How did contenders acquire their starting pitchers?

Author's Note: This article is based on the idea that we might learn something about how the Astros should construct a starting rotation, based on how contending teams built their starting pitching.

Ian Kennedy, with the Diamondbacks, became the first NL pitcher to achieve 20 wins this season.  Although he may be a dark horse, he has to be in the conversation for the NL Cy Young Award.  That's not bad for a 26 year old pitcher in his second full season in the majors.  For Kennedy's 20th win, the headline in my local Austin, Texas newspaper referred to Kennedy as a "Yankees Castoff."  That seems to have a negative connotation--as if he was just unwanted refuse.  And, if the headline writer really meant it that way, I think the headline was sloppy.  Kennedy was a first round draft choice of the Yankees in 2006, and as a Yankee minor leaguer, he was ranked as the 45th best prospect in baseball.

Why did the Yankees trade Kennedy?  That's unclear. Going into 2008, the Yankees had a starting pitcher prospect trio of Hughes, Chamberlin, and Kennedy.  Kennedy's fastball velocity isn't as fast as the other two, and the Yankees apparently viewed him as expendable.  As one Yankees' blogger said this year, the Yankees simply may have traded the wrong guy out of the three prospects.  Kennedy had suffered some injuries, but they were not directly related to his arm.  I don't recall any great concern or misgivings among serious Yankees' fans at the time of his trade.  It's true that Kennedy had some rough outings in his brief outings with the Yankees' major league club--but surely he wasn't dismissed based on a sample size less than 40 innings.  (However, I think there is some feeling that Yankees' fans and ownership can be unusually impatient with their pitchers.)  Kennedy's minor league record was good.  Kennedy kept his FIP close to 3 or below; his K/9 was 9.9 and his BB/9 was below 3.  With a fastball velocity in the 91-92 range, and good off-speed and breaking pitch offerings, it's hard to find a closely comparable Astros' pitcher; maybe Lyles if he can gain a tick on his fastball or even Paul Clemens.  In the end, I just have to chalk this up to a good trade by the D-Backs.  The Diamondbacks apparently had a better appraisal of Kennedy's potential than the Yankees.

Without spending too much time on the Diamondbacks, their rotation has benefitted greatly from trades.  The D-Backs set up their 2011 contending team by trading for two ML-ready starting pitcher prospects.  The D-Backs obtained Daniel Hudson out of the White Sox organization--and if you look at his minor league numbers they are quite similar to Kennedy's.  Yet Hudson was another young pitcher who seemed to be underrated among prospectors, but turned out to be a solid major league starter.

This article isn't intended as an evaluation of the D-Backs trading acumen.  I'm leading up my real question.  How did the contending teams construct their rotations?  Do they rely on the draft?  Do they have to use the free agent market?  Or do they make trades, like the D-Backs' example.  I think this can provide some context for what we should expect from the Astros if they are to put together a good rotation.

I looked at the top six starting pitchers for each of the following teams, all of which continue to be in contention for the playoffs: Phillies, Braves, Brewers, Cardinals, Diamondbacks, Giants, Yankees, Red Sox, Rays, Tigers, Rangers, and Angels.  I categorized the acquistion of each pitcher as follows: Amateur Signing (whether drafted or undrafted); Prospect Trade (trade of minor leaguer or young pitcher with less than rookie eligibility); Veteran Free Agent signing (excluding rehab projects); Buy Low signing (includes veteran rehab or comeback projects, Rule 5 picks, and minor league free agents); Veteran Trades (trade of major league veteran pitchers).  I'll summarize the results for each category, with the percentage of contenders' starting pitchers in each group.  The D-Backs' trades for Hudson and Kennedy fitted into the prospect trade category.

Amateur Signings (46%)

Almost half of the contending teams' starting pitchers were products of the team's farm system, whether drafted or signed as an amateur free agent.  There is quite a bit of variance among the 12 teams.  The Rays and Giants are outliers, with the Rays producing all six starting pitchers, and the Giants five of six starting pitchers, out of their farm systems.  Both teams relied upon a combination of high first round picks and late round picks.  Without those two teams, amateur signings would constitute only 36% of the contenders' rotations.  The D-Backs and Brewers relied the least upon amateur signings, with 1 each in their rotations.  Most contending teams had 2 or 3 amateur signings in their rotations.  The American League contenders were somewhat more likely to rely upon amateur signings than the National League.

Prospect Trades (6%)

Trading for pitching prospects is risky business, and fairly difficult.  So, perhaps it's no wonder that this isn't a significant source of starting pitchers.  The D-Backs reduced their risk by trading for two major league ready prospects, and they dominate this category.

Sign Veteran Free Agents (19%)

Signing veteran free agent starting pitchers can be expensive, but it's the second largest source of contenders' starting pitching.  The Red Sox have the most veteran free agent signings (3 of 6 starters) in their rotation.  Most teams have 1 or 2 free agent signings in their rotation.  If I hadn't split off some free agent signings into the "buy low" category, the Yankees would have the most free agent signings in their rotation.

Buy Low Acquisitions (11%)

This category is aimed at cheap acquisitions, whether in the form of "rehab projects" on the free agent market, Rule 5 draftees, or minor league free agents.  The Rangers, Yankees, and Brewers each have two pitchers in this category.  In most instances, the buy low pitchers are used as a spare (No. 6) starter or No. 5 starter.  However, the Yankees' signing of Bartolo Colon and Freddy Garcia (rehab projects) has produced almost 300 innings of 3.79 ERA starting pitching in the heart of the team's rotation.  Are the Yankees lucky, or did their scouts know something that other teams didn't?  American League contenders relied on buy low acquistions more than National League.  The Yankees and Red Sox needed some buy low sources, because their payrolls put them in luxury tax territory.

 Trade for Veteran Pitchers (18%)

Trading for a veteran starting pitcher is almost dead even with signing a free agent pitcher.  The Tigers relied upon trades the most, using 3 trades to bolster the rotation.  The Brewers may have been the most successful in using the trade, adding Marcum and Greinke to their rotation.  National League teams were somewhat more likely than American League teams to use trades.

Summary

The most likely path to a contending rotation is to acquire starting pitchers from several sources.  It is difficult to follow the Giants' and Rays' route of totally relying upon the farm system to produce winning rotations.  Both teams required multiple very high draft picks, plus both organizations have great skill in scouting pitchers.  Given the attrition rate among amateur pitchers, the Rays and Giants probably had some luck too. The Astros have not shown a propensity to use their highest picks on pitchers, and that alone means that the team is unlikely to follow the Rays' model.

  Trading for pitchers, whether prospects or veterans, requires being in the right place at the right time---and arriving with some trading ammunition.  The Brewers and D-Backs stand out as good traders, but it's noteworthy that both teams had good farm systems at the time, which provided enough depth to trade away value.  Undoubtedly, a good bit of luck is involved in buy low acquistions.  I say that without discounting the scouting skills that enable teams to make those bets.   Free agent signings may be second in frequency to amateur sources, but the impact of free agent pitchers may be greater.  For many contending teams, the farm system fills out the middle or lower portions of the rotation, with free agent signings more likely to bolster the top spots.  Look at the two free agents in the Phillies' rotation (Halliday and Lee), and you see what I mean. There are exceptions, though, where the team uses it farm system to supply the ace, and goes to other sources for the remainder of the rotation.  The Brewers provide this example: Gallardo, the ace, is the only member of the rotation to come from the farm system. The results of my tabulations are shown below.