It’s hard for me to comment from screenshots – if you post a snapshot via the “help” link next to the big Best in Bags or Upgrade Finder section headers and copy the 32-digit ID here, I can use that to replicate your case and see what’s up.
Thanks – I’ll fix that in the next update. It looks like it was mistakenly activating the set bonus at 2 pieces instead of 3 for that set.
edit: Note that I’m not exactly sure when we will get the next website update. There are just 2 weeks until the Midnight expansion launch, so we have a temporary code freeze on the main site right now as we are doing testing leading up to that. I should be able to do a small update in a couple of days with a few bug fixes. In the meantime you can lock in all 3 pieces of the set to get around the issue.
Regarding case 2. Are you sure that arcane/fire spell damage is correctly evaluated? Even without the setbonus the hands themselves offer significantly more damage than the optimizer suggests. And the off-hand upgrade recommendations don’t show 50 arcane spd + 14 int at all, even though it’s considered to be pre-raid BiS by pretty much every bis list.
We do calculate and apply school-specific spell power in the optimizer.
I can ask Swol (who wrote the TBC mage model) to comment more on why it might be making these specific choices, but I can at least verify that it is adding up spell power correctly and applying the proper spell power type to the proper mage spells.
As a follow-up… we have a pretty in-depth model of mana requirements for all specs in TBC. Usually the main reason you’ll see a different recommendation on our site is because you are running out of mana.
Try shortening the fight length – it’s OK to go really short. Then you might see more of the results that you want. Often different boss fights in the game will have downtime where you can regenerate some mana or can’t cast.
Yes, it is being handled correctly. As yellow mentioned, mana plays a big roll in the optimization for TBC (and arcane mages in general).
If you become mana-limited as an arcane mage with the build you have, you would swap in frostbolts as the alternative filler to save mana. Doing this will, of course, make arcane-specific spell damage worse and worse.
Think of the optimization as: this is how to optimize the damage I do for the given length of time, assuming I never stop casting. Set the “fight length” to the amount of time you want to be completely active. Maybe you are only interested in knowing how to maximize your damage as arcane up until you are totally out of options to get mana back.