After losing to Detroit on June 17 and dropping to 25-40, the Rockies announced what would become the defining topic for the rest of the season – they would switch to a four-man rotation with a 75-pitch limit on the starters. Furthermore, they introduced the concept of paired-pitching into the scheme – where a quasi-starter, or piggyback starter, would come in once the starter hit his pitch limit. He, too, would be on a pitch count, limited to 40-45 pitches so as to be available every three days. This whole idea was concocted by the Rockies’ GM, Dan O’Dowd, on the premise that it would combat the hypothetical effects of altitude adding extra wear on the pitchers and teach pitchers to throw more strikes. What it actually was was an admission that the Rockies’ starting rotation was so atrocious that they’d rather have relievers taking up more innings than continue to trot Jaime Moyer and Jeremy Guthrie to the mound every five days. This deployment of smoke and mirrors is a lot like what we see from our politicians every day and we can’t seem to get rid of any of them.
The hilarious part of this whole story is that by the end of August, people were actually buying what the politician was selling. The Rockies went 18-10 from August 6 to September 4 to fuel a change in many people’s opinions, even though they had gone 12-28 prior to that run. Even some of our own folks here at the Bulletin were starting to be persuaded, though not without a grain of salt.
On August 24, in 3 Up – 3 Down, Ned said, “Maybe, just maybe, the team has stumbled onto the solution to the eternal problem of pitching in Coors Field. And maybe the characterization of this pitching strategy will change from desperate to revolutionary.”
On September 1, again in 3 Up – 3 Down, Kevin stated, “Third, the four-man rotation is working and might even be a good idea.” As I said, giant grains of salt came included in those quotes, but you can see that sane people were grasping at straws when a glint of hope surfaced through the desolation of this season.
In other media outlets, the same thing was happening. On August 27, ESPN’s Michael Bertin said, “Still, the Rockies’ experiment seems to be working for now, if only a little bit.” This was preceded by a deluge of backhanded compliments, but this statement was not made in jest.
On August 30, Troy Renck of the Denver Post wrote: “Since the switch, the Rockies have pitched better. In the first 65 games, the starters were 13-20 with a 6.28 ERA, threatening the 1996 Tigers’ 6.64 ERA as the worst ever. In the last 64 games, they are 11-26 with a 5.61 ERA. The relievers have morphed into part piggybackers, part vultures, increasingly picking up wins. They were 12-10 with a 4.00 ERA before the switch, compared with 17-10 (4.52) since the move.” At this point, Renck’s level of expectations had dropped so far that he was arguing that a 5.61 ERA was a good thing. Decreasing ERA by more than half a run would normally be huge, but not when it still remains the worst in the league. He also fails to acknowledge that the relievers’ ERA had actually gotten worse.
On September 4, in his weekly power rankings, ESPN’s Jonah Keri said: “The Rockies are 40-36 since moving to a four-man rotation, and they’re planning to stick with it next season. There may very well be some merit to the plan, given the notion that fresh relief pitchers often pitch better than gassed starters, not to mention the potential benefit of giving hitters a whole new look sooner in games than they’re used to seeing.” Keri seems to be so delirious at this point that he mistook 30-38 (the Rockies were 55-78 at this point) to be 40-36.
Even players were susceptible to the madness. On August 30, in that same Post article by Renck, piggybacker Adam Ottovino said: “If you get good pitching out of it, and the starters give us four or five innings, then usually we will be in the game. It’s similar to a quality start (six innings, three runs)…We have shown we can win like this.” I suppose this statement is technically true, but the Rockies were losing more than they were winning. Maybe his last sentence was meant to be ironic.
The madness peaked on August 30 when the Rockies announced that the scheme would continue into 2013. The entire Rocky Mountain region groaned in unison, but whichever curse was responsible for this nonsense finally broke on September 14 when Bill Geivett, the new sort-of general manager, announced the team would be returning to a traditional five-man rotation, with pitch counts relaxed to 90-100 pitches. In other words, back to a normal pitching convention. The article by Renck that featured this news also contained this telling quote – “There were also practicality issues. Had the Rockies kept a strict four-man rotation, it would have necessitated mirroring the experiment in Triple-A, preventing fewer pitchers to be available on short notice because of rest issues” – as well as this one – “Rockies players said Friday that they were pleased with the decision to use a more conventional rotation, because the four-man staff had never been fully embraced by pitchers or position players.” So, not only did the players hate the idea, but they would have had a hell of time installing it on the minor league clubs.
