"Total Run Accounting" - 2008 Player by Player
As many of you have seen, last week I posted team by team results of my ongoing "Total Run Accounting" (TRA) analysis here. I won't repeat much of the method or team findings, but one of my primary findings was that the "little things", directional hitting and baserunning, have a real impact on runs scored, and the Twins are much better (over 40 runs above MLB average) at doing the "little things".
Those were team totals. How does TRA rate individual players? I've posted a spreadsheet with 2008 TRA data for all 1,038 MLB players with at least 1 plate appearance last season:
What did I find, a number of interesting details across MLB and for the Twins in particular.
The single best player in MLB at doing the "little things" is...Joe Mauer. Compared to the average MLB player, Mauer helped himself to the tune of +12.86 runs last year. Randy Winn, Matt Holliday and Ichiro Suzuki rounded out the top 4, each over 2 runs behind Mauer. Joey Votto, J.J. Hardy, Pat Burrell and Hank Blalock were the bottom 4, about 20 runs behind Mauer.
Interesting, Justin Morneau was 8th in MLB at +8.69 runs compared to average...
Looking at the player spreadsheet data field by data field we find:
Standard Batting
Compared to average, "Standard Batting" ER ranges from Albert Pujols (+68.68) to Jeff Francoeur (-38.18). This means that, due to hitting stats that directly contribute to OPS, there is over a 100 run (10 win) difference between the best and worst in baseball. That's a surprisingly large difference. Replace Pujols with Francoeur in the St. Lous lineup and the Cardinals could very well lose 10 wins.
The remainder of the top five does not surprise me. Lance Berkman, Grady Sizemore, Manny Ramirez (LA only) and Chipper Jones round out the top five. Pujols is head and shoulders above the rest, over 12 runs ahead of Lance Berkman.
The bottom five is interesting, Yuniesky Betancourt, Khalil Greene, Bobby Crosby and Robinson Cano just ahead of Francoeur. Cano surprises me, for someone many consider to be a batting champion in the near future. However, all four are over 10 runs ahead of Francoeur, that's how bad a season Jeff had in 2008.
For the Twins, Morneau (+26.50) and Mauer (+26.25) were (not surprisingly) the top two on the team. Denard Span (+22.99) was within four runs of each (RAA) and ahead of both on an absolute ER basis. Why? Chances are, Morneau and Mauer were hurt by 20 and 21 GIDP respectively.
Brendan Harris (-9.79) was worst on the team, with Delmon Young, Mike Lamb, Adam Everett and Alexi Casilla not far behind, none better than -8.92. In fact, five of our projected 2009 starters (Gomez and Punto, in addition to the above) were below average in "Standard Batting".
Across MLB, I was more surprised by the players at the bottom of the list, including Miguel Tejada (-13.57), Jason Bartlett (-14.96), Garrett Atkins (-8.91) all well below MLB average. For Tejada, his 32 GIDP and low walk rate likely hurt him for TRA.
Other Batting
On an absolute ER basis, the Twins were very well represented, with Justin Morneau (+24.20) tops in MLB and Joe Mauer (+22.12) at #3 overall. Compared to MLB average, Bengie Molina (+8.20) is #1 with Morneau and Mauer at #2 and #3 respectively.
"Other" Batting RAA ranged from Molina (+8.20) to Hanley Ramirez (-7.85). This tells me that across MLB, the total impact of directional hitting is up to around 16 runs. Pretty significant for a single player.
I find it surprising that so many right handed batters (half of the top 10) are tops in "Other" batting. I expected to see a heavy dose of left handed hitters, able to hit more easily to the right side, at the top of the list.
The bottom of the list had two all stars. Hanley Ramirez (-7.85) and Jose Reyes (-6.89) as the two worst in MLB, relative to average. Grady Sizemore (7th worst), Adam Dunn and Alfonso Soriano were all in the bottom 20.
As noted in my other post, the Twins were very good in this area. After Mauer and Morneau, Buscher, Lamb, Harris, Everett, Young, Casilla and Kubel all helped themselves by at least 2 runs in RAA. Only Cuddyer, Punto, Span and Gomez were below average (among regulars), and the worst was Gomez at -2.28.
Baserunning
Baserunning ER ranged from Willy Taveras (+13.59) down to Dioner Navarro (-8.98), over a 22 run range between MLB best and worst. Again, quite significant for a given player. Most of the top five was unsurprising, Ichiro, Holliday, Victorino and McLouth all at the top. Holliday surprises me a bit, but he is an underrated baserunner.
