One of the concerns about calling up a 20 year old Jordan Lyles to the big league club had to do with workload risk. The connection between pitcher workload and liklihood of injury is most clear for pitchers younger than 23. So, as the Astros head into the second half of the season, let's look at the implications for the remainder of 2011. I should give a hat tip to Chron.com's Chip Bailey for raising the question of when Lyles will be shut down on innings pitched limitations. The Astros plan to limit the Lyles' innings this season, but haven't disclosed the specific criteria.
So far this season, Lyles has pitched 59 innings in AAA and 47 innings in the majors for a total of 106 innings. Chip Bailey suggests that 170 innings may be close to the innings limitation we will see from the Astros. I think this is a reasonable guess. We know that the Astros shut down Bud Norris after 167 innings (AAA and ML combined) in his first big league season. There are few recent examples of 20 year old starting pitchers, and it may well be that Lyles' younger age should result in even fewer innings. The most recent example of a 20 year old rookie starting pitcher was Clayton Kershaw, and he was allowed to pitch 168 innings between the majors and AAA in 2008. That seems to be another point to validate the reasonableness of a workload in the 170 inning range.
Also, some teams rely upon concepts similar to the Verducci Effect, which posits that pitchers face greater risks if their workload increases by more than 30 innings over the previous year. Lyles pitched 158 innings between AA and AAA in 2010. A limitation in the 170 innings range would place Lyles comfortably below the threshold for the Verducci Effect. Therefore, a limit within that range would allow Lyles to increase his year-to-year workload in small increments.
Based on Lyles' 5.9 innings per game average so far, a 170 inning limit would give 10 - 11 more starts to Lyles this year. That could result in shutting down Lyles' work for this season sometime near the end of August. And that raises some interesting implications for the composition of the Astros rotation at the end of the season, particularly if Wandy or Myers were to be traded.
If no trades occur, presumably Aneury Ramirez could step into Lyles' rotation spot. However, if both Wandy and Myers were to be traded, the Astros will have difficulty replacing three empty slots in the rotation. Neither Nelson Figueroa or Ryan Rowland-Smith are on the 40 man roster, and even if they are added to the roster, their respective ERAs are 5.61 and 6.23 in AAA---which doesn't bode well for a call up. Frankly, the Astros' starting pitching depth at AA and AAA is exceptionally weak. Alternatively, the current depth situation could lead the Astros to ask for major league ready pitchers in any Wandy or Myers trade. This happened when the Astros asked the Phillies for Happ in the Oswalt trade, and I could see a similar type of request this time.
Going beyond the innings limitation question, I will briefly discuss other aspects of Lyles' workload in 2011 after the jump.
Research in this field indicates that damage from high pitcher workload is related to pitcher age, and that the connection between catastrophic injury and overuse seems to drop off significantly after age 25. The noted book Diamond Appraised by Tom House and Craig Wright studied pitcher work load impacts and recommended guidelines for the usage of young pitchers. The guidance from this 1989 book appears to have been effective in persuading baseball teams to reduce the workloads for their young pitchers. Young starting pitchers in the era prior to 1990 routinely exceeded the guidelines associated with that study, while there are relatively few examples in the most recent period. The book uses total batters faced per game and pitch counts as measures for evaluating starting pitcher workload.
Based on that source, 25 batters faced per game and 100 pitches per game are reasonable targets for younger pitchers. Over any reasonable period of consecutive starts during the season, teams should not allow pitchers younger than 23 to face an average as high as 128 - 130 batters per game or throw more than 105 pitches average per game. Given that these guidelines are more than 20 years old, these limitations may be on the high side.
So far this season, Lyles has thrown an average of 103 pitches per game in the majors. That doesn't appear to be unreasonable. However, individual games may raise a bit more concern. For instance, on June 11, Lyles threw 111 pitches. And, perhaps of more interest, Lyles has been allowed to exceed 100 pitches on a consistent basis in more recent games, with pitch counts of 109, 101, and 108 in the most recent three games. I wouldn't necessarily consider this a red flag---that is, unless it represents a pattern of intentionally increasing pitch counts over the course of the season. However, I think this may be random occurrences rather than a trend. I think the pitch counts have creeped up due to circumstances of the games rather than intentional decisions to increase the counts. Although strictly capping pitch counts is not a perfect approach, perhaps the Astros should consider a strict 100 pitch limit for Lyles this season, since it would prevent pitch counts from creeping upward due to game circumstances.
Below I have shown the batters faced per game by Lyles at various levels.
Total Batters Faced Per Game
Lexington-A (age 18) 23.15
2010 AA/AAA (age 19) 25.7
2011 AAA (age 20) 20.7
2011 ML (age 20) 25.9
Total batters faced reflects more of the context of the pitching situations than simple innings pitched. As more runners are put on base, the total batters faced will increase too. Lyles' batters faced per game at the major league level probably are higher than we would prefer. But they don't approach an unreasonable level (28 - 30). For comparison, a 20 year old Clayton Kershaw faced 21.4 batters per game in his rookie season.
Of course, pitch counts and batters faced per game aren't the whole story. Other factors are relevant, such as the stressfulness of pitching situations, types of pitches thrown, etc. The fact that Lyles' delivery exhibits clean mechanics is a positive. The impact of the types of pitches he throws is less clear. Some analysts have theorized that young pitchers who throw a lot of breaking pitches will face greater risks from high pitch counts. (Note that this is supposition--not a proven fact.)
Lyles throws a lower percentage of fastballs (49% compared to the major league median 56%) than the average starting pitcher. So, Lyles does throw more breaking pitches than the typical starter. But Lyles does not throw an above average percentage of sliders, which are the breaking pitches that are hardest on the arm.
Overall, I don't see many red flags for Lyles' usage so far this season. But the pattern of his usage in the future months bears watching. If Lyles continually exceeds 100 pitches in every game or if the pitch counts trend upward, then we should wonder if the Astros are tempting fate. Depending on the circumstances, I probably would be concerned if Lyles is allowed to continue pitching substantially beyond 170 innings or so.