Saying military strikes on June 21st triggered a 'no' outcome for prediction markets on June 19-20 feels like perhaps overstating the connection, doesn't it?
It may be worth considering that the resolution for those earlier dates would have simply been based on whether military action had actually occurred within their specified timeframe, as a standalone event.
It's a bit like suggesting that a bus arriving late today somehow made yesterday's earlier buses arrive on time; the past outcome is already set.
Perhaps the later event simply confirmed what was already decided for those days, rather than actively causing it, gently guiding us to look at causality with a little more precision.