As it turns out, Geivett himself never believed in this experiment, stating as much in an interview with Keri back in 2006 when Keri breached the subject of both a paired-pitching and a four-man rotation. Geivett says “Then you get into the tandem starting pitching system (where four pairs of pitchers work on a four-day rotation) and now you have the problem of a guy who throws four shutout innings, then a guy comes in and gets knocked around in the 5th. The appearance of your statistics, the fans’ reaction, players’ reaction, agents’ reaction, these things would be tough to manage. Both of these are certainly creative ideas, things that should be looked at and studied. But putting together a legitimate pitching staff with great balance in terms of starting pitchers getting relatively deep into the game, having a structured bullpen with roles, I think that’s the way to go. I study the game and have respect for everything that went on in the past. There’s a reason why it went from four to five.”
The miracle in all of this is that the Rockies stubbornly stuck with the system through the final three weeks of the season, even after announcing they were done with it. Though, what they stuck with was not a four-man rotation, nor did they strictly adhere to the 75-pitch limit. The only component of the system that made it to the end intact was the paired-pitching/piggyback concept. Thanks to that stubbornness, we have two extra weeks of data for a total of 97 games.
To begin with, many of the Bulletin’s readers, along with several of our writers, noted that the four-man rotation disappeared when Jhoulys Chacin rejoined the team after recovering from an injury and became the fifth man in the rotation. The funny thing is that the four-man rotation wasn’t employed all that much prior to that day. In the 55 games before Chacin returned, Rockies starters only pitched on three days of rest 20 times, or less than half.
Furthermore, the 75-pitch count was also not a strict limit. Rockies starters exceeded 75 pitches 54 times out of 97 games, or more than half. While quite a few of these only exceeded it by a couple of pitches (and the team acknowledged that it wasn’t necessarily an absolute limit), it still shows that the organization isn’t even capable of sticking with their own set of rules. This was just more evidence of how dysfunctional is the Rockies franchise.
The one component they stuck with was the piggybacking concept. However, they don’t deserve much credit for this since the starters commonly hit their pitch limits before the fifth inning, leaving the team with no choice in the matter. Now that the season is over, we can quantify the success of the system.
Back on September 11, Travis Lay laid out a comparison of the starters’ and piggybackers’ stats before and after August 1. He showed that the starters had virtually no change in any numbers, particularly their average number of pitchers per start. However, the most interesting statistic was that the piggybackers’ WHIP’s increased to well over 2.00 (Moscoso’s actually went down to 2.167). This shows that the piggybackers weren’t really providing an improvement over the starters, other than being much luckier in terms of stranding runners.
O’Dowd also claimed that the new scheme would reduce injuries to pitchers, among other things (check out this New York Times article for quotes).
Juan Nicasio, Christian Friedrich, and to a lesser extent, Drew Pomeranz (who was given lots of extra days off due to fatigue) all suffered injuries during the experiment. While those injuries can’t be directly attributed to the experiment, it’s definitely a strike against reducing injuries. If anything, it proves that injuries are unique and almost impossible to predict, let alone prevent.
Another reason quoted for the switch was to reduce the number of runs given up. In the 65 traditionally pitched games, the Rockies gave up 373 runs for an average of 5.74 runs per game. In the 97 remaining games, they gave up 506 runs for an average of 5.22 runs. Like those ERA numbers above, it’s an improvement over what they were allowing, but still terrible when compared to the rest of the league.
Finally, the main goal – to win more games. The Rockies were 25-40 when the experiment began and were on pace to finish 62-100 without it. Instead, they went 39-58 the rest of the way to finish at 64-98 (incidentally, 39-58 projects to 64-98 over a full season), gaining two whole wins. While most of us recognize this as no noticeable improvement, I’m sure O’Dowd and the owners (the Monfort brothers) are celebrating their success. Hell, they’re probably in a room right now trying to figure what seven-to-eight pitchers are going to get them to the World Series next year.
Now that we know the system provides no improved success and does not help prevent injuries, we should ask if there is anything to be learned from these 97 games. As it turns out, the results show that the only way to sustain success is for starting pitchers to complete at least five innings – a fact that everyone already knew – which removes the need for piggybackers in the first place. Here is the breakdown of wins by starting pitchers’ innings pitched:
- Less than 5 IP – 5-15
- 5 IP or more – 20-25
- 6 IP or more – 16-9
- Less than 5 IP – 21-45
- 5 IP or more – 18-13
- 6 IP or more – 4-3
Interestingly, in those 20 games where the starter pitched on 3-days rest, the Rockies went 9-11 and showed the same splits when comparing innings (less than 5 IP – 3-11; 5 IP or more – 6-3). It’s unfortunate the Rockies were so flaky in instituting the concept as the four-man rotation proved to be the most successful.
After looking at these results, the only conclusion we can come to is that the Rockies need better starting pitchers if they want to win; pitchers that can get through five innings and do it without getting lit up like Guthrie at Coors Field. Until that happens, we can look forward to more smoke and mirrors and snake oil from a politician we can’t even vote out.