Offensive performance, wRC+ and #PowerBananas: why Miguel Sano is the best prospect in the game

The art of evaluating performance inside a baseball diamond has undoubtedly changed the last few years with the infusion of science (math & statistic notations.)   In the "good old days" if someone "hit 300", with more than 30 home runs and either more than 100 runs scored or 100 "ribies", he had a great season.  That was the measuring stick that separated great from very good.  And it still does, mainly on the mainstream press, game broadcasts and talk radio.  The expressions "he hits two seventy five" and "he cannot hit his weight" will be there as performance measures, albeit as crude as an Amish scooter is for propelling devise when compared to a Tesla .  Still, even that is light years ahead than the Fred Flintstone mobile of a hitter's performance evaluation, the "eyeballing" method, which introduces measures such as "the special sound a ball makes when leaves the bat" or "the quickness of someone's wrists" or the way "someone is flying on the bases" or something.

There has been evolution.  OPS  as a concept was introduced in 1984, and then was refined to OPS+ or adjusted OPS, that normalizes for park and league effects, plus adds an easy to compare baseline of 100 that is the league average that season, for an easy reference.   So an OPS+ higher than 100 is better than average and less than 100 worse.  Much improved from "ribies" and other cumulative stats, but still one dimensional Gremlin-like.  What OPS lacks, is that it disregards the ability of someone to sacrifice runs, steal bases, score runs, avoid hitting into double plays, and all the other goodies that are happening in real life baseball with the bat or at the bases.

A year after OPS was created, in 1985, Bill James introduced a measure for offensive performance he called Runs Created . This concept has been improved (see previous reference) constantly to account for things like stolen bases and sacrifices.  Tom Tango, eventually improved the concept or Runs Created and morphed it into weighted Runs Created or wRC.  wRC is a good way of looking at someones performance, but (like RC) it is cumulative; think Runs, HRs and RBIs.  This is fine for evaluating a season (like who had the best season in an MVP type of consideration) or whether someone's career is HOF worthy (both cummulative questions; for the record I do prefer WAR over wRC to answer those, since it includes fielding, among other reasons); however it does not answer the question of who performed the best for the time he was on the field (and maybe play him more, or call him "up", if necessary.)  

So we moved from a Gremin to a 'cuda (Yes, I like 'cudas too, especially the Hemi version, but they are not without faults)  Enter wRC+ to the equation.  wRC+ is a lot like OPS+.  It is based on wRC (runs created with the bat and on the bases, independent of league and park,) but it normalized to 100 (denoting league average, like OPS+,) and furthermore it is a rate metric (think UZR/120 vs UZR) so you can look at chunks of performance, moving plate appearance variations (sample size) aside.    A good writeup in basic terms about wRC+ is found here:   This interesting article from Denver Post (an enlightened mainstream newspaper), uses wRC+ to argue the greatness of Carlos Gonzalez.

Here is a real life field test of the effectiveness of cumulative measures like wRC (and WAR) compared to rate stats, like wRC+.  Work with me: Let's rewind to the Twins' last good season (2010) and ask the question: who was the best Twins' player on the field (for the time he played, assuming that it was significant; I like the 200 PA mark as a cut-off point) and try to find an objective quantitative measure to support your argument.  "Ribies" and ball sounds and the like are not allowed.

Here are the top 6 in wRC, a cumulative measure, with Plate Appearances (and OPS for reference) in parenthesis, in descending order.

Joe Mauer 93 (584, .871)
Delmon Young 87 (613, .826)
Michael Cuddyer 82 (675, .753)
Justin Morneau 75 (348, 1.055)
Denard Span 73 (705, .679)
Jim Thome 71 (340, 1.039)

So basically, the cummulative stat, because of the disparity of sample size favors an average player (Cuddyer) and a below average player (Span) because they had twice as many PAs as Morneau and Thome; the "longevity" effect in HOF discussions. 

Let's look at the best wRC+ with a 200 PA minimum (to filter players like Luke Hughes and Carl Pavano who make the list but had less than 10 PAs) :

Justin Morneau 184
Jim Thome 178
Joe Mauer 136
Delmon Young 120
(Danny Valencia 118)
Michael Cuddyer 104
.
.
.
(4 players)
Denard Span 88

This is what wRC+ does.  And this was the season that every single Twins' fan was saying what that season would have been, "if" Morneau did not hit his head.  I hope this little example illustrates the value of wRC+ vs things like wRC and WAR.

