New patch soon?

[MaxWilson]

Originally Posted by AreaOfEffect

Even all of that still doesn't explain why some locations who don't have any magic sites found are skipped yet contain magic sites when manually selected and searched. I suppose this will still occur even after the next patch is released.

Sure it does, unless there are facts you're not telling me. Consider a hypothetical province Green Gallows, which has zero discovered sites, and one undiscovered level 3 death site. If I have searched manually with a D2 mage in Green Gallows, when I monthly-cast Dark Knowledge the automatic site-searching algorithm I outlined will skip Green Gallows because it's been partially searched already for D. The patch enhancement will make it go back and search Green Gallows after all the "prime" locations have been searched. This makes sense because a totally "fresh" province with no discovered sites and no Death searches yet is more likely to have a Death site than Green Gallows is (because a fresh province may produce a D1, D2, D3, or D4 site, whereas Green Gallows will yield only a D3 or D4 site. The number of discovered sites matters because it's relatively improbable to discover a D site in a province which is already known to have e.g. 2 air sites, since the average total number of sites in a province is only about 2 anyway, depending upon what site frequency you are playing with and what the terrain type is). If you crank the math, the site-searching algorithm is already a pretty good rule of thumb for searching. But Green Gallows is not a worthless province even if it is less probable, and in fact we already know that it *does* have a D3 site, so the enhanced algorithm will eventually get around to finding the hidden site.

-Max

[AreaOfEffect]

I'm apparently much better at reading your words then you are at reading mine. This is the part of the formula you say answers the question.

Originally Posted by MaxWilson

The number of discovered sites matters because it's relatively improbable to discover a D site in a province which is already known to have e.g. 2 air sites, since the average total number of sites in a province is only about 2 anyway, depending upon what site frequency you are playing with and what the terrain type is.

This is the part where I insist that this is not the case.

Originally Posted by AreaOfEffect

Note: The cases listed below are bypassed by all site searching spells, not just one or two. This tends to be a trend with this particular bug. They have also always been this way. Both before and after sites were found there. If they were searched by a spell it was manually selected.

I am 100% certain that site search was switched off for Tapanete (120) and Aenasia (294) before sites were discovered. I didn't bother to present any other locations because it is tedious and because I wanted to use example locations where sites were found spite the games desire to pass them up.

[MaxWilson]

Originally Posted by AreaOfEffect

This is the part where I insist that this is not the case.

I am 100% certain that site search was switched off for Tapanete (120) and Aenasia (294) before sites were discovered. I didn't bother to present any other locations because it is tedious and because I wanted to use example locations where sites were found spite the games desire to pass them up.

I'm sorry to have pushed your buttons. Please remember that you have an order of magnitude more data than I do about your situation; I only know what you post. The algorithm still obviously applies to Province Dumna, because it has two sites and there is no "before sites were found" since you never searched it (it's 0:0:0:0:0:0:0:0:0). Province Tapanete, Hamane, and Aenasia: you're saying that these were *never* auto-targeted even when there were zero known sites (1 known site in the case of Hamane). I would write that up in the bug discussion thread.

-Max