Just looked up the Baseball Prospectus run expectancy chart, which calculates how many runs actually score in every out/baserunner scenario. With no one out, and runners at first and second, the scoring expectation is 1.43. With one out and runners at second and third, the scoring expectation is 1.29. So Molitor reduced the scoring expectation by 10%, without even factoring in the chance of the bunt failing (pop up/failure to get it down). It's not a huge difference, but the job of the manager is to find ways to increase the team's odds. This was clearcut example of Molitor (micro) managing in a way that actively decreased the team's expected scoring, and thus, chances of winning. "All right boys, we've got ourselves a rally going. Keep it going, let's get some more hits, and do whatever it takes to avoid making an out."