Hence the “replication crisis”. Replication of experiment and differences in outcome highlight the “core” result. When the results are just totally different, obviously there was “gaming” or just a bad experiment.