Managerial risk aversion?

Why are American sports teams so seemingly risk averse in their selection of leaders? This subject has bothered me for some time and it bounced back into my mind last night while watching the Cubs/Cardinals game and enduring the endless dugout shots of the ever pensive Cardinals manager, Tony LaRussa.

Both LaRussa and his counterpart in the other dugout, Lou Piniella, have been fired in previous managerial jobs (a few times in Lou's case). Taking a cursory look around MLB, plenty of current managers come from the same stock of former firees: Leyland, Torre, Cox, Tracy, Garner etc. A survey of the NBA or the NFL produces similar lists of recycled leaders.

Two points that contribute to an explanation of this phenomenon:

1. Being fired as the manager of a professional sports team doesn't send the same signal of incompetence that you might suspect.
2. The risk aversion and short term outlook in the front offices of professional sports franchises makes selecting a "name" leader much more likely, regardless of that leaders previous record.

Point #1 could be true if a manager or coach only impacts the final outcome of a small number of games during a season. If so, then firing such a leader might be seen as a shot across the bow to the folks who do impact the success or failure of a franchise: the players. Additionally, following to point #2, the selection criteria for a new leader may focus on intangibles such as name recognition to send the proper signal of commitment to winning, rather than skills that directly impact on field play. A first time coach may send a low cost approach type signal and could hurt sales of tickets to see the team play.

Assuming that this line of thinking bears some resemblance to reality, are GM's too risk averse in their selections? Could the Triple A manager get the same results as the big name who has bounced around for 10 years, and at half the salary? Have GM's artificially limited their supply of talent? The flip side is the argument that there are a minority of superstar leaders with the ability to inspire the key talent, put fans in seats, and have a larger positive impact on outcomes than the next best leader.

Perhaps of more interest: do these same biases extend to corporate America? How much impact on performance does the CEO actually have and does the same risk aversion lead to a race to the top in terms of executive pay packages (leaving aside the fact that some CEO's basically set their own pay)? No company wants to be known as paying only at the 50th percentile for their leader. Promoting the unknown controller from within may send the wrong signal. For the rest of the team, 50th percentile may be just fine. Following this logic, the hypothesis would be that American companies are needlessly driving up CEO pay and perhaps exhibiting a similar recycling of "name" talent that is seen in professional sports despite the reality that leaders have limited impacts on performance, especially compared to the next best choice.