After the long but necessary introduction, back to the original subject, Miquel Sano.  If trying to find objective measures that evaluate performance on the field is difficult, finding objective quantitative measures that can estimate future performance, used to discuss who is the best prospect or to create prospect "lists" is borderline impossible (like trying to create a vehicle that is using air for fuel and emits water for exhaust.)  But we can dream and play.   Earlier this year, during the off-season, I looked at potential future performance of the starting pitchers in the Twins' organization using objective quantitative measures.  I will repeat the exercise this coming off-season looking at the position players and the whole slew of pitchers and not only starters.   I will be using wRC+ as the basis of that endeavor, based on the discussion above.  In addition, age and level of play will also be major factors. But for this piece here, before it turns into War and Peace, let's focus on Miguel Sano's performance this season in a quantitative way and putting it in perspective:

Miguel Sano in 243 PAs in A+ (average league age for hitters is 23) at age 20  had a 201 wRC+.  In 190 PAs in AA(average league age for hitters is 24.5) at age 20  he is having a 158 wRC+.   His AA numbers are in flex and we can discuss more after the season is over, but his A+ number are final.  To put that 201 wRC+ in perspective:

- It is the highest in the minors this season (second highest is Chris Colabello with 197 in Rochester, who is well in his prime and older than the average AAA player)
- It has been achieved by only 5 major leaguers in the past 30 years  (mininum 200 PAs): Miguel Cabrera,  Barry Bonds, Mark McGwire, Jeff Bagwell, Frank Thomas; and Cabrera's is this season, so it is not final.

For the fun of it, and the perspective of Sano's dominance, here is a list of all the MLB leaders in wRC+ (and some close to the leaders) with career highs indicated for the last 30 seasons:


2013: Miguel Cabrera, 210, in progress
2012: Mike Trout & Miguel Cabrera tied,  166
2011: Jose Bautista, 180 (career high)
2010: Josh Hamilton, 175 (career high)
2009: Albert Pujols, 180 ; Joe Mauer, 170 (career high)
2008: Albert Pujols, 184 (career high)
2007: David Ortiz & Alex Rodriguez tied, 175 (career high for both)
2006: Travis Hafner, 176 (career high)
2005: Alex Rodriguez, 174
2004: Barry Bonds, 233 (Todd Helton, 166, career high)
2003: Barry Bonds, 212 (Albert Pujols, 184)
2002: Barry Bonds, 244 (Jim Thome 189, Manny Ramirez 185); all three career high
2001: Barry Bonds, 235 (Jason Giambi 193, Sammy Sosa 186; career high for both);
2000: Manny Ramirez, 181
1999: Manny Ramirez, 172 (Mark McGwire, 168)
1998: Mark McGwire, 205 (career high)
1997: Mike Piazza, 183, career high (Frank Thomas, 179, Larry Walker 177, career high)
1996: Mark McGwire, 190 (Garry Sheffield 185, career high)
1995: Edgar Martinez, 182 (career high)
1994: Frank Thomas & Jeff Bagwell, 205 (Albert Belle 186); all three career high
1993: Barry Bonds, 193
1992: Barry Bonds, 198
1991: Frank Thomas, 179 (Chili Davis, 139)
1990: Rickey Henderson, 190 (career high)
1989: Kevin Mitchell, 184 (career high)
1988: Jose Canseco, 169 (career high)
1987: Jack Clark 176, Wade Boggs 171; both career high (Kent Hrbek, 134)
1986: Don Mattingly 160 (career high)
1985: Pedro Guerrero 181 (George Brett 168); both career high
1984: Mike Schmidt 154 (career high)
1983: Wade Boggs 155

And here are the wRC+ career highs and career averages for selected Twins' hitters

Harmon Killebrew, high: 176, 1967; career average: 142
Tony Oliva, high: 154, 1971; career average: 129
Rod Carew, high: 175, 1977; career average: 132
Kent Hrbek: high: 146, 1988 career average: 126
Kirby Puckett: high: 150, 1988 career average: 122
Dave Winfield: high: 161, 1975; career average: 128
Paul Molitor: high: 145 1991; career average: 122
Joe Mauer: high: 170, 2009; career average: 133
Justin Morneau: high: 184, 2010; career average: 118

Sano's career worst wRC+ is 146 last season in Beloit.  As you can see, his wRC+ numbers this season, in combination with his age, his level of play and the average age in that level, are totally out of the stratosphere. 

