October 20, 1998

Pitcher Usage and Result Patterns: Toronto Blue Jays

Focus on 1998 Blue Jays starting pitching

by Jeff Bower and Christina Kahrl

Pitcher Usage and Result Patterns? What are these, you ask? This is the end result of a conversation the two of us started up in the spring on how to evaluate starting pitchers in terms of how they're used by their managers, and what that means in terms of results. We wanted to generate a team-wide statistical snapshot on how starting pitchers performed based on how they were used by their managers. In doing this, we used the quality start as our qualitative measure of pitcher performance, and we recorded them against rest and usage patterns. What comes out on the end? Starters (and teams) are evaluated in terms of how many games they've started on 3, 4, 5, or 6+ days' rest. Among each of those sets of information, we see how many quality starts (QS) and blown quality starts (BQS) the starter recorded, and what their average pitch counts were. What's a blown quality start? A game in which a starter has already logged a quality start (six or more innings, three or fewer

	1998 Overall Pitcher Use Patterns Days rest -> 2 3 4 5 6+ CS Totals Starts 0 1 93 47 21 1 163 QS 0 0 53 22 12 1 88 %QS .00 .00 .57 .47 .57 1.00 .54 BQS 0 0 5 4 2 0 11 %QS+BQS .00 .00 .62 .55 .67 1.00 .61 Avg # pitches 0 105 109 105 107 107 108

runs) that either he blew by allowing a fourth run (or more) after the sixth, or that his bullpen blew by allowing baserunners that the starter left to them to score.

What can it tell us? It gives us a way to think about how a manager runs his rotation in terms of pitch counts and how regularly the starter pitches. As we work our way through the `98 season, we'll be able to come to some conclusions about some pitchers and some managers. How often does a manager have his starters take their regular turn? Is there a relationship between pitch counts and longer rest periods? How often did the starter give his manager a quality start? Are some starters more successful with longer rest periods than four days? Rooting through this information should begin to give us the answers to these questions.

Clemens            Days rest                 Hentgen            Days rest
                3    4    5   6+   Totals                    3    4    5   6+  Totals
Starts          0   22    8    3       33    Starts          0   15   10    4      29
QS              0   16    4    3       23    QS              0    6    3    3      12
%QS           .00  .73  .50 1.00      .70    %QS           .00  .40  .30  .75     .56
BQS             0    2    1    0        3    BQS             0    1    0    0       1
%QS+BQS       .00  .82  .63 1.00      .79    %QS+BQS       .00  .47  .30  .75     .45
Avg # pitches   0  117  107  123      115    Avg # pitches   0  104   96   96     100

Williams           Days rest                 Guzman             Days rest
                3    4    5   6+   Totals                    3    4    5   6+  Totals
Starts          0   17   12    3       32    Starts          0   13    8    1      22
QS              0    9    7    1       17    QS              0    9    3    0      12
%QS           .00  .53  .58  .33      .53    %QS           .00  .69  .38  .00     .55
BQS             0    0    2    1        3    BQS             0    0    1    0       1
%QS+BQS       .00  .53  .75  .67      .63    %QS+BQS       .00  .69  .50  .00     .59
Avg # pitches   0  111  110  112      111    Avg # pitches   0  109  110  124     110

Carpenter          Days rest                 Escobar            Days rest
                3    4    5   6+   Totals                   OS    4    5   6+  Totals
Starts          0   13    6    5       24    Starts          1    6    2    1      10
QS              0    7    3    1       11    QS              1    5    1    1       8
%QS           .00  .54  .50  .20      .46    %QS          1.00  .83  .50 1.00     .80
BQS             0    2    0    1        3    BQS             0    0    0    0       0
%QS+BQS       .00  .69  .50  .40      .58    %QS+BQS      1.00  .83  .50 1.00     .80
Avg # pitches   0  100  104  104      102    Avg # pitches 107  118  107  112     114

