Thank you for the update and the detailed patch notes guys. Much appreciated!)
mini note: unedited notes also include changes from previous update
What is the auto-attack setting and how is it broken?
There were two things: you can't auto-attack a team member, and this
Attribute Cap of 35 - To prevent the effects of Attributes from scaling away into infinity, they are now capped to 35. We chose this number based on both where we wanted the influence of Attributes to max out as well as what we believed was within the reach of aggressive min-maxers.
Feeling strange about this one ^^ because:
- a cap of 100, would achieve the same thing, i.e. prevent "from scaling away into infinity"
- in PoE1 we could achieve up to 43 dexterity; although do note: it was not worth it
- hard caps do smell bad. Sometimes it's indicative that the game loses it's balance outside of certain range; and that the system can't handle it gracefully itself. No hard cap in PoE1 was a healthy sign in my book, and it was really well handled, as dropping attributes was stinging, while maxing them to the roof was affected by intrinsic diminishing returns; not to mention that there were often wise to smooth the spread a little: think of MIG and DEX both contributing to dps, and at a very simplistic level: 15x15 > 20x10.It's no big deal though)
I was a lilttle skeptical of the attribute cap at first but 35 is pretty high -- that's about where the "soft cap" of effectiveness was anyway in the first game. If it were lower I'd have more concern but at 35 it seems almost more like a UI feature -- "it's ok, you can stop bothering now" -- and it might make character planning a little simpler and give the designers more room to play around with weird bonuses (like drugs, etc.)
I mean I see where you're coming from and it's definitely the most interesting change listed but it might work out to be a good thing iafter all.
I'm almost positive the cap was introduced just as an arbitrary restriction so that Resolve could be allowed to reduce hostile durations meaningfully per point without worrying about degenerative cases of high resolve (which JE Sawyer cited as a reason of why he was hesitant of resolve doing this effect). They could theoretically have gone up to 43 (any more and resolve would give >100% reduction), but i'm guessing the way resolve is implemented there are significant increasing returns (each additional 3% reduction is more of the remaining hostile duration). I haven't tried the new BB yet, so this is an assumption based on how intellect works.
EDITed to add: "while maxing them to the roof was affected by intrinsic diminishing returns"
Not actually true. By my calculations, while perception had diminishing returns, might, dex, int and con all had linear returns. Resolve actually had increasing returns, because deflection had increasing returns and resolve boosted deflection (similarly all the other stats have slightly increasing returns because they all boost a defense of some sort). As someone who did both triple crown solo and the ultimate achievements, the difference between a 19 resolve and a 20 base resolve can be a run-breaker (more so than 18 vs 19). Now, it is possible that to achieve really high levels in the stats you had to sacrifice a bunch where the net effect was diminishing returns, but the stats do not have "intrinsically" diminishing returns.
historical fact: back in the day intellect used to boost radius instead of total area. that was stupidly increasing returns, which they eventually fixed. (hint: area = pi*r^2)
Edited by thelee, 30 March 2018 - 12:41 PM.