How about Byron Buxton?   Buxton is only 7 months younger than Sano, and has been playing at a full level below Sano.  So the age/level combination is very similar (but still on Sano's favor).

Here are Buxton numbers:  in 321 PAs in A (average league age for hitters is 21.5) at age 19  had a 173 wRC+.  In 174 PAs in A+ at age 19 (average league age for hitters is 23) he is having a 129 wRC+.   For comparison's sake, the Angels' Mike Trout, at the same age (19), at the same league (Midwest), at the same team (Cedar Rapids) had an identical 173 wRC+.   Objectively, Byron Buxton, looks like he is the prospect Mike Trout was and had a great season in Cedar Rapids (and this is great news for the Twins), but Sano had a monumental season in Fort Meyers.  Better than Morneau's 2010 with the Twins.  I just cannot see how objectively anyone can justify ranking Buxton higher than Sano, or not rank Sano as the number one prospect in baseball this season; unless it is the sound the ball makes when it meets his bat (or something)

Sano is poised to be a centerpiece in the majors for years to come.  Fans will enjoy seeing his #PowerBananas for a long while.  I just hope that they are all with him in a Twins' uniform...

Assessing the Starting Pitching in the Twins' Organization

For a baseball team to content, one of three things need to happen:

1. An organization has to develop impact starting pitchers.
2. An organization has to trade for impact starting pitcher.
3. An organizations has to sign impact starting pitcher free agents.

I demonstrated earlier this month that the Twins' last two year abysmal record could have been predicted as earlier as 2008, based on the state of their starting pitching prospects.  Only 11 pitchers who have been in the Twins' minors in 2008 are still in the organization and none has been an impact starter.  The Twins have been adverse in doing numbers 2 and 3 above, so in order to compete, they have to develop starting pitching talent.

The Denard Span and Ben Revere trades infused the Twins with 3 young starting pitchers: Vince Worley, 24, and Trevor May and Alex Meyer, both 22, which makes the future a bit more hopeful.  How hopeful?  I will try to quantify, so the rest of the discussion here will be metrics and numbers based.  This will actually be somewhat of a logical continuation of this analysis, where in August last year, I tried to look positively into the 2012 Minnesota Twins pitching and draw conclusions based on potential.  What I am doing here is looking at the whole organization Starting Pitching, under very similar metrics and see what the future might look like.  This will include potential rankings of Twins' starters, but it is not a prospect list.  They are based on their 2012 performance (and adjusted for age and playing level) and not their potential.  Injured players, such as Wimmers and Salcedo will be higher on prospect lists that ranked here, because their numbers were awful.

The metrics I like to use to do this have been some simple things of my own device: Pitching Effectiveness or PE and Expected Pitching Effectiveness or xPE.  I fiddled around with PE in 2008 and with xPE in 2009.  Here is the reason I devised PE and here is the reason I optimized it to xPE.  My main arguments were a. I felt like xFIP and FIP and DICE weigh too much things like home runs (which anyone who watched the home runs by Miquel Cabrera and Delmon Young against the Twins yesterday cannot deny that they are a matter of inches and ballpark and luck and fielding performance from being a long fly ball).  Also these formulae are hard to memorize and I wanted something simple I can calculate looking at a stat sheet and also something that you can calculate using splits (e.g. how has Brian Duensing or Glen Perkins been as a starter vs as a reliever); you can find xFIP around, but not in a spit form.  So in 2008 I devised PE, which simple takes account three things:  Strikeouts, walks and hits.  So a pitcher who strikes out more people, walks fewer and gives lesser hits is more likely to succeed than someone who doesn't.  And all hits are counted equal because the difference between a single and a triple might be the difference of having Delmon Young or Ben Revere play Left Field or the difference between a fly ball out and a home run might be the difference of having Torii Hunter or Rich Becker playing Center Field.  And I use WHIP, K/9 and K/BB to calculate PE (a simple PE= (K/9*K/BB)/WHIP).  xPE further normalizes for BABIP (to league average .290) to account for "luck" with hits.  And unlike FIP and ERA, these two measures go the opposite directions (higher is better) and have a large variation (0.x to 100+) to allow for granularity in comparisons vs. compressing performance from 0 to 10 or so.  Over here, I show that xFIP and FIP correlate pretty well to the much more complex SIERA, which is way too complex to be able to calculate just with a cell phone calculate (which is my goal as far as metrics go.)

