Finding frequent issues with Upgrade Finder recommending particular items with Crests/Flightstones and after upgrading those items, not being recommended in Best in Bags.
See Snapshot below:
For Preservation (my m+ build), it recommended that I upgrade a Snipping Sleet Circle ring for Finger 1 and would be a 0.8% upgrade. Yet after upgrading BIB says it would be a 0.11% downgrade. Now it says another ring would be better than what I just upgraded.
It is recommending several items to upgrade and many of them I know will not be upgrades afterwards. This just leads to incorrect recommendations and wasted Crests/Flightstones.
I have had this happen on multiple toons. Please review.
a3abfe1ffba444f3bf2295d0e3416d5d
Just did this again on my Shaman - Resto; spec’d for M+
f01b2e662c1d482291d02b7473348517
Had 3 top items to upgrade using Whelp Crests (450 ilvl)
- Summer Soother’s Wrap - 1.68% upgrade or so
- Bramble Bark Ring - 1.3% upgrade or so
- Paracausal Fragment of Val’anyr ~1.x% upgrade
After spending ~500 flightstones to upgrade these, they now show as a combined 4%-6% downgrade.
I am not sure how they are being calculated to predict or recommend upgrades but something is clearly not working here. This is my primary spec/build with no custom settings.
I’ll see if I can reproduce the issue and take a look at it sometime this week or early next week.
I think that I was able to narrow the issue down… still working on it though. If you disable the “force tier set” option then the rankings should look better. There is some interaction with that setting and the upgrade finder that is making things difficult.
We did an update last night that should improve this issue – specifically that the Flightstone/Crest search was having trouble when the “force tier set” option is enabled for the active Best in Bags setup.
Please note that all upgrade finder searches have a small margin of error – they won’t exactly match the result you would see were you to obtain the item and run Best in Bags, but should be very close. The reason is performance – if you have a list of 30, 40, even up to 50 upgradeable items as people often do now, it would take way too long to run a full Best in Bags optimization on each. So we use techniques that are more thorough than the basic upgrade finder searches, but still not a “full-on” optimization.
That said, it’s good that you posted – the differences should be tiny, not large like you were seeing.
Appreciate it. I will try it out and let you know if I encounter any more issues.