. I am geared more for the 10-48-3 build, and will change it - half i like this spec, and half "need" it to trivialise Moroes, since my little guild doesnt have tanks to spare for the women that hang with Moroes so we use slow to zerg em.

1 little thing : mages with Arcane Power, on many encounters especially where threat isnt an issue, benefit alot more from activateable trinkets. Unsure if this has been stated before, if so sorry, skipped some of the more mathematical walls of text
Increasing spell dmg with trinket, and then using a %-increasedmg ability makes that increased damage worth alot more.

Otherwise, i find most trinkets handty in some situations. Keeping Badge of Endless Blessings in my bag still

Just like combustion, AP doesn't increase the benefit of trinkets enough to matter in most comparisons you'll make.

I think AP is kind of a wash. If the trinket has a 2 min CD and lasts 15 sec, you lose 33% uptime in exchange for 30% more benefit.

If you also sync'd up bloodlust or AB spam, saving such a trinket would be beneficial i think.

I've read through this thread and haven't really found the answer that I'm looking for. For my first trinket I'm using Mag's Eye and am wondering what my best bet is for the second. I've already ruled out SC-Braid because I'm not interested in giving myself anymore +hit, especially when there are other viable trinkets available to me.

I'm currently 10/48/3 and average around 40% crit during a boss fight with a fireball/scorch rotation at around 1000 spell damage. The two trinkets I'm really debating against are TLC and the Sextant. I've heard that the Sextant has a hidden 45 second cooldown and if that's the case I feel like TLC is a better choice. Keep in mind I have low hit because I'm using Mag's Eye and that I'm only concerned about my boss dps.

From what I can tell it's really kind of a toss up, leaning slightly towards TLC if the hidden cooldown exists or towards the Sextant if it doesn't. Does this sound right? and if so does that cooldown exist?

Just because you have Mag's eye doesn't mean you can roll with a low +hit.

If you're concerned about boss DPS, bump up that +hit - use the Serpent-coil Braid or if you have it, a Neltharion's tear.

Even with a hidden cooldown, you can see that the Sextant adds quite a bit of DPS. If you're straight fireballing, TLC is good, but nothing special.

Considering you have quag's eye and the icon, i would stick to that and just get more spell hit, and swap the icon for nef tear if its a possibility to gain more spell hit...

Has anyone done any math on the Ashtongue Talisman of Insight(http://www.wowhead.com/?item=32488)? Trinket seemed okay in theory, but I went over to Dr. Boom and tried to parse some data, and the trinket seems quite bad.

Anyone have any numbers regarding this trinket? It has no hidden cooldown, and thus can proc back to back if that helps anyone. Thanks!

Antale, can you ask your other guildies if their Ashtongue class trinkets have no cooldown and can proc back to back as well?

TC on mage trinket:
- 5s proc duration, 145 haste rating ~7% haste ~ 2.8 s fireballs - it will affect 2 fireball casts. (Haste affects the spell when you cast it, so it doesn't matter if the proc runs out while casting or not, you already got the haste benefit. Unlike +dmg procs, which need to last until you end the spell cast.)
- So, any fireball cast is affected, if the trinket procced on one of the last two casts. Assume 40% crit rate (generous), it's a 20% chance to proc, 80% to not proc per cast. 80%*80% = 64% to not proc during both casts, 1-64%=36% chance to have the proc when you cast the spell.
- So, 36% chance to have a 7% haste proc ~ 2.5% haste
- Assuming mana is not an issue (haste is rather pointless if you're mana limited), 2.5% haste ~ 2.5% hit ~ 40 spell damage, and about 0 if you're mana limited.

Yup, sounds pretty junk

Edit:
If you can stopcast your spells really really well, you can squeeze 3 Frostbolt casts out of 1 proc, bringing you to 48% uptime ~55 damage.
For Arcane Missiles (and ~30% crit), you get 55% uptime, or 80% if you can squeeze 2 casts in, bringing you to 65 and 95 damage respectively, but squeezing missiles will most likely give you clipping issues, and haste means even more mana burn on missiles.

For what it's worth, I ran a bunch of numbers on this a few weeks back when I first got the trinket.

With my gear and 10/48/3 spec here's actual numbers and some theorycrafting from real runs on Grul + Mag:

http://www.lossendil.com/wws/?report=rshwa6dkomstq&a=20

**Note, these numbers are based on a single night of raiding, and the theorycrafting is based on those number **


427 Crits = 142 Procs, actually got 133. 96.7% (You lose on procs when your spell crits and the mob is dead before the Lighting bolt lands)

The new buffed proc averaged 826 damage factoring in misses.

