I was about to sit down and write a perfectly conventional post addressing Jack Flaherty and interesting PITCHf/x data of his when the news broke: the bombshell announcement that Mike Matheny, John Mabry, and Bill Mueller were fired on Saturday night. With all due respect to Mr. Flaherty–who I might add is having a fine season of his own as [spoiler alert] you will probably read about in more detail next week–I chose to put that article on hold and explore a more timely topic.
A.E. Schafer penned his thoughts on the dismissals yesterday through the lens of what is in store for the Cardinals coaching staff going forward, but I want to focus on what effects firing a manager can have on the players. There is a perception that canning the head coach or manager can fire a team up and ignite a run of success in the immediate aftermath. In fact, that was one of the first questions I was asked when I joined host Matt Harab on SB Nation Radio yesterday.
Do you think [firing Matheny] was more done to fire up the team in an inspiration kind of deal to get them going for the second half of the season?
Make no mistake, I firmly believe that moving on from Matheny was not only the correct decision for the Cardinals, but one that was grossly overdue as well. Whether the Cardinals leapfrog teams ahead of them in the standings or not, the now ex-skipper’s well-documented shortcomings both as a tactician and leader in the clubhouse were undeniable and unfit for the manager’s helm. That said, I was still curious to see how past ballclubs have performed following a similar managerial change to determine if there is any truth to the myth about teams turning things around once their manager departs.
You thought you were going to make it through an entire post of mine without numbers? Think again.
I began by compiling game-by-game final score data for every team who has employed multiple managers in a season since 1998. Managers who did not complete the season due to medical reasons in addition to those who–either they or their replacement–managed fewer than 30 games that year were excluded. But before we launch these data points into the eternal abyss of deleted-from-the-spreadsheet-dom, let’s take a look at who has been handing out the most pink slips.
I won’t equate correlation with causation, but the more consistent, perennial contenders of the past two decades have not made nearly as many midseason switches at manager. To give the Matheny firing more historical context, St. Louis just became the last of the 30 teams to either fire a manager or refuse to renew their contract since Bill DeWitt purchased the franchise. Let’s not lose sight of the fact that we Cardinals fans have enjoyed more success over the past 15-20 years than arguably any other team in the National League.
But I digress. Returning to our original question, we now have an even 40 case studies to work with after filtering out the small sample size outliers. I first simply compared each club’s record and run differential before and after pulling the trigger on a managerial shakeup. For the seven clubs that rolled out three managers in one year, I simply merged the second and third manager stats together since the second was merely acting as interim during the search for the longer-term skipper.
Overall Team Performance Before vs. After Managerial Change
Stat | Before Manager Change | After Manager Change |
---|---|---|
Stat | Before Manager Change | After Manager Change |
Win% | 0.424 | 0.473 |
Pythagorean Win% | 0.440 | 0.463 |
Wins | 1,315 | 1,593 |
Losses | 1,789 | 1,778 |
Runs Scored | 13,251 | 14,805 |
Runs Allowed | 15,127 | 16,068 |
These teams improved their winning percentage by 49 points, or approximately 7.9 additional wins per 162 games. However, the row directly below that contains the before and after pythagorean winning percentages, which are what we would “expect” the win-loss records to be based on how many runs the clubs scored and allowed. The increase here is only 23 percentage points, working out to just under 3.7 more wins per 162 games. In other words, we can deduce that the axed managers called the shots for teams that, presumably unluckily, underperformed their expected records. Conversely, these same clubs appear to be the recipients of more generous good fortune after the prior manager made their exit.
It would require far more rigorous testing, but it is possible that the discrepancy between these expected and observed results shouldn’t be entirely chalked up to good or bad luck. Maybe the dispatched skipper was costing his team close games through suboptimal lineup construction, erroneous bullpen management, and a variety of other strategical blunders that got them fired in the first place whereas the replacement manager was a more adept in-game decision-maker.
What also makes these enhanced winning percentages more interesting is that these clubs should, in theory, deteriorate and become less talented as the season progresses. The average team in our data fired their manager after roughly 78 games, which generally falls in the last week of June. This means our average team fired their manager with around a 33-45 record, a figure that presumably points towards selling at the trade deadline and further depleting the big league roster the next manager is forced to work with.
You could argue that “Firing the manager fires up the clubhouse and galvanizes success” is at least a plausible conclusion according to these numbers. Given the error bars and typical standard deviations for both actual and pythagorean win totals, however, I would be inclined to suggest that these findings are clouded by statistical noise.
If there is a wave of momentum that these teams ride, we would expect it to be the most prominent immediately following the managerial change when the transition is still fresh in the players’ minds. This time, I juxtaposed team performance before swapping out managers with just the 10 and 20 games after the change.
Technical note before the incredibly astute reader notices this: the 2005 Reds tied against the Astros in their eighth game after firing manager Dave Miley, hence the win and loss totals adding to sums of 399 and 799, respectively.
Immediate Team Performance Before vs. After Managerial Change
Stat | Before Manager Change | 10 Games After Change | 20 Games After Change |
---|---|---|---|
Stat | Before Manager Change | 10 Games After Change | 20 Games After Change |
Win% | 0.424 | 0.444 | 0.451 |
Pythagorean Win% | 0.440 | 0.457 | 0.451 |
Wins | 1315 | 177 | 360 |
Losses | 1789 | 222 | 439 |
Runs Scored | 13251 | 1692 | 3375 |
Runs Allowed | 15127 | 1861 | 3759 |
The improvements in numbers here in the 10 and 20 game spans are even less pronounced than over the entire season. The changes in wins in the table above are only 3.2 and 4.4 wins per 162 games. Meanwhile, the higher and equal pythagorean records would seemingly contradict the notion that the new managers were savvy enough to net their clubs more wins than “deserved” as told by their run differential.
I’m sure there is a phycological aspect to replacing a manager and that higher workplace satisfaction can lead to greater productivity. There likely are tangible benefits to firing an ineffective skipper, but, if anything, this study would suggest that the narratives about “firing up the ballclub” are overstated.
For whatever it’s worth, the Cardinals currently have a 1.000 winning percentage in the Mike Shildt era. I personally wouldn’t mind if they maintained that.