PE and xPE have been fine to show performance and expected performance.  How about potential?  This is the many million dollar question, because if someone is able to guess estimate future potential of a player in single A, he/she will be having a great advantage in identifying cheap, future impact players, in a more objective way than scouting reports.  So yet a new measure in the PE family has been devised:  the adjusted expected pitching effectiveness or axPE.  I tried something similar the off-season after the 2008 season, but the resulting formula was too complicated (cannot fit on a T-shirt or be calculated using a cell phone calculator), so needed to be refined.  I hope I am close to this, since axPE is simpler.  It takes into consideration level of play.  Each level of play gets a number.  Here are these numbers:

All Rookie Leagues: 1
A: 2
A+: 3
AA: 4
AAA: 5
MLB: 6

the average of the levels a player participated is taken into consideration for axPE.  For example if a pitcher spent part of the season in high A and then moved to AA, the average level grade is 3.5.

The other adjustment involves someone's age.  Younger players in higher levels have higher potential; this is the premise here. axPE is defined as xPE* (level/age) *7 .  The 7 is a coefficient that makes it a number  in the neighborhood of PE and xPE.

A note of importance:  agPE is biased towards better performance in higher levels; this is by design, since there have been pitchers who blew away rookie leagues and then bottomed out when they went to AA.

The PE family metrics translators for starters (and relievers, for the sake of completeness, but RP are out of score here) are roughly translated to:

Rotation:
xPE/axPE
35+ Ace
25-35 #1- #2 Starter
15-25 #2- #3 Starter
10-15 #3 - #4 Starter
7.5-10 #5 Starter

Bullpen:
xPE/axPE
35+ Closer
25-35 Closer-Setup
15-25 Setup- Long Relief
8-15 Long Relief-Mopup

where axPE denotes intermediate/long term potential.

So, do the Twins have any potentially impact starters in their organization, based on their 2012 performance?

Without further ado, here are the numbers, that include pretty much every pitcher in the Twins organization, including the new ones, and those who pitched only in the major league level in 2012, under 30 years old.  The age indicated is their age in 2012.  I am including Nick Blackburn, for comparison's sake.  Raw data is taken from B-R and an (*) denotes LHP:

Thus,  it looks like the Twins have 3 potential impact starters in their organization; Kyle Gibson and Alex Meyer are not a surprise.  Cole DeVries is.  Cole DeVries' axPE is higher than his xPE level (which turns out to suggest a middle of the rotation starter), because he performed at the MLB level.  Whether or not potential is applicable to a 28 year old who has reached the majors, is a good discussion.   On the other extreme, some of the K/9 leaders in the organization, Josue Montanez, Taylor Rogers, Felix Jorge, Tyler Jones have repressed axPE, because they are still at the lower levels of the organization.  I think that they need to prove themselves at higher level.

Based on this, and if you cut the list at 25 years old or younger, the Twins have at least couple of pitchers who have impact starter (i.e. top of the rotation/ace) potential and several who have mid-rotation potential. 

A huge qualifier:  This list is of pitchers who were used mostly as starters (i.e. made more starts than relief appearances in 2012.)  This leaves at least one particular pitcher out who should be included, but he made 4 starts and 7 relief appearances:  Jose Berrios.  His numbers (albeit in 11 games and 30 some innings) are out of this world: 284.12 PE, 236.11 xPE, and 91.82 axPE.  He should be part of the discussion and definitely has top of the rotation potential, but there is an asterisk for the reasons mentioned.

Others who made few starts but mostly used in a relief role but definitely should be part of the equation are (in no order) : Matt Houser, Miguel Munoz, AJ Achter, Cole Johnson, Argentis Silva (the 16 year old high bonus singing) Elias Villasarra, Fernando Romero, Luke Bard, Jose Jimenez, Corey Kimes, Brett Lee and Mason Melotakis.

Again, this is an intermediate term discussion and does not really involve recent Twins veteran acquisitions Rich Harden, Mike Pelfrey and Kevin Correia.

