The Good, the Bad, and the Cheap: The Cost Efficiency of 162 Startsby J.P. McIntyre
January 22, 2007
In today's game, teams break Spring Training camp with 162 games ahead of them and plan for a five-man rotation to cover all the starts in those games. Of course, teams know there will be injuries, trades, and demotions because of ineffectiveness, and possibly an occasional double header to make up rainouts, so that original five-man rotation is not going to cover those 162 starts. In other words, stuff will happen.
Let's look at how both leagues did in terms of total starts by the five pitchers who had the most starts on their team. Note these numbers are not necessarily by spot in the rotation, just total starts. For instance C.C. Sabathia was the Indians' No. 1 one starter, but he had only 28 starts, fourth on his team.
#1 #2 #3 #4 #5 SUM DIFF AL AVG. 32.3 30.7 27.4 22.1 17.9 130.4 31.6 NL AVG. 33.3 31.4 26.0 19.6 16.4 126.7 35.3In fact, on average, teams from both leagues averaged 11 different pitchers to make those 162 starts (here is a full breakdown). We see that the difference between the top five's number of starts and the full 162-game season is about the equivalent of another rotation spot for the average team. Let's look at the outliers (for a full list of all teams, click here):
#1 #2 #3 #4 #5 SUM DIFF CHA 33 32 32 32 30 159 3 COL 32 32 32 31 27 154 8 TBR 25 24 21 16 16 102 60 KCR 29 21 20 15 12 97 65The White Sox covered almost all their starts with their original five starters. That wasn't necessarily a good thing when we look at their starters' overall performance, especially with Brandon McCarthy waiting for his chance. Meanwhile, the Devil Rays' and Royals' rotations were a dysfunctional mess in terms of stability, and their performances mirrored that. The Royals were last in the AL in starters' ERA (5.85), while the Devil Rays were 10th in the AL with a 4.96 ERA.
Let's look at how teams did in 2006 in terms of payroll efficiency with their starters. I am using Net Win Shares Value, created by Dave Studeman. To paraphrase Dave's definition, Net Win Shares Value essentially estimates the "expected" production from a player based on how he was signed (as a free agent, arbitration-eligible or not eligible for arbitration) and how much he was paid, then compares that to how he actually did. The difference is multiplied by the average amount teams paid for each Win Share Above Bench last year. If the number is positive, the player was a relatively good deal for the team; if not, not.
The 2007 THT Annual lists most of the players' Net WS Value, so one perk of purchasing the book is access to all the players' info. The following list includes all pitchers classified as starters (it doesn't include bullpen guys who made spot starts):
AL NET WS Value (in millions) NL NET WS Value DET 24.9 SFG 16.5 OAK 16.9 ARI 16.2 TBR 15.9 COL 14.3 LAA 13.0 FLA 14.2 TOR 11.8 CIN 12.5 MIN 6.9 SDP 11.3 TEX 3.4 MIL 7.9 CLE 2.9 PIT 6.6 CHA -.1 LAN 4.5 BOS -2.5 STL -2.3 NYA -4.1 ATL -4.8 SEA -5.7 NYN -5.7 KCR -7.4 PHI -5.8 BAL -8.4 HOU -8.7 CHN -10.6 WAS -14.0The Detroit Tigers had the most cost-efficient rotation by a considerable margin because not only did their starters pitch well, but most were signed to inexpensive contracts.
Net WS Value GS Verlander 8,003,000 30 Robertson 6,281,000 32 Bonderman 4,636,000 34 Rogers 3,282,000 33 Miner 1,378,000 16 Ledezma 1,362,000 7 Maroth -145,000 9Miner and Ledezma earned value in the pen also, so not all their positive Net WS Value came from starting. Roman Colon also made a start, but overall, the Tigers' rotation was stable in terms of who was starting while being cost effective. Let's now look at a team that was the opposite of the Tigers, the New York Yankees:
Net WS Value GS Wang 9,689,000 33 Karstens 1,399,000 6 Rasner 824,000 3 Wright -1,184,000 27 Ponson -1,807,000 3 Mussina -2,590,000 32 Chacon -3,930,000 11 Johnson -6,481,000 33Despite having one of the best values in baseball in Chien-Ming Wang, the Yankees were 11th in the AL in terms of payroll value for their starters, and that is not even including the $9 million they paid Carl Pavano to heal slowly. Mike Mussina had a solid season, but the Yankees were paying him so much that it was not cost effective. The Yankees received 125 starts from Wang, Johnson, Mussina, and Wright, but used eight other pitchers to make the remaining 37 starts, some with some rather inefficient results.
The Chicago White Sox were very stable in terms of who started their games. Unfortunately for the White Sox, they didn't receive a very efficient return from all their starters:
Net WS Value GS Contreras 2,644,000 30 Garcia 2,342,000 33 Garland 838,000 32 Vazquez -1,582,000 33 Buehrle -4,212,000 32The White Sox had no starter who truly shined, even though management was writing some large checks. A more efficient team was the Oakland A's:
Net WS Value GS Haren 7,062,000 34 Blanton 4,332,000 31 Saarloos 2,096,000 16 Zito 1,087,000 34 Harden 1,101,000 9 Loaiza 988,000 26 Windsor -585,000 3
Because Harden was not receiving a large salary, his injury did not hurt the A's financially, unlike Mark Mulder of St. Louis, who had a -$9.9 million Net WS Value. Oakland general manager Billy Beane's market philosophy allows him to get a high yield on the money the A's pay their starters.
Let's now look at the disaster known as Washington Nationals:
Net WS Value GS Perez 909,000 3 Patterson 482,000 8 Hill 301,000 6 O'Connor 19,000 20 Day -302,000 5 Traber -882,000 8 Armas -1,444,000 30 Hernandez -1,884,000 24 Drese -1,893,000 2 Bergmann -2,324,000 6 Astacio -2,838,000 17 Ortiz -4,068,000 33The Nats' rotation, which was not exactly constructed to last, was ravaged by injuries and ineffectiveness. Ramon Ortiz started the most games, which means the Nats had a replacement-level pitcher start 33 games. The Nats had 14 different pitchers start games, and none were effective. If this type of wreck occurred on a freeway, traffic would be shut down for two days while the carnage was cleared and the body bags counted.
In general, home-grown, inexpensive talent is the best route to building a cost-efficient rotation. Throwing good money after bad in hopes something sticks is the worst way to go. There is nothing novel in these premises. Teams that can develop their own pitchers have a huge financial advantage over those that build a rotation primarily through free agency or trading for established, expensive starters. That does not mean a team can't be successful with inefficient spending; the St. Louis Cardinals were not efficient, but defeated the most efficient team, the Tigers, in the World Series.
I'll wrap this up by showing the 10 most-efficient and least-efficient starters in terms of salary value:
Best NET WS Value Worst Net WS Value Webb 11,412,000 Pineiro -10,015,000 Liriano 10,376,000 Mulder -9,910,000 Arroyo 10,068,000 Od. Perez -8,447,000 Wang 9,689,000 B. Colon -8,310,000 Carpenter 9,561,000 Russ Ortiz -8,219,000 Je. Weaver 9,277,000 Chen -8,001,000 Capuano 8,229,000 Park -7,496,000 Verlander 8,003,000 Marquis -7,347,000 Bedard 7,902,000 Ra. Johnson -6,481,000 Haren 7,590,000 Towers -6,121,000
References and Resources
Resources: The Hardball Times 2007 Annual and BaseballReference
J.P. McIntyre is a transplanted Ohioan residing in California and authors the blog Many Go; Few Understand and co-authors Reds and (Blues).com. He welcomes comments via email.