So I started a thread earlier, whose initial question was quickly resolved, but which brought up another issue that seems thorny.
How, precisely, should we evaluate "<do some effect> to the 2* targets with <some condition>."
* Could be more than 2, as long as it is some specific number more than 1 and less than "all".
This seems obvious, but isn't, because of some well-established meta-rules. There are two very different ways of interpreting this:
A) This is how the App currently works, and is the simple english interpretation.
- Determine the 2 targets who fit the condition. If there are ties, players choose. Call these targets X and Y.
- Players choose whether X or Y gets affected first. Let's say it's X.
- X gets affected, and any appropriate side effects happen.
- Have 2 targets been effected? No. So we need to process another.
- Y gets affected, and any appropriate side effects happen.
- Have two targets been damaged? Yes. We're done.
B) Based on rulings involving End of Days and Forced Deployment, and the like.
- Determine the 2 targets who fit the condition. If there are ties, players choose. Call these targets X and Y.
- Players choose whether X or Y gets affected first. Let's say it's X.
- X gets affected, and any appropriate side effects happen.
- Have 2 targets been effected? No. So we need to process another.
- Determine the 2 targets who fit the condition. If there are ties, players choose. One (or sometimes both!) of these targets will be different from the original pair, so let's call them Y and Z.
- Players choose which target to damage. Let's say they choose Z. (If X had still been one of the 2, he would have been ineligible as a target due to already having been picked, so the "choice" would be forced.)
- Z gets affected, and any appropriate side effects happen.
- Have two targets been damaged? Yes. We're done.
(So, interpetation B would routinely allow players to react to situations such as "Damage the 2 targets with the highest HP", by damaging the (originally) 1st and 3rd highest!)
The argument for B is that each instance of damage is completely unique, and any variables involved are recalculated each time. The question is, does that recalculation include target selection in the general case of "<do effect E> to the 2* targets with <some condition>."
Interpreatation B *is* a bit counter-intuitive, but it seems to me to be most correct, in the absence of an official ruling. I hereby request a ruling along the lines of:
"When an card effects multiple (but not all) targets, those targets {ARE/ARE NOT} all determined before the actual effects start. The targets {ARE NOT/ARE} recalculated after each intance of the effect."