Hanson             Days rest                 Stieb              Days rest
                3    4    5   6+   Totals                    3    4    5   6+  Totals
Starts          0    5    1    2        8    Starts          1    2    0    0       3
QS              0    1    1    2        4    QS              0    0    0    0       0
%QS           .00  .20 1.00 1.00      .50    %QS           .00  .00  .00  .00     .00
BQS             0    0    0    0        0    BQS             0    0    0    0       0
%QS+BQS       .00  .20 1.00 1.00      .50    %QS+BQS       .00  .00  .00  .00     .00
Avg # pitches   0  104   96  110      105    Avg # pitches 105   92    0    0      96


Halladay           Days rest
                3    4    5   6+   Totals
Starts          0    0    0    2        2
QS              0    0    0    1        1
%QS           .00  .00  .00  .50      .50
BQS             0    0    0    0        0
%QS+BQS       .00  .00  .00  .50      .50
Avg # pitches   0    0    0   94       94

Tim Johnson arrived in Toronto with a reputation as a players' manager, particularly young players. It was expected that this would be a nice change from Cito Gaston, who by the time he left had established that he was apathetic about working with anybody in any age group. While it took Johnson about four months before he and the team committed themselves to their younger players (making themselves noticeably better in the bargain), the way he handled his starting pitching was established within the first three weeks of his tenure. Johnson works his starters relatively hard, averaging 108 pitches per start. He sticks with a five-man rotation, not skipping the fifth starter even in the event of an off day. In this way, all of his starters end up benefitting from the extra day of rest. Toronto's starters had eleven BQS for the season. An educated guess based on limited data projects that this figure will be around the league average. On that score, Johnson and his staff aren't falling asleep as the game progresses - something that Cito was frequently accused of.

First, let's look at the handling of the veteran pitchers and how they responded. Roger Clemens could probably throw a sixteen-pound shot 125 times a game every fifth day, and still finish the season with an ERA under 3.00. Johnson's usage pattern obviously was no problem for The Rocket. However, it was for Pat Hentgen. Hentgen, whom Gaston had worked harder than any starter in baseball during the previous two seasons, couldn't make it through a third. His health was the subject of rumors all summer before being shut down in September, and he only put up ten quality starts in his 25 starts with 4 or 5 days' rest, while logging three QS in his four starts on 6 days' rest. Woody Williams has a history of arm problems, but he lasted the entire season despite an average workout of 110 pitches. The injuries may have left him ill-suited to take his turn regularly and that he could benefit from a little extra rest. Williams tossed 9 QS+BQS out of 12 starts with 5 days' rest, while while struggling to put up a quality start half the time with a normal four days of rest. Juan Guzman threw four months of very good ball under Johnson's demanding workload before moving on to the Orioles. Erik Hanson was recovering from a lost season, and the Jays found out that he hadn't.

Was Johnson as demanding with his younger pitchers? Chris Carpenter started the year in the bullpen, and was pampered a little more than the other regular starters, averaging 102 pitches per start. That lighter workload paid off when the Jays mounted their late push for the wild card, as he gave the Jays with five consecutive quality starts in September. Kelvim Escobar was brought up from Syracuse in early August and inserted into the rotation. He immediately reeled off seven straight quality starts, but Johnson stretched him out to an average pitch count of 120 during that span. That takes a pretty heavy toll on any pitcher, but at 22, Escobar was especially at risk. Escobar finally collapsed at the same time as the Jays did, producing only one quality start in his last three outings. A note about Escobar: the "OS" column is for his original start, since we couldn't dig up the date for his last start at Syracuse to determine how many days' rest he had between that and his first start in the majors.

Conclusions? The workloads Johnson handed out didn't matter much to Roger Clemens, and it looks like his handling of Chris Carpenter was beneficial. Using Woody Williams on longer rest may generate better results for Williams, but it creates a new set of problems in terms of everyone else's rest patterns. Whether or not anything could be done to get good work out of Pat Hentgen or Erik Hanson is probably academic, given that they were both damaged goods. Kelvim Escobar was ridden hard and floundered after initial success, so it will be interesting to see how Johnson handles him (and Roy Halladay) over next year's 162 game season.

Christina Kahrl is an author of Baseball Prospectus. 
Click here to see Christina's other articles. You can contact Christina by clicking here