Something about pitching or why Kevin Slowey was more effective than Johan Santana

In order to evaluate position players more effectively, a couple of weeks ago I introduced a new statistic, bating and fielding efficiency (BFE). There are some established statistical measure that can tell you about a pitcher's performance, independent of the fielding of the team behind him. A couple of those are Fielding Independent Pitching (FIP) and eXpected Fielding Independent Pitching (xFIP) and Defense-Independent Component ERA (DICE). You can see the formulae for these measurement in this excellent Wikipedia article. I have 2 problems with these formulae:

1. They use arbitrary numericals to factor and add to the statistical measurements within their equations (3, 13, 2, 3.2)


2. For some strange reason, hits given are not included, but home runs are, factored by a huge 13-times factor


3. Bases on balls are factored by 3, strikeouts are factored by 2


4. Home runs are based mostly on a hitters capability and the park and not on the pitcher, whereas the defense does not have that much of a role for a hit.

Let's clarify the last point:
One of the statistics I will use to evaluate pitching is WHIP. Arguably, WHIP might be defense dependent, but how much?

Defense has 2 flavors:

a. accuracy - reflected by Errors. Errors do not count on WHIP, so that is out.

b. range - let’s use the plus minus system for this to understand the impact:

in 2008 the best defender in baseball as far as plus-minus goes was Chace Utley with a total score of 49 (i.e. he made 49 plays the average player does not make). He made a total of 803 plays (340 POs and 463 assists) even if we assume that all those +49 plays took a hit away (which is a stretch, because some of those were to get the lead runner in a double play or fielder’s choice, both of which do not take hits away from a pitcher’s WHIP).

49 is 6.1% of his total plays. If you divide that by 5 starting pitchers you get 1.2%.

So the best defender in baseball saved 1.2% of the hits for a particular starting pitcher (stretch). If you take the MLB average for 2008 pitchers 0.37 BB/Hits, 0.37 of a pitcher’s WHIP is a factor of BB and 0.73 a factor of hits. So the difference that the best ranging defender can potentially make on a pitcher’s WHIP is 0.88%

With examples:

Perkins’ WHIP in 2008 was 1.470, in that theoretical best case scenario would have been 1.457

Let’s go more extreme: here are the best plus minus numbers per position in 2008 (2B was Utley): 1B +24, 3B + 32, SS +23, LF +23, CF +32, P +16.
If you build a team with those people as defenders, the 0.88% difference above due to just Utley would become 2.8%

In other words, Perkins with the best defense in the MLB Universe of 2008 would have a WHIP of 1.429 instead of 1.470 (even with the best case scenario that all plus plays take hits away). Not much difference. Certainly not enough to discount the hits a pitcher gives as part of how effectively he pitches.

I created a new measurement, called Pitching Effectiveness (PE) which is simply: (K/9*K/BB)/WHIP. All these factors take into account how a pitcher is performing without extraneous factors. Here is a list of the 2008 MLB pitchers who pitched more than 20 innings and had a PE over 20, broken down by starters and relievers and sorted by decreasing PE (I included all Twins pitchers for comparison):

Starters

Dan Haren, AZ 39.13
Josh Beckett, BOS 37.83
Roy Halladay, TOR 37.81
Ervin Santana, LAA 35.79
CC Sabathia, MIL 34.08
Ricky Nolasco, FLA 31.68
Cliff Lee, CLE 30.84
Rich Harden, CHC 30.77
Kevin Slowey, MIN 30.67
Tim Lincecum, SF 28.29
Randy Johnson, AZ 26.85
Mike Musina, NYY 26.66
Cole Hamels, PHI 26.52
James Shields, TBR 23.23
Johan Santana, NYM 22.54
Javier Vasquez, CHW 21.46
Roy Oswalt, HOU 21.19
Scott Baker, MIN 20.98
Ben Sheets, MIL 20.97
Zack Greike, KCR 20.86
Jake Peavy, SD 20.50

Other Twins:

Boof Bonser: 13.44
Fransisco Liriano: 11.91
(Matt Garza 10.91)
Nick Blackburn: 8.09
Glenn Perkins: 5.69
(Livan Hernandez: 3.13)

