Roger,
A meteorologist warns of an dangerous front that might produce a local tornado within three days. Sure enough a damaging tornado occurs the next day, but nobody died because they had been warned and they responded appropriately.

What if that tornado had been followed by 5 others within the three day window?
Would you say that only the first one counted? At the end of the year would you discount the later five as they were only "aftertwisters"?

Meterological statistics often are concerned with the annual number of tornadoes. Your system would mess up the statistics as you only consider one per warning. Is that science?

job