I dislike second-guessing the manager. I really do. Usually, I can understand his reasoning even if I don't necessarily agree with it. I can figure out what he's probably thinking. I'm also aware that he knows lots and lots of stuff that I don't know.
- Me, while ripping John Gibbons a brand new orifice for a decision I didn't like a whole lot.
I blithely mentioned the other day that it made sense to expect the New Boss to run a ball game more or less in the same fashion as the guy he's been sitting next to on the Red Sox bench these past few years. Reasonable, no? But I then began to wonder - exactly how does Terry Francona run a ball game? How does it differ from what Joe Maddon would do? (I realize Tampa fans are wondering why anyone would be interested, as they firmly believe Clueless Joe is the only reason they haven't won the WS yet, but I'm not so sure... )
I've always said that Game management accounts for a rather small part of the manager's actual job - the thing is, it accounts for almost all of the job that is visible to us. And we're stuck with that. And while way, way too much is always made about any one particular game decision, if it's possible to uncover some larger patterns... maybe it'll be worth knowing. We never have much context for how a manager operates. What does one team actually do differently than another team? Sometimes we can actually find out.
It's always useful to remind ourselves that there is no right or wrong way to do these things. For example - most managers very much prefer to use a set lineup. And while it's possible that I was simply traumatized by watching Jimy Williams work here for three plus seasons, I certainly understand that kind of thinking. I have some sympathy for this view. I'm not sure players play their best when they're being juggled and moved around. Nevertheless.... only two managers have seven World Series titles and one of them is Casey Stengel, who was only the maddest juggler of ballplayers who ever lived. And who's right behind Casey? Joe Torre, a man cut from the exact same cloth as Casey. Granted, Stengel and Torre both spent the first few decades of their managerial careers wandering from town to town, losing wherever they went. But it's obvious (it's certainly obvious now!) that their approach could work, that you could win doing things that way, and you could win pretty big...
So I don't want to imply that one way is better than another. I'm just rooting through the numbers, I'm just looking for something bright and shiny that I didn't already know.
Let's begin with something about pitching. Is Terry Francona - or Joe Girardi, or Ron Gardenhire, or whoever - quick to pull his starter? Or does he have a slower hook? Let's find out. We will use Bill James' updated definition of the terms Slow Hook and Quick Hook (which he probably invented anyway, back in the day.) Here's what you do.
Pull out your database with every start in the American League in 2010. Everybody's got one, right?
Assign 1 point for every pitch thrown by the starting pitcher.
Assign 10 points for every run allowed by the starting pitcher.
You now have a Score for every game in the league. The 25% of games with the lowest scores are the Quick Hooks, the 25% of games with the highest scores are the Slow Hooks. These are very broad categories, obviously - we've got roughly 565 Quick Hooks, as opposed to about 1130 Normal Hooks and about 565 Slow Hooks. So naturally, the average for each team would be about 41 Quick, 41 Slow and about 80 that were Neither.
For example, on April 30, Brandon Morrow threw 109 pitches and allowed 2 ER - that scores at 129. As it happens, 129 is about as close to being in the middle of these 2000 or so games as you can get. In 2010, any game that comes in at 114 or less will be a Quick Hook; any game that comes in at 143 or higher will be a Slow Hook.
The extremes - well, you can imagine what they're like. The Quickest Hook of all was Ozzie Guillen pulling Gavin Floyd after 7 pitches last September - Floyd's shoulder tightened up. And the Slowest Hook of all was Mike Scioscia leaving Scott Kazmir in to allow 13 runs while throwing 103 pitches over 5 innings last July.
It may be helpful to see a couple of games near the margins. While every manager is very well represented in all three categories, I'll provide examples from the 2010 Jays because I figure there's a fighting chance you might remember the actual game. The quickest Toronto hook of all was actually June 1, when Brian Tallet was pulled after 5.2 shutout innings, 82 pitches. Some others:
Sep 22 - Kyle Drabek, 5 IP, 3 ER, 84 pitches. It scores 114 and it's (barely) a Quick Hook. But so is...
