A New Metric of High Unimportance: SCRAP
It’s something we hear all the time: “He’s a scrappy player” or “He’s always trying hard out there, I love his scrappiness.” Maybe chicks don’t dig the long ball anymore; maybe they’re into scrappiness. I’m not really in a position to accurately comment on what chicks dig though, so I don’t know.
Even from a guy’s perspective, scrappiness is great. It’s hard to hate guys that overcome their slim frames by just out-efforting everyone else and getting to the big leagues. It’s not easy to quantify scrappiness, though. Through the years it’s always been a quality that you know when you see, but there’s never been a number to back it up. Until now.
Scrap is a metric that is scaled on a similar scale to Spd, where 5 is average and anything above that is above average, and anything below 5 is below average. Here are the components that make it up (each component is factored onto a Spd-like scale, assigned a weight, and then combined with all of the other components to give a final number).
- Infield hit% — Higher is better.
- .ISO — Less power means more scrappiness.
- Spd –The ability to change a game with legs.
- balls in play% — (PA-BB-K)/PA — Go up there looking to fight.
- zSwing%. — Higher is better. Measures willingness to defend the zone.
- oSwing%. — Lower is better. These guys can’t hit the low and away pitch to deep center.
- zContact%. — Higher is better. These guys swing for contact.
Without further ado, here are the Scrap rankings of all qualified batters in 2013.
# | Name | Scrap |
---|---|---|
1 | Alcides Escobar | 6.31 |
2 | Eric Young | 6.27 |
3 | Leonys Martin | 6.25 |
4 | Jacoby Ellsbury | 6.24 |
5 | Starling Marte | 6.23 |
6 | Jean Segura | 6.19 |
7 | Ichiro Suzuki | 6.13 |
8 | Alexei Ramirez | 6.13 |
9 | Elvis Andrus | 6.08 |
10 | Denard Span | 6.08 |
11 | Jose Altuve | 6.08 |
12 | Erick Aybar | 5.93 |
13 | Adeiny Hechavarria | 5.9 |
14 | Daniel Murphy | 5.9 |
15 | Brett Gardner | 5.89 |
16 | Carlos Gomez | 5.89 |
17 | Gregor Blanco | 5.87 |
18 | Michael Bourn | 5.8 |
19 | Alex Rios | 5.76 |
20 | Will Venable | 5.72 |
21 | Norichika Aoki | 5.7 |
22 | Jimmy Rollins | 5.64 |
23 | Shane Victorino | 5.63 |
24 | Michael Brantley | 5.63 |
25 | Howie Kendrick | 5.63 |
26 | Gerardo Parra | 5.61 |
27 | Nate McLouth | 5.58 |
28 | Nolan Arenado | 5.54 |
29 | Torii Hunter | 5.53 |
30 | Austin Jackson | 5.53 |
31 | Chris Denorfia | 5.52 |
32 | Jon Jay | 5.52 |
33 | Brandon Phillips | 5.5 |
34 | Alejandro De Aza | 5.48 |
35 | Dustin Pedroia | 5.45 |
36 | Darwin Barney | 5.45 |
37 | Ian Desmond | 5.42 |
38 | Starlin Castro | 5.42 |
39 | A.J. Pierzynski | 5.4 |
40 | Eric Hosmer | 5.39 |
41 | Asdrubal Cabrera | 5.39 |
42 | Josh Hamilton | 5.39 |
43 | Alex Gordon | 5.39 |
44 | Adam Jones | 5.38 |
45 | Coco Crisp | 5.35 |
46 | Andrew McCutchen | 5.34 |
47 | Marco Scutaro | 5.34 |
48 | Ian Kinsler | 5.33 |
49 | Andrelton Simmons | 5.33 |
50 | Desmond Jennings | 5.32 |
51 | Jonathan Lucroy | 5.32 |
52 | Chase Utley | 5.3 |
53 | Brandon Belt | 5.3 |
54 | Hunter Pence | 5.26 |
55 | Jason Kipnis | 5.22 |
56 | Ben Zobrist | 5.21 |
57 | Alfonso Soriano | 5.2 |
58 | Pablo Sandoval | 5.19 |
59 | Manny Machado | 5.18 |
60 | Brian Dozier | 5.18 |
61 | Matt Holliday | 5.17 |
62 | Brandon Crawford | 5.17 |
63 | Allen Craig | 5.15 |
64 | Matt Carpenter | 5.14 |
65 | Michael Young | 5.13 |
66 | Yunel Escobar | 5.12 |
67 | Yoenis Cespedes | 5.11 |
68 | Yadier Molina | 5.11 |
69 | Nick Markakis | 5.11 |
70 | Zack Cozart | 5.1 |
71 | Mike Trout | 5.1 |
72 | Nate Schierholtz | 5.08 |
73 | Todd Frazier | 5.07 |
74 | Michael Cuddyer | 5.07 |
75 | Domonic Brown | 5.06 |
76 | Chase Headley | 5.03 |
77 | Salvador Perez | 5.03 |
78 | Marlon Byrd | 5.02 |
79 | James Loney | 5.0 |
80 | Neil Walker | 5.0 |
81 | Kyle Seager | 4.97 |
82 | Andre Ethier | 4.97 |
83 | Freddie Freeman | 4.96 |
84 | Mike Moustakas | 4.95 |
85 | Robinson Cano | 4.95 |
86 | Jed Lowrie | 4.95 |
87 | David Freese | 4.92 |
88 | Shin-Soo Choo | 4.91 |
89 | Adam LaRoche | 4.91 |
90 | Chris Johnson | 4.88 |
91 | Martin Prado | 4.87 |
92 | Carlos Beltran | 4.86 |
93 | Ryan Zimmerman | 4.85 |
94 | Victor Martinez | 4.83 |
95 | Justin Morneau | 4.81 |
96 | Adrian Gonzalez | 4.8 |
97 | Anthony Rizzo | 4.79 |
98 | Alberto Callaspo | 4.79 |
99 | Trevor Plouffe | 4.79 |
100 | Ryan Doumit | 4.77 |
101 | Brandon Moss | 4.74 |
102 | Mark Trumbo | 4.74 |
103 | Matt Wieters | 4.7 |
104 | Josh Donaldson | 4.69 |
105 | Adrian Beltre | 4.69 |
106 | Justin Upton | 4.68 |
107 | Daniel Nava | 4.67 |
108 | Paul Konerko | 4.65 |
109 | Billy Butler | 4.65 |
110 | Matt Dominguez | 4.64 |
111 | Jayson Werth | 4.62 |
112 | Russell Martin | 4.62 |
