From the Floor to the Ceiling of the 2017 Indians Rotation

Ubaldo Jimenez was a funny pitcher for the Indians. No other pitcher in recent memory had a greater divergence in potential outcome each start. For example, in 2013 he had one span where he allowed one run over six innings in his first game, then 14 runs over six innings in his next two starts combined, and came back two starts later with seven shut out innings. He ended that season with a 3.30 ERA and it lead to many great discussions including whether a team should prefer a pitcher who, all other things being equal, was consistently average every start or took a more Jekyll and Hyde approach to pitching.

Cumulative stats won’t identify these types of pitchers. A pitcher who allows four runs in two innings then none in eight next time out will have a 3.60 ERA over those two starts, the same as a pitcher who allowed two runs in both starts, but only went five innings each start. The pitcher who has great games offset by terrible games will likely have more decisions and for a low scoring team, would give his team a chance to win more often while the pitcher who was reliably mediocre would often hand the game off to his bullpen to decide and would be favored by a team with a potent offense.

Of course, these hypothetical pitchers don’t exist. In reality, even those pitchers who are consistent still have highly variable starts and those who go to extremes don’t do so with any predictability. To visualize what the Indians have going for them this year the chart below shows the game scores for the last 30 starts for each of the Indians starters heading into 2017.

If that looks like a jumbled mess, it’s because it is. Even so, we can still extrapolate some information from the squiggly lines. Mr. Consistency, Josh Tomlin, hasn’t had a game score above 70, but outside of the period last August when he his lack of innings the previous two seasons caught up with him, he only had two bad games. Trevor Bauer, on the other hand, had a high game score of 86 while bottoming out at 8 on August 3rd when he allowed eight runs (seven earned) and only struck out one in 2.2 innings against Minnesota. The chart below is a little clearer as it breaks down the average game score for each pitcher over their last 30 starts as well as their high, low and number of games in each range from 0-30 (terrible), 31-50 (bad), 51-70 (good) and 71-100 (excellent).

If you look at traditional statistics, Bauer and Tomlin had similar seasons in 2016. By record (Bauer: 12-8, Tomlin: 13-9), innings (Bauer: 190, Tomlin: 174) and ERA (Bauer: 4.26, Tomlin: 4.40) you could call them comparable although more modern metrics like FIP (Bauer: 3.99, Tomlin: 4.88) and fWAR (Bauer: 2.7, Tomlin: 1.0) would favor Bauer. You could easily use this to say that Bauer is the better pitcher and should be fourth in a rotation that otherwise makes sense. That is until you look at the game by game results.

While Tomlin essentially never throws a great game (his most recent game score above 70 was September 15th, 2015 vs KC), he throws a good game more often than any pitcher outside of Kluber. In addition, five of those seven terrible games came during his rough August before he was (temporarily) removed from the rotation. While generally it is fair to say that you can’t pick and chose which games to include in analysis, in this case Tomlin had just surpassed 120 innings after throwing just 95.1 the season before (combined MiLB and MLB) coming off shoulder surgery. Had the Indians not been in the heat of a play-off hunt, he likely wouldn’t have pitched those games at all. If you look back to Tomlin’s 2015 instead of the end of his 2016, he had only one game score below 50 (27 on September 26th vs KC) and nine in the good to great range.

At the same time, Bauer’s inconsistency is readily apparent. His best is among the best in baseball (86 game score from June 22nd vs Tampa when he struck out 10 in a complete game, allowing one run), but so is his worst. In his last 30 starts, nearly half were below what is considered an average start (50) including the Indians worst start of 2016. He’s still a talented pitcher (and definitely has a higher overall ceiling than Tomlin), but for a team looking to repeat as AL Champs, the 5th spot is the safest place to hide his variability.

As mentioned in the opening, a team that has trouble scoring runs would prefer a pitcher who alternates great and awful games over one who is consistently average. It’s better to win a game 2-0, then lose one 10-2 than to lose both 5-2. Back in 2013, Ubaldo had 12 games with a game score of 64 or above and the Indians won all 12. He also had ten below 50 and the Indians went 4-6, losing all that his game score was 30 or below. Obviously, any team would prefer a rotation full of pitchers like Kluber who average a game score above 60, but since that’s not possible, Tomlin may be the perfect back end man. Beyond the fact that his home run rate from last season is unlikely to repeat (at least to that extreme), he’ll give the Indians improved offense a chance to win each time he pitches and for a fourth starter, that’s really the most you can ask.