Aug 1 - Jesse Litsch, 4 IP, 4 ER, 74 pitches. Also 114. As is...
Jun 11 - Ricky Romero, 5 IP, 4 ER, 74 pitches.
And some Slow Hooks from the 2010 Jays:
April 16 - Shaun Marcum, 7 IP, 4 ER, 105 pitches (scores 145)
June 29 - Brandon Morrow, 6 IP, 5 ER, 96 pitches (scores 146)
May 9 - Ricky Romero, 5.1 IP, 4 ER, 110 pitches (scores 150)
The slowest Toronto hook was August 11, when Shaun Marcum went 4 IP, threw 92 pitches, and allowed 8 ER. Score of 172.
Now these games, and games like them, may not seem like Slow Hooks or Quick Hooks to you, or to me for that matter. Here's what they are:
In the context of how AL managers were actually running their pitching staffs in 2010, games like these really do represent Quicker and Slower Hooks than what is currently generally accepted Standard Operating Procedure.
Now that was a mouthful of something, no?
A final question of method - what about the Complete Games? How can it be a Hook at all, when the hook never came? Well, obviously the hook could have come, and indeed if the hook had come for Gil Meche last May 8, after 128 pitches and 3 ER (score of 158) it would have been a Slow Hook. So why not count it? It does mean that on the other end, we really do have a couple of complete games that actually qualify as Quick Hooks - Braden's perfect game, Galarraga's imperfect game, Marcum's game against Oakland. Lumping them in with the Quick Hooks is obviously kind of silly but in a data set of more than 2200 games, I'll live with a dozen anomalies. What the hell... But the presence of these games, as well as games when the manager was obliged to quickly pull a starting pitcher whose arm wasn't feeling good that day, probably does give each manager an extra Quick Hook or two...
And now, at last.. .you know you want it.... ZE DATA TABLE !!!
The teams and managers really committed to the Quicker Hook were Oakland, Minnesota, and Seattle. And the most interesting, to me anyway, is Seattle. To find out why I think that, you'll have to wait until I get around to doing something similar about how everyone was running their bullpens. As for the other Quick Hooks - Cito Gaston in Toronto was managing a very raw pitching staff (only Ricky Romero had ever thrown enough innings in a season to even qualify for the ERA title) and Ron Washington was managing in Texas, where the baseballs fly out of the yard and the heat grinds you down. So I don't think there's much to see there.
Now the Slower Hooks.... let's just re-sort the numbers
Francona and Mike Scioscia are both veteran managers, and many veteran managers have tended to go longer with the starters. Jim Leyland is also among the managers more likely to have a Slower Hooks, and I suspect Leyland would rank much higher here if he wasn't watching out for Porcello (and maybe Scherzer as well.) But on the other hand... Ron Gardenhire's been doing this for a while and he's got one of the Fastest Hooks around...
Finally, let's sort them one more time - this time by most games that are neither Slow nor Quick. By now of course we've identified almost every manager as tending in one direction or the other. With a couple of prominent exceptions...
Ladies and gentlemen, let me commend to your attention the fascinating cases of Joe Maddon and Mike Scoscia.
Maddon has both fewer Quicker Hooks and fewer Slower Hooks - he actually defines the conventional wisdom! The conventional wisdom is what Joe Maddon does! (For now - wait til we get to the bullpen!). Ned Yost, with his Slower hook and Cito Gaston with his Quicker one are pretty conventional as well.
But at the other extreme - Mike Scioscia has far more Slower Hooks than the league average - but he also has more Quick Hooks than the League Average. He defies the conventional wisdom, whatever it is. Well, him and Trey Hillman. Which didn't work out so well in Kansas City, but I doubt this was why.