The bottom is dominated by catchers, with four of the bottom six (Navarro, Molina, Varitek and McCann). Magglio Ordonez and Prince Fielder were also at the bottom. I was surprised to see a few players generally considered good baserunners at the bottom of the list, Cristian Guzman, Julio Lugo and David Wright all around the -4 range.
For the most part, the Twins ranged between +2 and -2 runs on the basepaths. The two exceptions were Matt Tolbert (+4.93) and Joe Mauer (+4.67), who bring the overall team ER above zero. Bottom five were Monroe, Young, Pridie, Kubel and Buscher, with Punto right behind. The fastest baserunner, Carlos Gomez, came in a little above average (+0.52), with a bunch of caught stealing and other outs on the basepaths cancelling out a ton of steals and extra bases.
Opponents Fielding
Going in, I did not expect this value to vary a great deal, as it is mostly out of the batter's hands. However, the top of the list was dominated by speedy contact hitters such as Ichiro, Pedroia, Crawford, etc. I'll dig into this statistic at a later time.
Overall "Little Things"
Adding up the Other Batting and Baserunning ER gives us a total for what I call the "little things". Comparing to MLB average, these values range from Joe Mauer (+12.86) down to Joey Votto (-8.74). Overall, a player can help (or hurt) himself by about 21 runs doing the "little things".
A few players I haven't mentioned at the top of Other Batting or Baserunning appear near the top of this list. Raul Ibanez (#6), Carlos Beltran (#7), Howie Kendrick (#13) and Jason Bartlett (#14) all helped themselves by about 8 runs or more.
The bottom of the list is dominated by slow sluggers like Votto, Hardy, Burrell, Fielder and Dunn. I was surprised to see Iwamura, Chris Young, Hunter Pence and Johnny Damon on the bottom page, none better than -4.30 runs.
Total Offense
The next few sections focus on rolling up all of these offensive contributions into (hopefully) meaningful statistics that can be compared to other measures of value such as Runs Created (the previous versions, not the version I propose below) or wOBA.
Total Offense, on an absolute basis, correlates more with Standard Batting than any of the other ER categories due to the largest overall contribution. Just as we would expect. The top five in MLB in 2008 were Pujols, Berkman, Holliday, Youkilis and Sizemore. The bottom five were Francoeur, Greene, Varitek, Betancourt and Lugo.
Total Offense Run Average (tORA)
Another way to look at Total Offense is to look at offensive runs on a per plate appearance basis. To do this, I have defined a new statistic called "Total Offense Run Average, or tORA.
Calculating tORA is pretty simple, done in three steps.
- Calculate total offense runs above average, per plate appearance. For Albert Pujols, this means +66.03 / 643 = 0.103.
- Add a factor to account for MLB average offensive runs per plate appearance. To do this, I used the MLB team average runs scored (753) and PA (6254) and added a factor 753/6254 = 0.120 to each player's tORA.
- In order to normalize tORA around the same point as wOBA (0.335), I added an additional factor of 0.214 for each player to center around 0.335.
I haven't done a detailed comparison to wOBA, but tORA gives us a good idea of total offense using a rate statistic. Manny Ramirez (.490 in his time with LA) was tops among players with at least 200 PA, followed by Shin-Soo Choo (.443), Pujols (.437), Mark Teixeira (.431 with LA) and Berkman (.421).
Bringing up the rear were Wily Mo Pena (.231), Andruw Jones (.240), Tony Pena (.244), John McDonald (.245) and Julio Lugo (.258).
Below is a table showing all Twins players with at least 200 PA. Players were roughly 50-50 above and below MLB average, with Span, Mauer and Morneau far above the rest. Just about every player's tORA was higher than their wOBA, perhaps reflective of the Twins and "little things", or simply a bias in my calculation of tORA.
|
Player |
tORA |
wOBA |
Total Runs Created |
|
Denard Span |
.397 |
.364 |
75 |
|
Joe Mauer |
.397 |
.378 |
116 |
|
Justin Morneau |
.389 |
.369 |
125 |
|
Brian Buscher |
.356 |
.317 |
35 |
|
Michael Cuddyer |
.350 |
.311 |
38 |
|
Jason Kubel |
.346 |
.345 |
68 |
|
MLB Average |
.335 |
.335 |
|
|
Alexi Casilla |
.331 |
.316 |
51 |
|
Brendan Harris |
.326 |
.318 |
55 |
|
Delmon Young |
.322 |
.324 |
67 |
|
Nick Punto |
.322 |
.324 |
41 |
|
Carlos Gomez |
.318 |
.294 |
64 |
|
Mike Lamb |
.316 |
.254 |
27 |
Total Runs Created
Finally, I can apply tORA to calculate the total number of runs (absolute, not deltas up and down from average, like ER) that a player created. To do this, I simply subtract the .214 "normalization factor" from a player's tORA and multiply by number of plate appearances. Obviously, those batters with more PA will create more runs, but this gives us an idea of the individual contributions from each player over the season.
