Snapshot ID: a280ca992105431c933a068cb341a4d7
When I click on the neck item to open the necks list, the Elemental Lariat with item level 486 is missing.
You can choose any variation of an item in the lists: press the icon to the left of the rank to show controls for choosing variants, then choose the item level (and any other properties you wish), then press Apply. The new version will appear in the list, click it to equip it.
That the Elemental Lariat 486 is not in the list seems to invite the question if it was considered. Because Elemental Lariats with lower item levels are in the list, but the 486 one is absent. Would be great if it would be considered and in the list if it’s considered.
Here (snapshot id: bd128fb235e647a5963628d623b02cd5) it still suggests Ironshell Pendant as neck even though Elemental Lariat 486 with Crit and Haste (and 3 sockets) seems better (snapshot id: 82a3f114322844b3b94c4396c63f192e).
Is that because the algorithm only searches in a subspace and therefore misses a more optimal solution that is outside of the search space? Or in other words it considers only certain configurations of crafted gear instead of all?
The crafted filters determine the quality of crafted items in the solution. As you can see in your snapshot, it is choosing other ilvl486 crafted items. The per-slot item list is a generic list independent of the Best in Slot optimization itself – it is just there for you to browse other gear choices in that slot. With your settings, it will only try the ilvl486 lariat with 3 gem slots.
In your screenshot you were modifying the ilvl463 version (which can only go up to ilvl476 at crafted tier 5). You would need to modify the ilvl473 version to get up to the ilvl486 version.
We don’t try every variant of every crafted item for Best in Slot – there are too many. They can have different stat combinations, embellishments, etc… would be too slow. The lariat effect itself introduces a dependency that significantly slows down the optimizer as well. We aim to find a solution that is within ~1% of the theoretically optimal possible solution, usually closer. This case happens to be right on the edge. Sometimes you can lock in a specific item and find that extra 0.5% here and there. But sometimes… the optimizer finds a solution that most people would not have considered that is slightly better. On the average things even out.
For all practical purposes, you won’t be able to tell the difference between those two solutions in-game. They are too close to tell, and no model of the game is perfect enough to tell you definitively one way or the other (and there’s no good way to empirically make that decision in-game). So use whichever you prefer.
Thanks for the explanation. I first have crafted what Ask Mr. Robot has suggested. After taking a closer look at the effects it seemed easier to maximize the effects of the Lariat, Wrist combo. Because as far as I get it, the head requires target swapping regularly to maximize its effect and the boots required moving as much as possible to maximize its effect.