Did I mention that there is no right or wrong way to do these things? (Repeatedly, I'm sure.) There's only whatever works for you, whatever you can win with. The work of Francona and Gardenhire and Maddon and Scioscia sure looks like an argument over the best way to do these things, because they certainly don't seem to agree with one another. But it really is possible to win every which way. (Does anyone really want to say one of them's right and one of them's wrong? That Francona's hook is too slow, that Gardenhire's is too fast, that Maddon is too conventional, that Scioscia is too unconventional?) Let's remember that most of the managers working in the AL last season have been able to win doing things whichever way they happen to do them. Six of them (Francona, Gaston, Girardi, Leyland, Guillen, Scioscia) have managed WS winners - four more managed teams into the post-season (Showalter, Maddon, Gardenhire, Washington)
And Ned Yost came really, really close.
- Me, while ripping John Gibbons a brand new orifice for a decision I didn't like a whole lot.
I blithely mentioned the other day that it made sense to expect the New Boss to run a ball game more or less in the same fashion as the guy he's been sitting next to on the Red Sox bench these past few years. Reasonable, no? But I then began to wonder - exactly how does Terry Francona run a ball game? How does it differ from what Joe Maddon would do? (I realize Tampa fans are wondering why anyone would be interested, as they firmly believe Clueless Joe is the only reason they haven't won the WS yet, but I'm not so sure... )
I've always said that Game management accounts for a rather small part of the manager's actual job - the thing is, it accounts for almost all of the job that is visible to us. And we're stuck with that. And while way, way too much is always made about any one particular game decision, if it's possible to uncover some larger patterns... maybe it'll be worth knowing. We never have much context for how a manager operates. What does one team actually do differently than another team? Sometimes we can actually find out.
It's always useful to remind ourselves that there is no right or wrong way to do these things. For example - most managers very much prefer to use a set lineup. And while it's possible that I was simply traumatized by watching Jimy Williams work here for three plus seasons, I certainly understand that kind of thinking. I have some sympathy for this view. I'm not sure players play their best when they're being juggled and moved around. Nevertheless.... only two managers have seven World Series titles and one of them is Casey Stengel, who was only the maddest juggler of ballplayers who ever lived. And who's right behind Casey? Joe Torre, a man cut from the exact same cloth as Casey. Granted, Stengel and Torre both spent the first few decades of their managerial careers wandering from town to town, losing wherever they went. But it's obvious (it's certainly obvious now!) that their approach could work, that you could win doing things that way, and you could win pretty big...
So I don't want to imply that one way is better than another. I'm just rooting through the numbers, I'm just looking for something bright and shiny that I didn't already know.
Let's begin with something about pitching. Is Terry Francona - or Joe Girardi, or Ron Gardenhire, or whoever - quick to pull his starter? Or does he have a slower hook? Let's find out. We will use Bill James' updated definition of the terms Slow Hook and Quick Hook (which he probably invented anyway, back in the day.) Here's what you do.
Pull out your database with every start in the American League in 2010. Everybody's got one, right?
Assign 1 point for every pitch thrown by the starting pitcher.
Assign 10 points for every run allowed by the starting pitcher.
You now have a Score for every game in the league. The 25% of games with the lowest scores are the Quick Hooks, the 25% of games with the highest scores are the Slow Hooks. These are very broad categories, obviously - we've got roughly 565 Quick Hooks, as opposed to about 1130 Normal Hooks and about 565 Slow Hooks. So naturally, the average for each team would be about 41 Quick, 41 Slow and about 80 that were Neither.
For example, on April 30, Brandon Morrow threw 109 pitches and allowed 2 ER - that scores at 129. As it happens, 129 is about as close to being in the middle of these 2000 or so games as you can get. In 2010, any game that comes in at 114 or less will be a Quick Hook; any game that comes in at 143 or higher will be a Slow Hook.
The extremes - well, you can imagine what they're like. The Quickest Hook of all was Ozzie Guillen pulling Gavin Floyd after 7 pitches last September - Floyd's shoulder tightened up. And the Slowest Hook of all was Mike Scioscia leaving Scott Kazmir in to allow 13 runs while throwing 103 pitches over 5 innings last July.
