Science would improve dramatically if peer review were replaced with reproduction. Every scientific grant would need to budget for reproduction by an independent party, and there were journals for novel results and journals for reproduction. Grad students should do reproduction of established results as part of their studies as well.