Regular site commenter Steven Hazani sunk 50 hours into testing on the PTR, and he’s filed a lengthy report detailing his opinions on everything. Topics include Grift density and balance, item progression, gold and legendary item rewards, ancient item upgrade issues, experience and lacking paragon point bonuses, and much more.
Here’s a very partial quote with some conclusions, to give you the (highly-critical) flavor of his opinions. Guest Article: Diablo 3 PTR Mega-Test Report
The rewards of course are still atrocious, ignoring test realm specific things. And that’s my first point. See, when this buff originally went live I was mid GR 36. Result? At the end I got 6 legendaries (because GR 3x guardians only have 6 item drops, and the buff makes every item that falls legendary). That’s how it should be. All the time. Because in that same length of time I could have done 2 or 3 rofflestomp runs and got the same or more legendaries without the considerable risk, so 6 legendaries would make it slightly worse instead of considerably worse.That’s a huge step in getting some risk vs reward in the game. I also got 300 something shards. Which is also how it should be all the time since that’s about how many I’d get off normal farming for the same timeframe.
Leaving the enhanced drops and shards for higher GRs only would go a long way in correcting one of the biggest problems in the game. Because guardians have a cap on the number of items they drop it’s not abusable.
If more item types were actually useful and there were not so many mandatory items the current live drop rates would feel much more reasonable.
I probably smashed at least 20 Ancient items that were of the correct type (read: the 0-3 items per item type anyone actually wants) because they had 2+ stats incorrect, and were therefore not actually better.
If Kadala could fix more than one stat, the current live drop rate would feel much more reasonable. This would also let people burn those Hoarder stashes and saved mats, and it’d still be insanely expensive if you had 3 stats wrong but that’s better than finally finding rare item x after 600 levels and then getting terrible stats on it. Alternately you could make more than a few stats “correct”, but this is Blizzard so let’s have reasonable standards.
1: GR guardians should flat out drop all legendaries/sets with a number of slots based on level beginning at level 26. Remember, they take longer so even more legs/run does not necessarily = more legs/hour. 2: Double shards in GRs only beginning at level 26. 3: Trial rewards should be the same as current test realm GR rewards. 4: GR experience rewards should be the same as current test realm trial rewards. 5: GR gold should be current GR values x3. 6: 1-5 are lost if you die (death means you get the current worthless live rewards). This is because SC has no death penalty now, and will otherwise death zerg for lots of free stuff. 7: If we must stick with the current model than make all those unused legendaries supercharge skills people aren’t using, make the base skill good (don’t expect people will take Magic Missile/Shock Pulse/Energy Twister/Arcane Orb seriously), and make sets not mandatory for creating a functional character. Suddenly there’s a lot fewer auto souls. Ideally we scrap that model and then nerf enemy life as no one will be doing 10,000+ weapon damage anymore as skills are somewhat well balanced with themselves, the lack of build diversity comes from only a few narrow legendary gimmicks existing and those being the only way of dealing with the insane life on everything. But again, Blizzard. 8: If you find an Ancient it should never be bad. Make it always have the correct stats, make it always max stats, or perhaps best of all scrap the current implementation and just give us Ancient Gifts. It’s still just MOAR NUMBERS, but at least we ditch the RNG trolling and don’t automatically invalidate every existing item. 9: Make Kadala change any number of stats, the cost still scales based on cumulative changes. Fixing an item with 3 wrong stats will be expensive, at least it can be done.
The current live drop rates of roughly 3 legs a run would be a lot more reasonable if you didn’t smash > 99% of them. This also fixes risk vs reward by making high level GRs done well more productive than rofflestomping trivial content, which makes the game much more enjoyable.
See his full post for much more commentary on practically everything, and there’s a lot more in the follow up comments and rebuttals in the thread. I’ve done a lot of PTR testing this time as well and while I disagree on many of his points, I think others are right on.