Evaluating Closers- Part 2

Tuesday, January 30 2007 @ 07:00 AM EST

Contributed by: Mike Green

In Part 1 last week, I looked at factors that are relevant for the comparison of closers to starters. This week, we'll look at the existing Hall of Fame closers using the adjustments we looked at in Part 1.

There are four pitchers, known at least primarily for their relieving, in the Hall of Fame: Hoyt Wilhelm, Rollie Fingers, Bruce Sutter and Dennis Eckersley. Sutter did not make a start in his career, and Eckersley had half a career as a starter. We have leverage figures for Wilhelm beginning in 1960, which covers most of his relieving career, and complete leverage figures courtesy of Baseball Prospectus for the other relievers. We are using 80% of leverage to account for the effects of chaining., and assuming leverage of 1.00 for starting innings.

Here are their statistics, broken down into starting and relief roles. I have made the adjustments to the relief lines described in Part 1.

Pitcher IP W W/9 W Ind. K K/9 K Ind. HR HR/9 HR Ind ERA+ LEV.
Wilhelm-starting 383.3 115 2.70 130 250 5.87 111 25 0.59 153 140 1.00
Wilhelm-relief 1871 663 3.18 110 1360 6.54 123 125 0.60 150 147 1.5
Wilhelm-relief adj. 1871 729 3.50 100 1224 5.89 111 138 0.66 136 133 1.2

Fingers-starting 195.7 63 2.90 114
116 5.33 101
26 1.20 63
82 1.00
Fingers-relief 1505.7 429 2.56 129
1183 7.07 133
97 0.58 129
127 1.61
Fingers-relief adj. 1505.7 472 2.82 117
1065 6.36 120
107 0.64 117
114 1.28


Eckersley-starting 2478.3 612 2.22 151
1609 5.84 123
268 0.97 82
109 1.00
Eckersley-relief 807.3 126 1.40 250
792 8.83 152
79 0.88 102
143 1.64
Eckersley-relief adj. 807.3 139 1.54 225
713 7.95 137
87 0.97 93
129 1.31

The walk, strikeout and HR indices use the pitcher's rates compared with league averages at the time they pitched. The league averages can be found here.

Let's take a litttle closer look at each of the pitchers.

Hoyt Wilhelm

Knuckleballers tend to give up fewer hits on balls in play than other pitchers in part (it seems likely) because of higher pop-up rates. Wilhelm's ERA+ is a little more impressive than his peripheral statistics, but this is typical for knuckleballers. His ERA+ (and adjusted ERA+ for his time as a reliever) is a fair marker of his effectiveness. Can we find a comparable starter to him, taking into innings/leverage and effectiveness? We'll multiply his innings as a reliever by his leverage and add it to starting innings. That amounts to 2628 innings. His composite starter ERA+ and adjusted relief ERA+ is 134.

There are quite a few comparable starters to Wilhelm, and they are in the Hall of Fame. Short careers and very effective- Rube Waddell, Sandy Koufax, and Hal Newhouser. Koufax and Newhouser, of course, had tremendous peaks which differentiate them from Wilhelm, but Waddell was to my mind a pretty comparable pitcher both in quantity and quality, albeit that Waddell's innings were compressed into many fewer seasons. In any event, there were no starters, who threw as much as Wilhelm, and with about the same effectiveness who are not in the Hall of Fame. The closest is probably Harry Brecheen.

Rollie Fingers

Fingers is to relief pitchers what Joe Rudi is to left-fielders. A very good player in a number of phases of the game and a key contributor to a great ballclub (the Moustache Gang), but not one of the greats of the game. Fingers had consistent numbers across the board once he was converted to relief, as reflected in his adjusted line, but they are just not that unusual. To make the most charitable case for him, one would discard his innings as a starter. He then emerges as a reliever who threw 1,500 innings with an adjusted leverage of 1.28 (for an adjusted figure of 1925 innings) and an ERA+ of 114 with peripherals to match. There are numerous starters who threw more innings with more effectiveness- Dave Stieb, Brecheen, Virgil Trucks, Kevin Appier, Chuck Finley, Billy Pierce, Luis Tiant, and many others.

Bruce Sutter

The evidence that Bruce Sutter was a better pitcher than Dan Quisenberry does not survive. Quiz threw precisely 1 more inning than Sutter in his career, and this makes comparison easy. They were different pitchers. Sutter struck out many more, but Quiz walked many fewer and gave up many fewer homers. Winning two of the three true outcome contests was enough to give Quiz a 12 point lead in ERA+ and push him into Wilhelm territory. Anyways, I think Quiz falls short anyway, but Sutter was a very poor selection, one of the worst ever made by the writers.

Even if he is given the benefit of all of his leverage (disregarding the effect of chaining), he accounts for 1875 adjusted innings and an ERA+ of 122. He pretty clearly wasn't as valuable as Stieb, Brecheen, Appier, Pierce and Tiant.

Dennis Eckersley

Eckersley was a very good pitcher as a starter, and then turned it up a notch in all departments when he was made a closer and continued into his 40s. For his career, he ends up with total adjusted innings of 3535, with composite adjusted ERA+ of 115, a smidge better than Luis Tiant in both departments. There are no starters that I can find who are ahead of him in both departments who are not in the Hall of Fame. Through age 40, Tommy John had thrown 3900 innings with an ERA+ of 115, but hung on for a number of years afterwards, and Rick Reuschel, like Tiant, threw a comparable number of innings to Eckersley with a little less effectiveness.

A word about peak performance and closers

One of the difficulties about measuring closer performance results from the relatively few number of innings pitched. It is difficult enough to evaluate over a career, but over a season of 60-120 innings, the problems become enormous. Strength of opposition, home/road or day/night splits, fielding support, and scoring decisions, can affect the superficial statistics for a season to a great degree. For instance, look at Dennis Eckersley's great 1990 season. His ERA+ for 1988-1992 read 160, 237, 606, 130 and 196. By any objective measure, Eckersley did pitch much better in 1990 than in any other season, but not by that much. The key seems to be that almost 1/2 his runs allowed in 1990 were unearned.

To take another example, look at the ERA+ (or if you prefer RA+) of Billy Wagner over 2001-2006- 166, 170, 248, 180, 300, 193. He has actually been a remarkably consistent pitcher over those years. It seems to me that one gets a much more accurate picture of the effectiveness of a closer by looking at a career than a peak.

In closing...

I would like to thank the Academy...nevermind. The choices made by the writers among the possible Hall of Fame closers provide little assistance to us. Fingers was a poor selection, while Sutter was really quite strange. In fairness, the change in reliever usage to the modern closer, and the save rule, made evaluation difficult.

As for the current crop of closers, the evaluation is even harder. They're throwing less, but with more effectiveness than their predecessors of the 70s and 80s. Here is Billy Wagner's age 34 comparable BBRef comparable list, to give a flavour. Next time, we will take a closer look at Goose Gossage, before moving on to Mariano Rivera, Trevor Hoffman and Wagner.