Again, Pujols, Sizemore, Berkman, Hamilton and McLouth top the list. Morneau tops the Twins with 124.96 runs, followed by Mauer (115.75). In other words, the M&M boys contributed a total of about 240 of the Twins 829 runs last year.
Upcoming Work
I hope you all found this player by player analysis interesting. I plan to delve a bit deeper into many of the categories, as well as compare tORA to wOBA in a bit more detail. I'm still working on defense (pitching and fielding), don't know if I'll get done with defense first or run comparisons for TRA between 2008 and previous years to look for repeatability. Suggestions?
11 comments
|
4 recs |
Do you like this story?
Comments
Once again, awesome
I love this stuff.
"You're thinking too much. Just have fun." -- Bennie "The Jet" Rodriguez in Sandlot
Thanks
This time around it was less time on the software, more time with “spreadsheet engineering”, defining ways to normalize many of the ER values relative to MLB average, as well as coming up with tORA as a rate statistic. Feel free to download the entire spreadsheet and manipulate however you want.
by Adam Peterson on Feb 7, 2009 12:00 PM EST up reply actions
I'm in awe,
because this is the kind of thing I’d love to get into, but simply don’t have the time. Thanks again for putting this together.
I’d like to see some numbers on how batters and pitchers perform on balls outside of the strike zone. You’d think guys like Vlad can still excel by swinging at balls out of the zone, but I’d like to see what happens to “normal” players and, specifically, how the Twins do.
Just another reason to be patient and stop hacking for Delmon and Carlos…
I'll queue it up, very interesting
Should be just a matter of an additional report, since MLB Gameday has this data. My question is how to consider from the batter’s perspective. Do I count every AB that finishes by swinging at a ball outside the zone? Contact rate on balls outside the zone? Etc…
by Adam Peterson on Feb 7, 2009 12:23 PM EST up reply actions
Longer periods?
How easily can this be done over longer periods? I’m thinking both from team and individual player perspectives. As I understand it, this was done based on 2008 stats, and a lot of the results match intuitively to what I would expect. I’m curious to see how consistently people are the best and worst at these things, or if it just bounces around from year to year. I’m also curious to then see whether certain coaches or teams consistently breed good/bad little things, or whether it is the players that make the difference.
As long as the MLB Gameday data is available...
I’ll run the software. It took a while to get the 2008 data right, as I had to deal with each special case, e.g., force out at second followed by a throw to the plate that goes wild and results in an out at home. I’ll get the data for players and teams.
by Adam Peterson on Feb 8, 2009 3:00 PM EST up reply actions
Calling all those people...
…who bashed Span for a few years before early last year. There are lots of em out there, some on this site as well as others. Congratulations Denard Span!
I was probably one of them
how about that turnaround. I hope he can keep it up this year. Maybe Plouffe can do something similar.
by Adam Peterson on Feb 8, 2009 5:17 PM EST up reply actions
I was one of them
But his numbers didn’t support much enthusiasm before last year, except for one half season. Once he got his eyes fixed, he started showing a better eye at the plate and making better contact.
"You're thinking too much. Just have fun." -- Bennie "The Jet" Rodriguez in Sandlot
Great stuff
Again, this is great work. Interesting, too, that it comes out basically as we would expect (fast guys and contact hitters at the top of the list), which gives me much more confidence that it’s a bona fide useful stat – I realize that part of the usefulness of statistics is to challenge conventional wisdom, but it’s much easier to find them believable if they jibe with our perceptions at least a little bit (this is why fielding stats have been met with so much resistance).
The correlation between errors and speedy guys does make quite a bit of sense – I would guess that a fast player is more likely to beat out a mildly-bobbled grounder than a slow guy, and fast players are also more likely to force the fielder to hurry the play, which could make both fielding and throwing errors more likely.
Just curious – have you checked the 2007 and earlier Gameday logs to see how easy it will be for your tooling to parse?
"There are only two things that are infinite, the universe and human stupidity, and I'm not sure about the former." - Albert Einstein
Not yet
I plan to download and check out the 2007 Gameday logs within the next week or two and see if there are any major differences in the data.
by Adam Peterson on Feb 10, 2009 7:55 AM EST up reply actions

by 



















