Readers of Fire Brand of the American League have been aware of my fandom for manager Terry Francona. Simply put, I feel he’s an excellent manager who could be considered one of the greatest managers in the game today.
That perception of mine has taken a hit since I viewed three metrics used to determine how well a manager strategically performs in games. As is wont with most numbers, this discounts the impact a manager has in a clubhouse amongst other variables, but viewed through the lens of strategy, the numbers that result are quite eye-opening.
Out of 20 managers that managed at least one full season prior to 2007 — 20 managers total — Francona ranks 16th in terms of strategy, ahead of only Joe Torre, Eric Wedge, Jim Leyland and Clint Hurdle. You can view the chart of the managers here, courtesy of the Wall Street Journal.
The three categories are “Close Games, Wins Above Expectation Player Performance.” And those are…
Assessing managers is difficult, since their win-loss records are closely tied to the team’s talent level. Because of that, we used three metrics designed to reveal a manager’s acumen — performance in close games, “Pythagorean” win expectations and whether they got the most out of their players — then averaged each manager’s ranking in those categories for a final score.
For the close-games category, we subtracted each manager’s overall winning percentage from his winning percentage in games tied after the sixth inning, thus determining whether he performs better or worse in close games. The win-expectations chart compares the number of games each manager’s team won versus how many it should’ve won based on how many runs it scored and allowed (also known as the team’s Pythagorean expectation). The third category states how many additional games a manager won or lost for his team per season. David Gassko, a contributor to the Hardball Times Web site, arrived at those figures by comparing how players’ statistics improved or declined under different managers (with adjustments for age and other variables).
The data for the first two categories is for the past five regular seasons and is from Baseball Info Solutions, a firm that provides statistical information. The data for the third category is from all of major-league history through the 2006 season. The 20 managers listed are current major-league skippers who managed a full season before 2007.
Francona ranks ninth in Close Games with a 0.031 record, meaning that for games tied after the sixth inning, he usually comes out on top… this is good, right? I mean, the talent level DOES have something to do with this, as WSJ alludes, but still… take for example Joe Torre who shows an astounding -0.053 for 18th place.
Wins Above Expectation depend on the Pythagorean record, also another viable way of determining a manager’s effectiveness. In this, Joe Torre ranks a fantastic No. 1 with 16 wins above expectation. Francona ranks eighth with two wins over expectation.
Let’s stop here for a second. What would be more indicative of a manager’s effectiveness out of these two categories — Close Games or Wins Above Expectation?
I’m going with close games. If the pitcher of the day stinks it up and we lose 10-2, there’s not much the manager could do, and if he’s the manager of a high-octane offense, the Pythagoren expectation will inevitably show that the team should have won. Managers affect games in the clubhouse, in the lineup card, and when the game matters — and a tie game in the sixth certainly matters.
Last one, player performance. What’s the statistical difference in how a player performs for a manager? Joe Torre ranks dead last, meaning that “Torre” lost 1.82 games for his team. This was measured by “comparing how players’ statistics improved or declined under different managers (with adjustments for age and other variables).” I put quotation marks around Torre because… ultimately, the player is responsible for his play. Who’s to say the market didn’t affect the player, the shock at being traded… there’s way too many variables in here.
“Francona?” He’s just behind Torre, costing the team 19 games. “Bobby Cox” gives the Braves an amazing 2.29 games each year.
Now, let’s turn to the elephant in the room. Let’s look again at who brings up the rear behind Francona in overall ranking: Torre, Wedge, Leyland, Hurdle.
Let me get this straight. In 16th place is someone with two World Series rings in four years, four rings as a Yankee, a manager who brought his team to one out of the World Series last year, , a manager who has rings and turned the Tigers from decrepit to superpower and is the best manager in the game according to Nick Cafardo and basically everyone else and Hurdle, who brought the Rockies to the World Series in the midst of an amazing run.
Other notable managers include Mike Scoscia at 15, Lou Pinella at 12, Tony LaRussa and Bobby Cox both at 9. Those are some pretty heavy hitters who don’t do well in this metric.
So who does? Let’s look at the top five. Leading the pack is Ron Gardenhire, Bruce Bochy, Ned Yost, Charlie Manuel and Bob Melvin. Gardenhire is not considered a great manager, Bochy is by all accounts a fine manager. Yost is on the hot seat, Manuel was on the hot seat for two years before Jimmy Rollins intervened and had a MVP season and Melvin wasn’t even his team’s first choice as manager! (Remember Wally Backman?)
Fine effort by WSJ, but it seems like the order is reversed; Francona is the fifth best manager in the bigs.
Loading ...
Don’t look now, but Lugo’s leading the league in hitting! lol
I’m just not sure the statistics in question do a good job of objectifying performance.
Yeah, this looks like a load of crock to me. These statistics in no way separate player performance from manager performance.
Unless a player is under a manager for a couple of years, I’m not sure you can fairly evaluate the effect being under a different manager had. Like Evan said, changing location or even a change in age can have a large effect on a player’s performance.
