In 2013, the Brewers’ rotation featured two of the National League’s dependable starters. Yovani Gallardo pitched his fifth consecutive season with 100+ IP in the NL, while Kyle Lohse pitched his third consecutive season with 100+ IP in the NL. Lohse was one of the club’s pleasant surprises in 2013, as the control-oriented righty once again limited the damage. Although the Brewers surrendered a first round draft pick for Lohse’s services, it now seems difficult to imagine how 2013 would have unfolded without his veteran production. Gallardo worked one of his most disappointing seasons of his career, although it did follow his previous pattern of jumping between the top-and-middle rotation. Taking their stretches of 100+ IP into consideration, the Brewers’ rotation featured two of the Senior Circuit’s solid #2 starters.
Dependable Starters and Roster Spots
The refrain that teams need dependable starters is a difficult one to transcribe and breakdown. In the abstract, the idea that teams need starting pitchers that work 100+ IP year-in and year-out seems self-supporting; clubs simply cannot afford to go to the well of replacement starters time and time again. The expectation is that a dependable starter will be a consistent, serviceable starter. In reality, however, the picture is much murkier. Even if a pitcher consistently works a high level of innings, his performance is not necessarily like to remain the same. In fact, while analyzing the value of dependable starters last offseason, I found that the standard deviation of dependable starters’ runs prevented rankings was rather high; given any two years, a dependable starter was likely to shift at least one rotation spot. Last year’s class of dependable pitchers included 30 NL examples to choose from; this year, a class of 25 dependable starters worked in the NL and experienced similar fluctuations.
Why does this fluctuation matter? From the perspective of a GM, the purpose of signing players (especially pitchers) would seemingly be to build a reliable core that can produce year after year. Yet, given the extent of performance fluctuation for dependable starters, one might argue that teams need not sign dependable starters for their respective rotations. However, this extreme might be as unrealistic as the expectation that a pitcher should produce at a reliable, consistent level each year. Frankly, it appears that GMs keep dependable pitchers around to avoid the black hole of transactions that follow replacement starters; when one pitcher goes down, GMs almost always need to use two or more starters to replace that pitcher’s starts.
This is one of the shortcomings in replacement theory; although one might expect a below-replacement starter to be replaced by his club, if that starter is dependable, the resulting lack-of-transactions will make him a favorable candidate for a roster spot (regardless of production). The San Francisco Giants are the best example of this phenomenon; despite paying large sums for several replacement level starters in 2013, they have once again handed out contracts to feature those starters in their 2014 rotation. The simple reason? Keeping those starters around could limit the potential unknown replacements that might otherwise be necessary in the absence of Tim Lincecum, Ryan Vogelsong, and maybe even eventually, Matt Cain (who was notably below average in 2013).
The general scarcity of dependable starting pitchers could also explain their ability to keep jobs. Edinson Volquez is a particularly good example of this issue. Despite pitching no better than a group of replacements over the last three seasons, Volquez’s durability landed him a $5 million spot in Pittsburgh. While the Pirates could probably recall three minor leaguers and match Volquez’s production for $1.5 million or less, they paid $5 million for his service and durability. If he stays in their rotation for the season, that’s one less spot that their club needs to worry about.
If you might protest that Volquez’s presence in the Pirates rotation hurts the Pirates’ contention chances, consider the Cardinals and Jake Westbrook. Westbrook hasn’t been better than a #4 starter in three consecutive seasons, but that hasn’t hurt the Cardinals’ ability to get to the playoffs. Dependable starters are so rare that even those “replacement level regulars,” like Volquez and Westbrook, will probably face some replacement starters throughout a season, anyway. Even a contending club can survive a 15-run below average performance (the last three NL champs have had three such pitchers, as well as Jaime Garcia (-14 runs prevented) and Barry Zito (-13 runs prevented). Contenders do not need good starters at every spot; heck, they don’t even need good rotations sometimes).
Brewers: Flexible Rotation
In this context, the prices for Lohse and Gallardo seem downright reasonable (the Brewers owe a maximum of approximately $46 million for their services in 2014 and 2015). While questions remain about whether Gallardo will be able to correct his 2013 slide, he still works at an above-replacement level for a sizable number of innings. Lohse’s price tag might seem more extreme, given that it must include the surrendered draft pick. However, it is his spot in the rotation that allowed the Brewers to stay put with their rotation for 2014, while the organization’s youngsters found their way in three other shuffling replacement spots.