So each crit was worth = 826 / 3 * .967 = 266.

Fireball spam: 266 * .29 (shity crit rate that night, usually around 35%) = 77.14 damage. Hard to compare to a trinket, but here's a shot:
A regular 77 damage trinket is worth: Fireballs receive .90% from +damage (assuming improved FB) = 77*.9 = 69.4.
Crit is ugly considering the mob has to stay alive for the ignite to tick, but lets assume it does:
Crit is then worth an addition 1.1 = 69.4*1.1 = 76.36 * Crit chance .29 = 22.14 for a total worth of 91.57. Same as above, have to factor in debuffs on the mob that do not effect the Lighting Cap (Scorch and CoE) would add another 10% or so damage (scorch is not always up) ~ 100.7

Now have to take out misses and partial resists.

My miss rate was 6.7 for fireballs, which gives us 100.7 * .933 = 93.9. My partial resist mitagation was another 4.2%. 82.5 * .958 = 90.03

So a regular +77 damage trinket is worth +90.03 damage when spaming fireballs with my current gear/ build.

Scorch spam: 266 *.36 = 95.76 damage. 95.76* .4286 = 41.04. Working crit the same way as above:
41.04 *1.1 = 45.14 * .36 = 16.25 for a total worth of +57 damage. Work in the debuffs and it would add in roughly 6 more damage for roughly + 63 damage trinket. After miss rates (4.6) and partial resist (4.1 mit), its worth roughly 57.6.

So in summary, the Lightning Cap is worth a flat +77 damage on fireballs and 95.76 damage on scorch for that night. If I'm just spaming fireballs, to replace the Lighting Cap I would need a trinket that is +66 Damage.

When spamming Scorch, I would need a trinket that is +169 Damage to replace it.

Since I'm usually stacking Scorch for the debuff, the value of the trinket is somewhere in the middle. Assuming a 7:1 ratio and the initial stacking of 5 and reapplying when the stack is lost (silenced, thrown etc), I usually cast scorch 4 to ever 6 fireballs throughout an entire run, putting the trinket worth roughly +108 damage. Given, I cast scorch more then most deep fire mages, I find myself spaming it when I'm trying to watch other parts of the encounters etc.

2.5% haste ~= 2.5% DPS increase (or exactly 2.5% if that's your only haste trinket and you only cast spells affected by haste when it's up), and according to my spreadsheet 1% DPS is20.5 spell damage when I'm fully buffed and flasked with wizard oil and well-fed. Actually my spreadsheet is missing some buffs but they're neglicible, probably won't change it to any more than 21 damage per 1%. So 2.5% DPS is 50-52.5 spell damage - still not good enough. Wowhead is down right now to verify if it has any other stats but if it's only 2.5% haste on average it is kinda crappy. Not as bad as 40 damage though in terms of dps.

I wonder if it was tuned before they nerfed spell haste?

All spell haste items seem to have been. Quag's was overpowered and rather than fix that 1 item they fucked every other spell haste item in the game.

Regarding Quag's Eye:

This is the way i was modeling Quag's Eye too, but i was just thinking...do we really cast 3 hasted fireballs?

If the proc is on hit, we've already started our next cast before the buff is applied and will only have two hasted fireballs.

If the proc is on cast, does it affect the current spell (would give 3 hasted fireballs) or the next spell (just 2 hasted fireballs).

The proc is on cast not on hit (polymorph and AOEs that aren't hitting anything also proc this as far as I noticed). However it's hard to tell if the first spell is affected by the haste - but it should be.

Stein: Thx for the info and hard work. Your short changing the Lightning Capacitor by a good bit though. The average damage is a lot higher then your formula of 750. I've used it for the last few weeks and had an average proc damage of 892 (Maulgar + Gruul), 923 (Mag), 883 (Lurker), 905 (Hydross).

In regards to the trinket having worse resist rates. They are a bit worse, but nothing crazy. For the above raids, the trinket was resisted:

Lurker: 3.7 Miss 3.0 Mitigated
Mag: 13.2 Miss 3.5 Mitigated
Hydross: 3.0 Miss, 2.5 Mitigated

After factoring in misses and mitgation, I figured the trinket to be worth around 825 per proc.

If you wish, you can view a bunch of WWS files of myself with the trinket on here: http://www.lossendil.com/wws/?guild=htrtxvqu2gfx3

Yeah, i wasn't giving it the benefit of +hit gear or % based talents (leads to the discrepancies i believe). At the time, it wasn't clear if it benefited from either. I'll try to redo the math and bring it closer to reality.