You will miss the second runner, if their timing is less than 100 ms apart,
which happens quite often.
I see further problems identifying the tags that correspond to measurements
if more than one tag is in reach.
Good luck with using delay while trying to get precise measurements of events. 