113 | Jay Bruce | 4.62 |
114 | J.J. Hardy | 4.6 |
115 | Joey Votto | 4.59 |
116 | Buster Posey | 4.59 |
117 | Dan Uggla | 4.57 |
118 | Nick Swisher | 4.55 |
119 | Kendrys Morales | 4.52 |
120 | Carlos Santana | 4.51 |
121 | Pedro Alvarez | 4.49 |
122 | Mark Reynolds | 4.48 |
123 | Jedd Gyorko | 4.48 |
124 | Paul Goldschmidt | 4.47 |
125 | Prince Fielder | 4.47 |
126 | Edwin Encarnacion | 4.45 |
127 | David Ortiz | 4.45 |
128 | Adam Lind | 4.4 |
129 | Jose Bautista | 4.38 |
130 | Justin Smoak | 4.37 |
131 | Miguel Cabrera | 4.37 |
132 | Mitch Moreland | 4.36 |
133 | Joe Mauer | 4.34 |
134 | Evan Longoria | 4.24 |
135 | Chris Carter | 4.23 |
136 | Giancarlo Stanton | 4.1 |
137 | Mike Napoli | 4.09 |
138 | Troy Tulowitzki | 4.07 |
139 | Chris Davis | 3.94 |
140 | Adam Dunn | 3.81 |
That’s quite a bit to look at. Here are a few of my takeaways:
- The general perception of a player’s scrappiness is pretty close to what this metric spits out.
- There are some surprises, such as Tulo being near the bottom. In his case it’s caused by an extremely low speed rating and a low z-swing%.
- Little dudes that run hard tend to be scrappy (duh).
- Big oafy power guys tend not to be scrappy (duh).
- Upon removing the qualified batter restriction the ‘Scrap’ leader is Hernan Perez. Tony Campana is a close second. I think we can all agree that Campana is more or less the definition of scrappiness.
This isn’t a stat that’s going to forever change how we view baseball. But this does give us a way of quantifying, however imperfectly, a skillset that we haven’t been able to before. Now we not only know that Jose Altuve is scrappy, we know just how scrappy he is. I’ll let you decide how important that is.
If you have any suggestions regarding different ways to calculate Scrap let me know in the comments. It’s a metric that requires a good amount of arbitrary significance since, well, what does it even mean to be scrappy? We’ve always had an idea, and now we have a number.
The idea for this metric was spurned on by Dan Syzmborksi on this episode of the CACast podcast, somewhere around the 75-minute mark.
Brandon Reppert is a computer "scientist" who finds talking about himself in the third-person peculiar.
Yet another thing that Neil Walker is extremely average at.
This is my favorite comment.
Ha, love it. How about including position? Up-the-middle guys are more scrappy, right? I mean, 1B’s and DH’s certainly aren’t. That would help Tulo…
I thought about this, but it should show up in the other numbers if the player is indeed scrappy. Tulowitzki has a negative career number for BsR, which is odd for a guy that plays the shortstop position. It seems that if anything Tulo should get punished for being an athlete capable of playing up the middle but not wreaking havoc on the base-paths with scrappiness.
Of course, a decrease in scrappiness could also mean a decrease in injury opportunities, which would be good for him. It’s hard to get hurt if you’re not playing hard.
Next up, measuring a pitchers’ craftiness. I think the measures can be low velocity, high ERA, a decent amount of wins, and a big plus if left handed.
Good idea, I may well do this. Big pluses for lots of change-ups, etc. Measuring the mostly useless perception is fun.
My money’s on Cole Hamels
Reminds me of this community article from a little while ago:
http://www.fangraphs.com/community/an-introduction-to-grit/
Interesting, it’s a similar concept. The difference between what GRIT entails and what SCRAP entails is interesting in and of itself.
Word-based communication systems are weird.
Height and weight need to be factored into this somehow. As an extreme example, Jose Altuve is only 1 Altuve tall – he could slug .600 and he’d still be scrappy, just on principle. If anything, it would make him seem scrappier if he somehow put up elite power numbers with that body.
Agreed, maybe I could use BMI as one component and then just straight-up height as another. Or I could make up some sort of metric that measures .ISO relative to height/BMI. Heck, that may be a good idea for an article in and of itself: Little Sluggers™
Brett Gardner!
Isn’t good defense also a sign of being scrappy?
To some extent, but not all parts of defense are scrappy. For example I would consider a guy who runs really hard out in the outfield to be scrappy even if he takes terrible routes to the ball.
The closest thing we have to a scrappy defensive metric would be range runs, but that incorporates a lot of being “smart” defensively with positioning and route running. I decided not to include it based on this reasoning, but it’s not like I was confident in that decision.