Talk: 40/40

For anyone who's brain isn't currently molten

[1]. Which is correct? To me, it makes no sense 2 20% HSRs are different from 2 20% HCTs, and thus should have the same odds for 1 of either proccing. --- -- (contribs) (talk) 15:59, October 11, 2009 (UTC)

I agree with the original, though my brain is slightly molten as well. The chances of both failing to proc should be 80% x 80% if I'm remembering my probability math correctly, and that comes to 64%, and 1-p here is 36%. The chances of both procing would be 20% x 20%, so that would be 4% chance of quartering casting time. So unless my brain is failing me at the moment, the original is correct. Nwash 16:10, October 11, 2009 (UTC)

The HCT and HSR triggers should, in theory, be independent from each other.

For each dimension (casting or recharge time),

Wand\Focus	Triggers	Does not trigger
Triggers	Quarter time (4%)	Half time (16%)
Does not trigger	Half time (16%)	Normal time (64%)

• Quarter time odds: chance wand triggers * chance focus triggers = 20% * 20% = 4%
• Normal time odds: odds that nothing triggers = (wand fails trigger) * (focus fails trigger) = (1-20%) * (1-20%) = (80%^2) = 64%
• Half time odds: odds that either one (but not both) triggers, for each it's: (chance the other fails) * (chance this one triggers) = 80% * 20% = 16%
• Consistency check: 4% + 16% + 16% + 64% = 100%
• Gets more complicated if you want to consider getting HCT and HSR benefits from same spell.
• Finally, recharge times can't go lower than 50%, so 2xHSR ends up providing HSR 36% of the time (32% + 4%) while 2xHCT provides HCT 32% and QCT 4%

So, technically, a 40/40 set ought to be called a 36/36 set.   — Tennessee Ernie Ford (TEF) 21:20, October 11, 2009 (UTC)

Ah, now I see where the % difference comes from. It has little to do with odds; rather, it has to do with stacking. You can't benefit twice from a HSR proc, so effectively dual procs and single procs are equal. 16+16+4=36%. --- -- (contribs) (talk) 21:26, October 11, 2009 (UTC)

Stack cap? (follow-up of historical note)

There's a stack cap? Does that mean that 40/40 sets, which are technically 32(4)/32(4) are nerfed to 36/36? (For what it's worth, I think the historical note is fine; I doubt it can be compacted much.) If that is how the cap works, then we should also update the rest of the page to reflect it.   — Tennessee Ernie Ford (TEF) 23:16, October 11, 2009 (UTC)

Erm, I'm looking at the update and I can't actually find a reference to effect stack capping. Are we sure that this was the correct update? And that we've correctly explained what happened? The official wiki says that casting time can be quartered, but skill recharge time is limited to 50%. In which case, these are pedantically 36/32(4) sets.   — Tennessee Ernie Ford (TEF) 23:24, October 11, 2009 (UTC)

Game_updates/20060713#All, look at the red link :) --- -- (contribs) (talk) 14:14, October 12, 2009 (UTC)

It reads a lot better now, thank you TEF. --Macros 16:49, October 12, 2009 (UTC)

Great! appreciate your saying so.   — Tennessee Ernie Ford (TEF) 17:22, October 12, 2009 (UTC)

I'm entirely confused as to how the math kicked out a 32/4/36 for the math, but I can attest that after having spammed about 100 2-second cast spells, I was unable to get a .5 second cast. So no quartering the cast time anymore. As I don't understand the whacky resultant numbers, someone else had best ammend the article (because despite the note, the page still mentions quartering cast times). Skippster 06:46, May 23, 2010 (UTC)

Edit: Haha, just noticed it was about recharge (this discussion). I still don't understand the math, but someone who does needs to ammend the math and the article as quarter-casting doesn't work. Skippster 06:48, May 23, 2010 (UTC)

There was a game update that changed how these stack only a couple days ago, so the math was correct. (And technically, it still is correct in that there's a 4% chance of a double proc. It's just that a double proc now has the same effect as a single proc due to the stacking cap, so you wouldn't even notice it.) I'll amend the article to reflect this. Nwash 07:27, May 23, 2010 (UTC)