The problem with Defensive is the stability loss means we have no good options. (Well, that and it's massively far behind in vote count, but ignoring that)
Play around with it in
the program. We can do:
['KingYes', 'OrderNo', 'Defensive']
for 90% chance of -2 stability, or
['KingYes', 'OrderYes', 'Defensive']
which has that 7% chance of death that people don't like.
If we do KingNo, then we end up dead with extremely high probability.
Both options leave us with 1 Legitimacy and a war-oriented leader without a war to fight. If we had a bit more stability it'd be ok, but as we are, going to war and boosting it as much as possible is almost certainly better than needing half those boosts anyway and still ending up in a bad situation.
Huh? I'm saying literally the opposite once again. That we should care about game mechanics, and people are doing so. That stats and game mechanics are important, and the vast majority of voters care about them.
Perhaps you need to go back and check how the argument started? I was arguing against the "You're the only one who really cares about stats" assertion.