May 3, 2004
Rockies Try the Four-Man
The Right Team for the Test?this article in the Denver Post, the Rockies are switching to a four-man rotation.
Let me repeat that: the COLORADO ROCKIES are going to a FOUR-MAN ROTATION. In one stroke, Dan O'Dowd has mixed together two of the most compelling issues in baseball analysis today--how to win at altitude, and how to optimize the usage of your pitchers.
First of all, the Rockies should be commended for being so innovative. After an off-season that mostly involved inviting old Rockies of yore back to town, and made it look like the Rockies were giving up on solving the whole how-to-win-at-Coors-Field thing, it's reassuring to see them trying to adapt to their home environment by once again thinking outside the box. As this series of articles (Part I, Part II, Part III) suggests, the four-man rotation can be a winning strategy that should be considered, if not adopted, by all major league teams.
That said, if I were to recommend one team to make the switch, the Rockies wouldn't be my first pick. They might be my last.
It's not that the four-man rotation can't work in Colorado. On the contrary, because more pitches are required to get through each inning in Colorado, the Rockies have always dealt with limitations on the number of innings they can get out of their best starters, limitations that other teams don't have. If done right, the four-man rotation will help the Rockies get those innings back by extracting an extra six or seven starts from each of their front four.
The key phrase there is, "if done right". The pressures to conform, and the very real risk that an unsuccessful fling with the four-man rotation can cost a general manager his job, make the decision to ax the fifth starter a dicey one even in the best of circumstances. Throwing in that little wrinkle known as Coors Field adds another layer of risk to the equation.
To their credit, the Rockies are saying all the right things. They aren't simply eliminating the position of fifth starter; they are keeping that roster spot for a pitcher, giving Clint Hurdle no less than eight relievers he can call on.
Because of the number of pitches that must be thrown at Coors Field, along with the reality that no lead is truly safe in that ballpark, the Rockies are probably the only team in baseball that can legitimately claim they need to carry 12 pitchers. The physical drain on the Rockies' pitching staff is a very real problem with no great solution, although Joe Sheehan's suggestion that MLB approve 26-man rosters for all games played in Denver is a good one.
The Rockies already carry more pitchers than other teams; now that they have to carry one fewer starter, they go into every ballgame with eight relievers, while most of their opponents carry only six. This has the obvious benefit of making it easier for the Rockies to strictly limit their starters' pitch counts--a vital ingredient to the success of the four-man rotation.
If the Rockies want to take full advantage of the move, though, they should use their advantage in bullpen depth to carry a couple of "specialist" relievers--say, sidearm pitchers with extreme platoon splits--whose value is highly dependent on facing only batters on one side of the plate. Few teams can afford the luxury of a reliever who can be spotted for just one or two batters at a time, which is why even left-handed specialists typically face more right-handed hitters than left-handed hitters over the course of a season. The Rockies are in a position to have, not only a LOOGY, but a ROOGY as well.
In a sense, the Rockies already do this with Steve Reed, who routinely faces 65% right-handed batters over the course of a season, and has always enjoyed great success in Colorado. (On the other side of the rubber, Mike Myers' best seasons were spent in Colorado.) Going to an eight-man bullpen allows the Rockies to extend this tactic even further.
But this gamble will succeed or fail based on one simple factor: whether the Rockies can keep their starting pitchers healthy working on three days' rest. Which is why their #1 priority has to be the strict monitoring of their starters' pitch counts. The Rockies have announced a limit of 85-90 pitches for all their starters, and if they keep to that promise--and also try to limit high-stress, high-pitch count innings--they will be doing everything in their power to pull this off.
It would be very tempting for the Rockies to set a more moderate limit on their starters, like 105 or 110 pitches, and in any other ballpark that might work. But one thing the Rockies have figured out--a finding backed up by medical science--in their decade in the mountains: as a result of the thin air, the body recovers from physical exertion slower than it does at sea level.
Mike Hampton threw eight shutout innings in his first start at Coors Field after signing with the Rockies, throwing only 98 pitches. The next day, he said, "I felt like I had been hit by a truck when I got up." The difficulty in recovering from each start was so debilitating that, before he was traded to the Braves, Hampton was planning to outfit his bedroom in Colorado with a pressure chamber so that his muscles might heal faster between starts.
More than any other characteristic of Coors Field, it is this factor--the difficulty pitchers have in recovering between starts--that increases the risk that the Rockies' four-man experiment will fail.
The Rockies are aware of this. O'Dowd discussed this factor at length during his speech at the SABR convention in Denver last summer. Last year, the Rockies had the lowest team Pitcher Abuse Points score in all of baseball, averaging just one point of Stress for the season. This year, even before implementing the four-man rotation, no Rockies starter exceeded 108 pitches, and only twice has a starter faced a new batter after his pitch count hit 100.
It may fail anyway, but if the Rockies do what they say they'll do, they will give the four-man rotation every opportunity to succeed.
And if it does? Not only will the Rockies get the benefit of skipping over a fifth starter, a huge relief for a team struggling to find even four decent starters. There may be some ancillary benefits as well. Two of the Rockies' starters, Jason Jennings and Shawn Estes, are groundball pitchers, and Joe Kennedy's stunning start with the Rockies is associated with a huge increase in his G/F ratio (he's currently at 1.83 for the season; his career ratio through 2003 was less than one.) If the conventional wisdom that groundball pitchers are more effective on short rest--which is at least partially supported by the evidence--is true, the Rockies could get the added benefit of increased effectiveness from three-fourths of their rotation.
The best part of the Rockies' decision, though, is that they really have nothing to lose. Through Sunday, the Rockies' starters have combined for a 7.98 ERA. Scott Elarton has a 9.88 ERA, Jennings is at 10.52, and they're in the four-man rotation. (If you have to ask why, it's because the fifth starters the Rockies have used, Denny Stark and Jeff Fassero, had ERAs of 16.20 and 21.00 in the rotation.)
So all things considered, the Rockies may just pull this off. They've got the strict pitch limits in place, and the 12-man pitching staff to make it feasible. They've got a rotation that can't pitch any worse than it has to this point. And most of all, they have the perfect excuse to try something radical: the desperation that comes from 10 hapless, fruitless years of trying to find a way for their pitchers to beat Coors Field.
It's not the way I would have done it. But the revolution has to start somewhere. Even if "somewhere" is at 5,280 feet.