Wrath of God targetting question

[Ironhawk]

So, in my past two games, one of the players has cast Wrath of God. It only seems to kill a handful of units so I basically ignored it. But two times now, the lightning bolts from it have *just happened* to kill the most valuable mage in my capital. The first time I wrote it off to unlucky coincidence, now I am not so sure. Does anyone else have any experience with how this spell targets your units? If it chooses powerful mages/priests over others? Or commanders over troops?

Additionally, if anyone has any clue how to find out which units have been killed instead of the completely useless "3 units were killed this season" message, I would be much obliged.

[Agrajag]

I *think* the target selection is completely random, I've had many cases where for several turns I had a lot of units killed by "Wrath of God" and not a single commander died (and in each province with any commander, there were plenty of troops).
Though I don't know how many troops died, its so hard to keep track of who dies and why =P

[alexti]

AFAIK target selection is random, which means that it is random in MP, but in SP the randomness is corrected by Murphy laws, so the most valuable units are usuall targeted

[Cainehill]

ANd it would be _so_ nice if Illwinter would make it give a breakdown on casualties - if nothing else, at _least_ which commanders were killed / wounded.

[Vicious Love]

Empirically baseless hypothesis here, what if WoG randomizes damage in two stages per "ouch check", first choosing a province at random(Probability modified by dominion, as stated in spell description), then choosing targets within said province at random(a la Fires From Afar)?
This might explain why researchers/artificiers/conjurers/etc, who spend most of their time alone in a fortified laboratory, often seem to be the spell's favorite targets.
Then again, this might just be a question of selective observation, since one seldom overlooks the death of one's prize lapmage.

[Peter Ebbesen]

...I always thought that it affected the entire world, and performed separate checks in each province based on dominion; i.e. that several provinces (belonging to one or more players) could take casualties each turn and that each province had a chance based on dominion strength to be affected each turn. Hard to gather enough empirical data except through a lot of work, though.

[Vicious Love]

Peter Ebbesen said:
...I always thought that it affected the entire world, and performed separate checks in each province based on dominion; i.e. that several provinces (belonging to one or more players) could take casualties each turn and that each province had a chance based on dominion strength to be affected each turn. Hard to gather enough empirical data except through a lot of work, though.

If true, that'd definitely account for the apparent bias. It'd be very hard to test precisely, but it shouldn't take too many turns of testing under controlled sandbox conditions to get a general, not-too-fallible impression of whether or not mages are more vulnerable to WoG when alone in a province.

[NTJedi]

Cainehill said:

ANd it would be _so_ nice if Illwinter would make it give a breakdown on casualties - if nothing else, at _least_ which commanders were killed / wounded.

Definitely this would be great information... hopefully Dominions_3 will have this.

[Taqwus]

It might be quite a bit of information.

I wonder if, in addition to a summary in the message, it would work better if there were a visible, clickable sign on the affected provinces (where commanders were killed through non-normal-battle events) linking with text indicating who was killed and what they were doing at the time. A province-based note system might be extremely useful.

[Vicious Love]

Seconded. All vociferous-like.