Where It Went Wrong for the Tribe (Part 4)

In 2005, general manager Mark Shapiro patched together a starting rotation that included three pitchers at the age at which many pitchers start to come into their own, a journeyman, and a top-of-the-rotation type coming off serious injuries. One of the three youngsters, C.C. Sabathia, was supposed to develop into an ace; another, Cliff Lee, looked as if he could be a number two starter; while the third, Jake Westbrook, looked to be a number three starter. Kevin Millwood was a mystery, coming off a lackluster year and a serious injury, and Scott Elarton looked as if he could hold down the five spot well enough for the Indians’ offense to have a chance to win his games. Let’s look at what transpired:

                PRC    ERA+     WSAB     GS    IP
Millwood       100     145     9        30    192.0
Lee             88     109     7        32    202.0
Sabathia        85     103     7        31    196.7
Westbrook       69     92      2        34    210.7
Elarton         62     90      2        31    181.7

Because of Millwood and Lee’s performance, the Indians had a very solid five man rotation, totaling 404 Pitching Runs Created over 1006.2 innings pitched. Also important, these five starters combined for all but four of the Indians’ starts (which went to Jason Davis).

However, Millwood and Elarton became free agents, and their performances couldn’t be expected again. Millwood led the 2005 AL with a 2.86 ERA, but his Fielding Independent Pitching (FIP) was 3.75 and his Expected FIP (xFIP) was 3.98. Elarton’s FIP was 5.07 and his xFIP was 5.18. Both were benefactors of luck or a solid defense behind them. Both were going to command too much money on the free agent market, so Shapiro cut them loose (he did try to re-sign Millwood). As with the bullpen, Shapiro was faced with a reconstruction project for his starting rotation, and like the bullpen, the pieces Shapiro brought in to shore up the rotation crumbled.

In early December, Shapiro signed Paul Byrd to a two-year, $14.25 million dollar deal with a club option for the third year. Shapiro had been negotiating with Kenny Rogers and almost signed him, but the deal broke down because Shapiro did not want to commit to a two-year deal. Three days after Byrd signed with Cleveland, Rogers signed a two-year, $16 million dollar deal with the Detroit Tigers. In January, Shapiro signed Jason Johnson to a one-year, $3.25 million dollar deal, completing the Indians’ rotation.

It is doubtful that Shapiro thought Byrd and Johnson would replicate Millwood and Elarton’s 2005 performance. Instead, he expected his other three starters to improve while Byrd and Johnson shored up the middle and back of the rotation. Here’s how it played out in 2006:

            PRC     ERA+    WSAB     GS     IP
Sabathia    100     138       9      28    192.7
Westbrook    84     109       6      32    211.3
Lee          74     103       4      33    200.3
Byrd         52     92        0      31    179.0
Johnson      20     90        -2     14    77.0
Sowers       40     125       4      14    88.3

Part of Shapiro’s plan worked, Sabathia and Westbrook had much better years. However, Lee dropped significantly from his 2005 performance while Byrd was a replacement-level pitcher and Johnson was below replacement.

In fact, Johnson was traded to Boston in late June, and Jeremy Sowers took his place in the rotation until he was shut down late in the year. The emergence of Sowers tempered the fallout of a very bad design by Mark Shapiro. The five starters that began the season only put up 330 PRC, 74 PRC less than the starting five in 2005. Even with Sowers’ strong performance, the Indians’ starters only mustered 370 PRC, 34 short of the previous year—about 3.5 team wins below 2005. The 2006 Indians had 10 starts from three other pitchers who finished the season with a combined -2 WSAB.

Most importantly, Shapiro spent just under $10 million dollars on Byrd and Johnson (the teams that picked up Johnson paid the league minimum for his services) in 2006, plus $7.25 million for Byrd next year. Let’s look at what might have happened if Shapiro had forgone the Byrd/Johnson plan, signed Rogers and started the year with Jason Davis as his fifth starter. It is easy to use a general manager for target practice when armed with hindsight, so I am not really faulting Shapiro for not finalizing the Rogers deal, but instead looking at what might have been:

             PRC    ERA+   WSAB  GS    IP
Sabathia     100    138     9    28    192.7
Rogers       84     115     8    33    204.0
Westbrook    84     109     6    32    211.3
Lee          74     103     4    33    200.3
Davis*       25      90     0    14     77.0
Sowers       40     125     4    14     88.3

* pitching at a replacement level

That combination would have amassed 407 PRC, slightly more than the 2005 squad. Signing Rogers for two years at $16 million would have been cheaper than the $17.25 million Byrd/Johnson failure. Kenny Rogers wasn’t a sure bet coming into 2006, let alone 2007, so Shapiro having reservations about signing him is understandable. After all, Rogers might have choked Carl Monday with a microphone cord by July.

Rogers was on the wrong side of 40, plus he had had a rather steep decline in the second half of 2005. However, Byrd and Johnson had more red flags than a construction crew assigned to re-route traffic from a bridge that collapsed in a river. Shapiro was already gambling with $17.5 million and would have been better off rolling the dice with Rogers.

The signs that Byrd and Johnson would be gas cans juggling hand grenades were in plain slight last offseason. In 2005, Byrd posted a 3.74 ERA, but his xFIP was 4.77, which was very close to his 2006 ERA. Most importantly, his K/9 IP, which was never that good at the major league level, dropped to 4.6. Pitchers with strikeout-rates that low usually don’t sustain success, plus Byrd gave up more hits than innings pitched in 2005, another bad sign.

Johnson was a 31-year-old journeyman who pitched only three WSAB in 2005 and had a K/9 IP under 4.00. He gave up 233 hits in 210 innings. He projected to be around a replacement level pitcher, but Shapiro felt the need to send $3.25 million his way. The Indians could have found at least a replacement level pitcher from their own system and saved the money they wasted on Johnson. Paying veterans seven-figure salaries to receive less than replacement level pitching is not to way to build a successful team.

Because of Sabathia, Westbrook, and Sowers, the Indians starting rotation did not fall into Lake Erie like the bullpen and defense did. Having a guy like Kenny Rogers instead of Paul Byrd and Jason Johnson would not have saved the season; the bullpen and the defense created too much damage for that to happen. However, the Indians starting rotation that Shapiro built was not that of a contender—the Tigers received 372 PRC from their top four starters, plus picked a fair share of PRC from their various fifth starters.

The good news is that the starting rotation is probably the easiest thing to fix for next year. Sabathia looks to be coming into his own and Westbrook is a solid middle-of-the-rotation guy. Lee scuffled in 2006, but finished strong and has the type of stuff to bounce back in 2007. Sowers will probably experience some growing pains next year as he appeared to receive some luck this year (his FIP was a full run better than his ERA), but barring an injury, he should develop in a middle-of-the-rotation guy for 2007 with more upside for the future. The Indians are a quality starter away from a very solid rotation. Byrd would be an expensive, but probably somewhat effective long reliever/spot starter, but the mistakes of the past cannot be undone. However, they can be fixed for a price, and most importantly, avoided in the future.

Using Recurrent Neural Networks to Predict Player Performance
Technology is rapidly advancing possibilities in decision-making.

Print This Post

Comments are closed.