2014 hardly hangs on Lohse and Gallardo, alone, but having these arms at the top of the rotation will help Doug Melvin to rotate his youngsters and present the most effective rotation. With the knowledge that he may not need to replace at least two starters (for, Wily Peralta might prove rather durable), Melvin may be more aggressive with his fourth and fifth rotation spots. Marco Estrada, Tyler Thornburg, Will Smith, Mike Fiers, and Jimmy Nelson might not be likely candidates to work a full slate of starts, but given the dependable top rotation, they do not need to work full seasons. Here, the “five man rotation” becomes something more flexible, to the point where an eight- or nine-man rotation might not be problematic for the club.
Compared to the 2012 NL, in 2013 the Senior Circuit lost fifteen dependable starters (pitchers with 100+ IP in three, four, or five consecutive years). Several of those starters simply did not have jobs (such as Carlos Zambrano and Randy Wolf), while others suffered injuries (such as Garcia and Wandy Rodriguez). A handful of these starters also left the NL for the Junior League.
|Starter (Dependable Years / Rank)||2013 IP||2013 Runs prevented|
|J. Cueto (5 yr. #23.8)||60.7||8|
|R. Dempster (5 yr.. #28.6)||AL||AL|
|W. Rodriguez (5 yr. #30.2)||62.7|
|C. Billingsley (5 yr. #31.6)||12||1|
|C. Zambrano (5 yr. #38.0)||DNP||DNP|
|R. Wolf (5 yr. #42.2)||DNP||DNP|
|A. Harang (5 yr. #46.2)||23*|
|K. Correia (5 yr. #60.6)||AL||AL|
|T. Hanson (4 yr. #32.0)||AL||AL|
|C. Volstead (4 yr. #61.8)||8.3||-6|
|R.A. Dickey (3 yr. #10.7||AL||AL|
|R. Halladay (3 yr. #15.7)||62||-20|
|A. Sanchez (3 yr. #31.3)||AL||AL|
|J. Garcia (3 yr. #38.3)||55.3||-2|
|B. Norris (3 yr. #60.7)||AL||AL|
|*AL not included|
However, the NL gained nine new dependable starters. This group included some starters that worked back from injury (like Chris Capuano, Volquez, and even Jordan Zimmermann), as well as pitchers that were dependable in the AL, and now have three years in the NL (such as Zack Greinke and Cliff Lee). The top four pitchers in this group hardly fluctuated from their 2012 performances, but the bottom five easily made up for that. One could argue Volquez did not fluctuate much, but I’m not sure that argument looks as appealing at the bottom of the rotation.
From 2012 to 2013, a group of nine starters graduated to four- and five-consecutive 100+ IP seasons, including Gallardo. With more years under their belts, everyone in this group except the Cincinnati Reds’ youngsters fluctuated rotation spots. Bailey solidified himself as a #2 starter after several years as a #4 starter, and Mat Latos made another argument to be considered as one of the league’s top righties (as someone like Max Scherzer receives potential contract extension hype, it seems that Latos is a much better case for big money, for many reasons).
Finally, in 2012 there were 15 starters that worked five consecutive seasons with 100+ IP. Only seven of those starters made the jump to six-consecutive years in 2013. Save for Clayton Kershaw and Tim Lincecum, this group also fluctuated wildly between this year and last.
Fluctuation and Dependability
Ultimately, we ought not equate a dependable starter with a consistent performance on the mound. Furthermore, a pitcher that produces at levels that fluctuate over several seasons is not necessarily unreliable. A starting pitcher can be dependable and produce at levels that fluctuate several rotation spots over a couple of season. Among this group of 25 starters, one standard deviation in performance ranking is approximately 21 spots; this is easily the equivalent of shifting between one rotation spot to another rotation spot within the span of one season. Moreover, this level of fluctuation does not necessarily change as time marches forward; 3-year dependable starters fluctuated 23 ranking spots per season, 4-and-5 year dependable starters fluctuated 17 ranking spots, and 6-year dependable starters fluctuated 22 ranking spots per season. Ultimately, at any end of the rotation, and at nearly any level of dependability, one can expect a regular starting pitcher to fluctuate between rotation spots from one season to the next.