First Pitch - "Sabermetric Tuesday - The Setup Man"
Good morning and welcome to the Fantasy Week 6 edition of "Sabermetric Tuesdays." I've had a thought in my head for quite some time, but haven't had a chance to accumulate the stats. That is, until today. Just how effective and accurate is the Hold statistic in determining a setup man or middle reliever's value? Digging through some of the more popular league services around the web, somewhere around 10% of all leagues utilize the Hold as a way of incorporating middle-relief into a quantifiable fantasy measure. Yet, the metric remains an enigma to most baseball fans outside of a few fantasy circles.
Before we get into the quantitative analysis, let us revisit the rules of acquiring a Hold. Last year, I wrote a First Pitch about the obscure stat (you can find it here) and listed the Hold leaders for the year. In summary, the stat is defined:
"A Hold is credited any time a relief pitcher enters a game in a Save Situation, records at least one out, and leaves the game never having relinquished the lead."
And, as a refresher, a save situation is defined as a pitcher who satisfies at least one of the following conditions:
1. He enters the game with a lead of no more than three runs and pitches for at least one inning
2. He enters the game, regardless of the count, with the potential tying run either on base, or at bat or on deck
3. He pitches for at least three innings
As I asserted last year, I find it ridiculous that a relief pitcher can enter a game in a blow-out, relinquish a dozen runs, record an 9 outs and earn a Hold. The same holds true for the "Save," as we evidenced last year in the August 22, 2007 Baltimore/Texas game where Wes Littleton earned a save in a 30-3 game. What exactly did Littleton save in that game? Please explain.
This analysis begins with a look at the Holds leaders over the past 3 seasons, in aggregate. I have listed those pitchers and ranked them by the top 50 in cumulative Holds from 2005-2007. These are the guys that consistently get handed the ball in a setup role. Of course, some of them have spent brief stints as a team's closer (when the usual closer either gets hurt or becomes ineffective). But a look at these names on the list shows that they are true middle-relievers who have consistently earned holds over the last 3 years.
Closers have a blown-save category that, when added to saves earned, equates to their total save opportunities. There is no "blown-Hold" stat and therefore, no total Hold Opportunities measure that is easily defined. Therefore, I turn to Inherited Runners Scored (IRS) as a measure of middle-relief performance. IRS is a simple concept = Runners who are already on base when the reliever entered the game later score while that reliever was pitching. Assuming no fielding errors, this run would be attributed to the ERA of the pitcher who allowed the runner to reach base.
IRS is a great number but is rarely used in popular analysis. Afterall, if we can measure power hitters based on their HR/AB rate, closers based on their saves percentage, and starters based on quality starts, we should hold middle relievers accountable for a key part of their job description: the ability to strand inherited runners on base and not allow them to score.
To fully capture the number of runs scored, I added earned runs to IRS to get a blended number of runners allowed to score that are attributable to the reliever's performance. From there, I took the classic ERA formula of (ER * 9 / INN) and turned it into a reliever adjusted run average (RERA) of [(ER + IRS)*9] / INN. I performed this exercise simply to give us a better idea of pitcher performance relative to innings pitched and appearances (although I do include those raw numbers below as well).
And now, the list and the results:
A lot of data, but the chart below does show some linearity (even though it is very gradual). In other words, as we move down the list of 1 to 50, we see, in general, that RERA increases. A scatter chart below shows this linearity visually:
While this analysis just scratches the surface it does tend to show that fantasy owners, in general, can use Holds as a decent measure to determine a pitcher's effectiveness in the middle innings. Even if you don't count Holds and you decide to fill a roster spot or two with middle relievers for WHIP/ERA purposes, Holds can be used as a decent indicator of appearances and pitcher success. An in-depth regression by a true statistician would
need to be performed to truly determine if Holds and RERA are
correlated. Another theory that could be tested would be to see if an increase in RERA by a certain percentage in one season leads to a decrease in holds in subsequent seasons. In other words, allowing inherited runners to score more often could be a leading indicator of a future decline in accumulated holds and middle-relief effectiveness.
If nothing else, hopefully this article has you thinking about the forgotten men of baseball - the middle reliever - and their impact on the game. Perhaps you will decide to go with the hold as a stat for next year's fantasy leagues. Have a great Tuesday. --- Joe
First Pitch - "Sabermetric Tuesday - The Setup Man"