Mismatched experimenter results

Hi,

I am having a problem with the experimenter. In the model attached, I have set up 3 scenarios. The first and third are exact copies of each other and as you can see, most of the replications give the exact same results as each other. But for some reason, the first 4 replications of scenario 1 give very different results. This is not an isolated incident and was happening when I was running all 16 scenarios with 20 replications each; some of the replications of a particular scenario gave very different results to the other replications (a factor of 10 out).

Does anyone know why this might be happening?

Move your code from "Start of Replication" to the model trigger "On Reset"

Thank you very much! :-)