Relievers

Mariano Riviera, NYY 189.22
Jonathan Papelbon, BOS 101.06
Billy Wagner, NYM 57.94
Sergio Romo, SFG 51.05
Kerry Wood, CHC 49.00
Hong-Chih Kuo, LAD 48.76
Grand Balfour, TB 48.49
Trevor Hoffman, SD 45.02
Joe Nathan, MIN 44.89
Scott Eyre, PHI 43.89
Matt Thorton, CHW 41.92
Brian Fuentes, COL 39.87
Mike Adams, SD 38.14
Joey Devine, OAK 37.91
Tagashi Saito, LAD 36.16
Joakim Soria, KCR 35.56
Carlos Marmol, CHC 35.22
Frank Francisco, TEX 32.67
Octavio Dotel, CHW 32.43
Rafael Perez, CLE 32.16
Scott Linebrink CHW 32.00
Chad Qualls AZ, 31.91
Jonathan Broxton, LAD 31.87
Jose Valverde, HOU 31.71
Mike Gonzalez, ATL 31.14
Jon Raunch, AZ 28.83
Max Scherzer, AZ 27.06
Brad Lidge, PHI 25.61
Joba Chamberlain, NYY 25.50
Neal Cotts, CHC 25.10
Taylor Buchholz, COL 24.89
Wade Corey, LAD 23.65
Joe Nelson, FLA 23.01
JUan Cruz, AZ 22.52
Jeremy Affeldt, CIN 22.37
Damaso Marte, NYY 22.36
Ramon Troncoso, LAD 22.10
Edwar Ramirez, NYY 21.89
Jesse Carlson, TOR 20.91
Doug Brocail, HOU 20.90
Arthur Rhodes, FLA 20.45
Brandon Morrow, SEA 20.12
Manny Delcarmen, BOS 20.08

Other Twins:

(Pat Neshek: 31.64 less than 20 innings)
Dennys Reyes: 16.59
Craig Breslow: 13.59
Jesse Crain: 10.90
Matt Guerrier: 7.00
(Juan Rincon: 6.27)
(Brian Bass: 4.58)

The MLB averages for 2008 were:

MLB starter average WHIP 1.39
MLB starter average K/9 6.20
MLB starter average K/BB 2.06
MLB Starter average PE: 9.19


MLB reliever average WHIP 1.39
MLB reliever average K/9 7.80
MLB reliever average K/BB 1.94
MLB reliever average PE: 10.89


It is expected that relievers would have higher PE numbers than starters and they do. A couple of observations: Slowey and Baker were the best Twins starters. Slowey is among the pitching elite. BTW, for those who are lamenting the Delmon Young trade, Matt Garza's PE was 10.91, placing him as the 5th best potential Twins starter. Glenn Perkins is trailing the starting group and Nick Blackburn's numbers are below MLB average, making Perkins and potentially both expendable. Bonser should get another chance to make the rotation and Liriano's PE was hurt by his horrid first few starts.

The Twins need at least 2 relievers above 20 in 2009 now that they lost Neshek. Mijares could be there, but at least another right handed reliever would be welcomed. From the list of relievers here, Jeremy Affeldt (22.37), Doug Brocail (20.90), Juan Cruz (22.52), Brian Fuentes (39.89, probably a closer someplace other than the Twins), Trevor Hoffman (45.02, probably a closer someplace other than the Twins), Bobby Howry (23.40), Damaso Marte (22.36), Arthur Rhodes (20.45) and Kerry Wood (49.00, probably a closer someplace other than the Twins) are free agents; so are Kyle Farnsworth (16.55), Brandon Lyon (15.23), Will Ohman (15.74), Darren Oliver (15.62) and Russ Springer (17.26), all with PE higher than the remaining Twins. Given the facts that the Twins would probably like a right hander and some of the above players will sign as closers, the list gets smaller, but there are still more than a few potential targets remaining. Other than Guerrier, the remaining bullpen arms were above league average in 2008.

One last parting thought: The PE is a useful tool to identify future closers. I think that PE > 40 equals good closer material. From the above list:

Sergio Romo, SFG 51.05
Hong-Chih Kuo, LAD 48.76
Scott Eyre, PHI 43.89 and
Matt Thorton, CHW 41.92

show closer potential.

Do stolen bases and GIDP matter for Gardy's Twins to win?

A couple of weeks ago I examined what makes the Gardenhire Twins win, looking for potential improvements in that caterogy. I was recently asked to examine the effect of SB and GIDP and its correlation with Twins wins. The following should be familiar. It basically lists all the statistical measurements examined in the previous post with the addition of SB and GIDP and their correlation to wins.