It may be helpful to see a couple of games near the margins. While every manager is very well represented in all three categories, I'll provide examples from the 2010 Jays because I figure there's a fighting chance you might remember the actual game. The quickest Toronto hook of all was actually June 1, when Brian Tallet was pulled after 5.2 shutout innings, 82 pitches. Some others:
Sep 22 - Kyle Drabek, 5 IP, 3 ER, 84 pitches. It scores 114 and it's (barely) a Quick Hook. But so is...
Aug 1 - Jesse Litsch, 4 IP, 4 ER, 74 pitches. Also 114. As is...
Jun 11 - Ricky Romero, 5 IP, 4 ER, 74 pitches.
And some Slow Hooks from the 2010 Jays:
April 16 - Shaun Marcum, 7 IP, 4 ER, 105 pitches (scores 145)
June 29 - Brandon Morrow, 6 IP, 5 ER, 96 pitches (scores 146)
May 9 - Ricky Romero, 5.1 IP, 4 ER, 110 pitches (scores 150)
The slowest Toronto hook was August 11, when Shaun Marcum went 4 IP, threw 92 pitches, and allowed 8 ER. Score of 172.
Now these games, and games like them, may not seem like Slow Hooks or Quick Hooks to you, or to me for that matter. Here's what they are:
In the context of how AL managers were actually running their pitching staffs in 2010, games like these really do represent Quicker and Slower Hooks than what is currently generally accepted Standard Operating Procedure.
Now that was a mouthful of something, no?
A final question of method - what about the Complete Games? How can it be a Hook at all, when the hook never came? Well, obviously the hook could have come, and indeed if the hook had come for Gil Meche last May 8, after 128 pitches and 3 ER (score of 158) it would have been a Slow Hook. So why not count it? It does mean that on the other end, we really do have a couple of complete games that actually qualify as Quick Hooks - Braden's perfect game, Galarraga's imperfect game, Marcum's game against Oakland. Lumping them in with the Quick Hooks is obviously kind of silly but in a data set of more than 2200 games, I'll live with a dozen anomalies. What the hell... But the presence of these games, as well as games when the manager was obliged to quickly pull a starting pitcher whose arm wasn't feeling good that day, probably does give each manager an extra Quick Hook or two...
And now, at last.. .you know you want it.... ZE DATA TABLE !!!
QuickThe league average, by definition, is 25% - so Mike Scioscia (just barely) and everyone above him, has more Quick Hooks than the average.
MGR Games Hooks Pct.
Showalter BAL-3 57 21 .368
Geren OAK 162 54 .333
Gardenhire MIN 162 51 .315
Brown SEA-2 50 15 .300
Wakamatsu SEA-1 112 32 .286
Gaston TOR 162 45 .278
Washington TEX 162 43 .265
Girardi NYY 162 42 .259
Hillman KCR-1 35 9 .257
Scioscia LAA 162 41 .253
Acta CLE 162 39 .241
Maddon TBR 162 39 .241
Guillen CWS 162 38 .235
Samuel BAL-2 51 10 .196
Francona BOS 162 30 .185
Leyland DET 162 29 .179
Yost KCR-2 127 20 .157
Trembley BAL-1 54 5 .093
The teams and managers really committed to the Quicker Hook were Oakland, Minnesota, and Seattle. And the most interesting, to me anyway, is Seattle. To find out why I think that, you'll have to wait until I get around to doing something similar about how everyone was running their bullpens. As for the other Quick Hooks - Cito Gaston in Toronto was managing a very raw pitching staff (only Ricky Romero had ever thrown enough innings in a season to even qualify for the ERA title) and Ron Washington was managing in Texas, where the baseballs fly out of the yard and the heat grinds you down. So I don't think there's much to see there.
Now the Slower Hooks.... let's just re-sort the numbers
SlowEveryone from Ned Yost on up has more Slow Hooks than the 25% average and Terry Francona, by a comfortable mile, leads the pack. With John Farrell sitting by his side, of course. Not that this really tells us how Farrell will handle the group of starters he has in Toronto. In Boston Francona was working with a veteran staff for the most part - Lackey, Becket, Matsuzaka, even Lester I suppose. Buchholz being the obvious exception.