But statistics exist exactly to objectively rate people. If you removed the names and looked at these people just as a collection of numbers would you come to the same conclusion? Is “everyone” saying Leyland is the best manager in the game any different than “everyone” saying Juan Pierre is a useful player or “everyone” handing Jeter gold glove after gold glove?
Also, I’ll say it again: Francona can be a good statistical manger once he stops using Lopez as a lefty specialist.
But stats are only as good as the data that backs them up. I would think a player’s performance is influenced by too many independent variables to fairly access a managers influence.
The article does not contain methodology for the player performance metric, but with enough of a sample size, and by comparing these players to those who stayed with the same manager, you could eliminate the noise of players aging. What you couldn’t cancel out is the other effects of changing franchise, such as medical staffs, weather and the like.
I’m not saying this chart is perfect by any means, but you can’t dismiss it by saying “it says my favorite managers are bad so its wrong.”
No, you can’t dismiss it simply because your favorite manager scores poorly… and yes, sometimes reputation belies actual performance… but come on now. This list has several of the best managers in the game at the very bottom, and some very, very questionable managers at the top.
That’s overwhelming circumstantial evidence that either (a) the methodology is fundamentally flawed, or (b) there is no purely statistical way to measure a manager’s performance. I suspect that a lot of it is the latter; in-game tactics are a vastly overstated aspect of the manager’s job.
I haven’t closely examined the WSJ study, but just from the results I infer that managers on good teams fare poorly, while managers on “surprise” teams fare well. Sorta like the balloting for Manager of the Year, most of the time.
Go Gagne!
Gagme lives!
JV, I didn’t even look at the rankings. I simply dismissed it because it’s not methodically sound. All of these measures have as much to do with the team itself as they do with the manager.
Maybe the WSJ article is complete rubbish. But that’s not my point. The point is that more than a third of this post is devoted to how the WSJ article is wrong because you and others feel it is wrong. That is Fire Joe Morgan worthy.
If you want to dig up the Hardball Times article that lays all this out and debate that you’re welcome to.
Evan, one of the problems with Pythagorean is that it somewhat exaggerates blowouts. It’s a good metric, but definitely not one that should be taken as gospel. Since it measures runs scored and runs allowed, it does not take into account that a 15-1 loss is still just a loss and a 1-0 win is still just a win. Plugging those numbers into a like formula would show that only 2 runs were scored whereas 15 were given up, which would make for two bad games; in actuality it was a 1-1 record.
BTW, in my sabermetrics community, this WSJ article is being debated. Go to insidethebook.com and click on the blog link. There is a discussion over there that can provide good insight into what a lot of this article does and doesn’t tell us about managers.
What’s wrong with feel? Not everything has to be about statistics.
Evan and Aubrey, not to be-labor the obvious, but it seems historically, in most sports, most of the time, good managers get good teams to perform well; the variables more often effecting team performance than not.
We are not claiming this proposed research protocol is flawed because it places our favorite managers in the bottom of the heap. Not at all. It is because it places the winningest managers and the miracle workers at the bottom, and consistently poor performers at the top. Statistics have to make sense, reflect and sometimes even explain reality, and this just doesn’t.
Evan, thanks for listing last year’s article “Is Tito hurting WS chances”? What comments. Like reliving the series. Lots of angst, and lots of second guessing.
On this subject, wasn’t everyone happy when Tito pulled Wake the other night, putting runners on base after a great performance? Did my heart good. And, IMO, wasn’t he correct in giving Corey and Snyder another shot . . . and didn’t they produce, again?
Sean:
At least the Brewers bailed him out. First he almost kills the Sox, then he goes and starts wreaking havoc on my adopted NL team. Thanks, Gags.
There’s nothing wrong with feelings, but if you disagree with a statistical argument, arguing from your gut isn’t really responding in any meaningful way to the article.
Side-note: KNBR here in SF reports that Bobby Kielty has accepted his assignment to Class-AAA Pawtucket.
I’m not sure if the extrapolation is accurate, but they believe that “this is an indication that the Red Sox have made progress on a potential deal involving Coco Crisp.”
Either way, I’m glad Kielty’s staying in the organization. He hit the GW homer in Game 4 of the WS last year, and I think there’ll be a few times this year when we’re happy he’s on our side.
High fives: MLB’s best, worst, most overrated
http://msn.foxsports.com/mlb/story/7969984/High-fives:-MLB’s-best,-worst,-most-overrated
Five best managers
1. Terry Francona, Red Sox – His steady hand amid the roiling Red Sox Nation keeps his team insulated from chaos.
2. Jim Leyland, Tigers – Perhaps the game’s finest tactician, with an excellent feel for his players.
3. Mike Scioscia, Angels – No team reflects its manager better: aggressive, hard-playing and maximizing its abilities.
4. Joe Torre, Dodgers – Will have to prove in a new forum that his serenity and leadership translate to a dysfunctional clubhouse.
5. Bobby Cox, Braves – His passion remains unchanged after all these years to give him the nudge ahead of Cleveland’s Eric Wedge, Arizona’s Bob Melvin, St. Louis’ Tony La Russa, the Cubs’ Lou Piniella and Minnesota’s Ron Gardenhire