One surprise: Stolen bases have a negative correlation to wins for the Gardy Twins, which is actually stronger than the positive correlation of runs scored (compare the pink boxes). In plain english, this means that this team has less chance to win if they steal. By first glance this looks like a paradox, albeit an interesting one. OPS still has the higher correlation to wins and GIDP does have an expected negative correlation (albeit not extremely strong) with wins.

I dug further and examined the correlation of both GIDP and SB with the other offensive statistic measurements as well as with wins:



GIDP (top line in the bottom of the table) has the highest negative correlation with SLG and the highest positive correlation with SB. SB, has the highest positive correlation with GIDP and the highest negative correlation with wins.

In other words:

1. The higher the number of stolen bases the less wins the team has
2. The higher the number of stolen bases the more GIDP (and the reverse) the higher the number of GIDP, the higher the number of SB
3. The higher the number of GIDP the lower the SLG.

Statement 3 makes absolute empirical sense. Statement 2 is fine also: the more a team grounds into double plays, the more it wants to run to prevent DP so the more SB. Statement 1 is the kicker that defies empirical knowledge and bit of further discussion is necessary.

Regardless the perception that stolen bases increase the probability of a team to win, James Click of Baseball Prospectus has shown that this is not a case in an article called What if Ricky Henderson Had Pete Incavilia's Legs?, Published in the Baseball Prospectus' book Baseball Between the Numbers. The previous link is a link to the whole book (pointing at the pertinent chapter) available free at google books. Great read. A must for stat fans. So despite the popular empirical opinion, stolen bases have been proven to decrease win probability and the 2002-2008 Twins, confirm this fact.

As a conclusion, OPS is still the best correlating measurement to Wins for these Twins. And OPS (and projected OPS) could be used as a leading indicator to predict wins as I did here. This is fine for these Twins, but how about the rest of the league? What if you wanted to start a team from scratch or select a fantasy team? What would be the best league indication away from the context of the Gardy Twins?

To answer this question I looked at the same statistic measurements I examined for the 2002-2008 Twins, with the addition of Pythagorian Wins a Bill James measurement that predicts wins based on runs scored and allowed. Pythagorian Wins is a lagging indicator, which it means that it confirms trends and events rather than predict them (in other words, a game has to be played before you get the RS and RA measurements, whereas you can use historical or predicted OPS values to predict future performance). The following chart will probably look like an eye-chart:




Here is the summary. Looking across the MLB teams in 2008. The best correlating measurement with Wins is Pythagorian Wins (0.922). The correlation of other categories were: BA:0.405, OBP:0.521, SLG:0.566, OPS: 0.592, RS: 0.588, GIDP: 0.033 (practically non correlated), SB: 0.489 (surprisingly a + correlation), ERA: -0.649 (the higher the ERA the fewer wins), WHIP: -0.687 and RA: -0.641. Unfortunately, not a single statistic that could be used as a leading indicator has similar correlation to wins as Pythagorian Wins (runs cannot be used as leading indicators). Plan B: create a composite measurement. I created 8 measurements dividing each of the 4 offensive categories (BA, OBP, SLG, OPS) with the 2 pitching categories (ERA, WHIP). Here is their correlation to Wins: SLG/WHIP: 0.907, OBP/WHIP: 0.815, BA/WHIP: 0.779, OPS/WHIP: 0.892, SLG/ERA: 0.867, OBP/ERA: 0.763, BA/ERA: 0.765 and OPS/ERA: 0.835.

As you can see SLG/WHIP has a correlation of 0.907, close to that of Pythagorian Wins 0.922) and could be used as a leading indicator for team wins. In other words, if you assemble a team from scratch, real life or fantasy look for batter with high SLG and pitchers with low WHIP.

But how about them Twins? Well, for the Gardy Twins, the correlation of wins with Pythagorian wins was 0.768 and the correlation of SLG/WHIP with wins was 0.804, both lower than the correlation of OPS with wins (0.886). So, in other words, if you want to make the Twins better, look for batters with high OPS, esp the SLG part of OPS, because that correlates with OPS for these Twins at a close to absolute 0.959 rate.