MGR Games Hooks Pct.
Francona BOS 162 61 .377
Hillman KCR-1 35 13 .371
Scioscia LAA 162 52 .321
Samuel BAL-2 51 16 .314
Guillen CWS 162 49 .302
Leyland DET 162 49 .302
Acta CLE 162 48 .296
Yost KCR-2 127 36 .283
Washington TEX 162 40 .247
Trembley BAL-1 54 13 .241
Girardi NYY 162 35 .216
Gaston TOR 162 31 .191
Maddon TBR 162 31 .191
Wakamatsu SEA-1 112 21 .188
Geren OAK 162 29 .179
Gardenhire MIN 162 26 .160
Brown SEA-2 50 8 .160
Showalter BAL-3 57 6 .105
Francona and Mike Scioscia are both veteran managers, and many veteran managers have tended to go longer with the starters. Jim Leyland is also among the managers more likely to have a Slower Hooks, and I suspect Leyland would rank much higher here if he wasn't watching out for Porcello (and maybe Scherzer as well.) But on the other hand... Ron Gardenhire's been doing this for a while and he's got one of the Fastest Hooks around...
Finally, let's sort them one more time - this time by most games that are neither Slow nor Quick. By now of course we've identified almost every manager as tending in one direction or the other. With a couple of prominent exceptions...
MGR Games Neither Q/S Pct.
Trembley BAL-1 54 36 0.667
Maddon TBR 162 92 0.568
Yost KCR-2 127 71 0.559 Slow
Brown SEA-2 50 27 0.540 Quick
Gaston TOR 162 86 0.531 Quick
Wakamatsu SEA-1 112 59 0.527 Quick
Showalter BAL-3 57 30 0.526 Quick
Gardenhire MIN 162 85 0.525 Quick
Girardi NYY 162 85 0.525 Quick
Leyland DET 162 84 0.519 Slow
Samuel BAL-2 51 25 0.490 Slow
Geren OAK 162 79 0.488 Quick
Washington TEX 162 79 0.488 Quick
Acta CLE 162 75 0.463 Slow
Guillen CWS 162 75 0.463 Slow
Francona BOS 162 71 0.438 Slow
Scioscia LAA 162 69 0.426 Slow/Quick
Hillman KCR-1 35 13 0.371 Slow/Quick
Ladies and gentlemen, let me commend to your attention the fascinating cases of Joe Maddon and Mike Scoscia.
Maddon has both fewer Quicker Hooks and fewer Slower Hooks - he actually defines the conventional wisdom! The conventional wisdom is what Joe Maddon does! (For now - wait til we get to the bullpen!). Ned Yost, with his Slower hook and Cito Gaston with his Quicker one are pretty conventional as well.
But at the other extreme - Mike Scioscia has far more Slower Hooks than the league average - but he also has more Quick Hooks than the League Average. He defies the conventional wisdom, whatever it is. Well, him and Trey Hillman. Which didn't work out so well in Kansas City, but I doubt this was why.
Did I mention that there is no right or wrong way to do these things? (Repeatedly, I'm sure.) There's only whatever works for you, whatever you can win with. The work of Francona and Gardenhire and Maddon and Scioscia sure looks like an argument over the best way to do these things, because they certainly don't seem to agree with one another. But it really is possible to win every which way. (Does anyone really want to say one of them's right and one of them's wrong? That Francona's hook is too slow, that Gardenhire's is too fast, that Maddon is too conventional, that Scioscia is too unconventional?) Let's remember that most of the managers working in the AL last season have been able to win doing things whichever way they happen to do them. Six of them (Francona, Gaston, Girardi, Leyland, Guillen, Scioscia) have managed WS winners - four more managed teams into the post-season (Showalter, Maddon, Gardenhire, Washington)
And Ned Yost came really, really close.
