:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/8674429/gyi0060719333.0.jpg)
After looking at batting average and its flaws last week, I decided next to focus on reforming slugging percentage.
Slugging percentage is considered to be one of the most important stats for evaluating a player at the plate, as it supposedly measures the quality of a batter's hits. I find however, that it fails in doing such in the optimal way, but that the idea is certainly valuable.
For reference, here is the equation for slugging "percentage."
SLG=1B+(2x2B)+(3x3B)+(4xHR)/AB
This stat of course has many flaws, which I will kindly list below.
- ABs in the denominator. Let's just stop looking at that stat entirely.
- It weights the different hits incorrectly. We know this obviously, as the work involving wOBA and wRC+ has advanced offensive statistics way past this point. A home run is not 4 times as valuable as a single, a triple is not 3 times as valuable, etc.
- It isn't a percentage.
So I propose an alternative that is a true measure of the quality of hits for a batter. I took the wOBA constants for singles, doubles, triples, and home runs, and put them all together and put hits (H) in the denominator. The stat that came out should be an effective measure of the "quality" of a batter's hits, and do a much more accurate job of measuring it than slugging percentage.
Ideally a leaderboard of hitters in this statistic would be guys who hit lots of extra-base hits, but would not be negatively affected by other aspects of their game. Luckily we have a leaderboard, and can check out the top 10 in this statistic below. I subtracted the singles-weight in the "scaled" column, at the urging of fellow BtBer Matt Hunter, which scales it so that the minimum is 0. The stat is called "wH/H."
| Season | Name | wH/H | wH/H scaled |
| 2010 | Jose Bautista | 1.372311 | 0.482311 |
| 2009 | Carlos Pena | 1.355327 | 0.465327 |
| 2012 | Adam Dunn | 1.353045 | 0.463045 |
| 2002 | Jim Thome | 1.326212 | 0.436212 |
| 2010 | Mark Reynolds | 1.320707 | 0.430707 |
| 2011 | Mark Reynolds | 1.307059 | 0.417059 |
| 2005 | Andruw Jones | 1.301903 | 0.411903 |
| 2004 | Barry Bonds | 1.300504 | 0.410504 |
| 2007 | Carlos Pena | 1.299522 | 0.409522 |
| 2003 | Barry Bonds | 1.298767 | 0.408767 |
That is a pretty good looking leaderboard, for a full one you can go to this Google doc.
For fun, here is a bottom 10 leaderboard.
| 2008 | Chone Figgins | 0.932272 | 0.042272 |
| 2006 | David Eckstein | 0.931425 | 0.041425 |
| 2010 | Juan Pierre | 0.931179 | 0.041179 |
| 2012 | Ben Revere | 0.930793 | 0.040793 |
| 2007 | Reggie Willits | 0.927841 | 0.037841 |
| 2010 | Elvis Andrus | 0.926949 | 0.036949 |
| 2012 | Jamey Carroll | 0.925492 | 0.035492 |
| 2005 | Scott Podsednik | 0.92415 | 0.03415 |
| 2005 | Jason Kendall | 0.920798 | 0.030798 |
| 2006 | Jason Kendall | 0.919761 | 0.029761 |
As much as I like the results of this stat, I still find it concerning that the denominators are differen between wH/H and H/PA. I do not mind this however, because the goal for this stat was not to go side-by-side with H/PA, but rather to show what kind of hits the hitter accumulated, as opposed to how many.
Now that I have revealed my first two changes to the triple-slash line, I will reveal my final change to it next week.
What do you guys think? Better? Worse?