Why is that discrepancy between the Gardy Twins and the rest of the league? Here is my theory and deemed to be vastly unpopular but not surprising to the people who have been following this blog. Look at this table for the Gardy Twins:

year         actual wins      pythagorian wins    AL Central Record- Twins

2002            94                  87                 .421
2003            90                  85                 .432
2004            87                  88                 .452
2005            83                  84                 .492
2006            93                  94                 .502
2007            79                  80                 .502
2008            88                  90                 .492

The harder the division has gotten in the Gardy ERA (right column is the record of the rest of the division) the harder time the Twins have to achieve their predictive record. They fell short a win or two, but a win or two would have gotten the Twins in the postseason this year. Why does this happen? Methinks is the manager who does not realize the full potential of the team. But this is another long discussion.

Next: evaluating pitchers.

Middle infield revisited:

After devising the BFE measurement to apply to third basemen, I am examining the ranking of MLB second basemen and short stops based on their 2007 BFE numbers (Twins in bold, SS free agents in italics) Minimum 200 innings at the position and 250 AB:

Short Stops:

Rafael Furcal, LA: 1.331 (fewer than 250AB)
Hanley Ramirez, Fla: 1.194

Jerry Hairston Jr., Cin: 0.983
Mike Aviles, KC: 0.959
Jose Reyes, NYM: 0.949

J.J. Hardy, Mil: 0.890
Stephen Drew, Ari: 0.871
Jhonny Peralta, Cle: 0.854
Jimmy Rollins, Phi: 0.853
Derek Jeter, NYY: 0.852
Yunel Escobar, Atl: 0.848
Cristian Guzman, Was: 0.845
Jed Lowrie, Bos: 0.824
Clint Barmes, Col: 0.821
15.Nick Punto, Min: 0.808

Michael Young, Tex: 0.773
Miguel Tejada, Hou: 0.766
Ryan Theriot, ChC: 0.758
Marco Scutaro, Tor: 0.746
MLB AVERAGE: .745
Troy Tulowitzki, Col: 0.717
Maicer Izturis, LAA: 0.716
22.Brendan Harris, Min: 0.701

Erick Aybar, LAA: 0.690
Orlando Cabrera, CWS: 0.683
Jason Bartlett, TB: 0.682
Yuniesky Betancourt, Sea: 0.672
Edgar Renteria, Det: 0.661
David Eckstein, Tor/Ari: 0.645
Jack Wilson, Pit: 0.629
Julio Lugo, Bos: 0.622
Bobby Crosby, Oak: 0.610

Cesar Izturis, StL: 0.558
Jeff Keppinger, Cin: 0.539
Khalil Greene, SD: 0.527
Angel Berroa, LA: 0.507

36.Adam Everett, Min: 0.497
Omar Vizquel, SF: 0.407

Juan Castro, Bal/Cin: 0.279

Second basemen:

Mike Fontenot, CHI: 1.131 (246 AB)
Chase Utley, Phi: 1.105
Ian Kinsler, Tex: 1.069
Dustin Pedroia, Bos: 1.042
Dan Uggla, Fla: 1.024

Brian Roberts, Bal: 0.967
Mark DeRosa, ChC: 0.953

Ray Durham, Mil/SF: 0.867
Kelly Johnson, Atl: 0.858
Placido Polanco, Det: 0.849
Orlando Hudson, Ari: 0.846
Alexei Ramirez, CWS: 0.834
Joe Inglett, Tor: 0.825
Jose Lopez, Sea: 0.824
Kazuo Matsui, Hou: 0.823
Ronnie Belliard, Was: 0.812
Clint Barmes, Col: 0.804

Howie Kendrick, LAA: 0.796
Aaron Miles, StL: 0.794
MLB AVERAGE: 0.792
Mark Grudzielanek, KC: 0.782
Mark Ellis, Oak: 0.776
Akinori Iwamura, TB: 0.771
Mark Loretta, Hou: 0.770
Brandon Phillips, Cin: 0.760
Jeff Baker, Col: 0.755
Edgar Gonzalez, SD: 0.745
28.Nick Punto, Min: 0.744
Rickie Weeks, Mil: 0.739
Felipe Lopez, Was/StL: 0.737
31.Alexi Casilla, Min: 0.736
Marco Scutaro, Tor: 0.721
Asdrubal Cabrera, Cle: 0.720
Jeff Kent, LA: 0.719
35.Brendan Harris, Min: 0.716
Robinson Cano, NYY: 0.701

Adam Kennedy, StL: 0.693
Jamey Carroll, Cl: 0.689
Juan Uribe, CWS: 0.654
Damion Easley, NYM: 0.629
Freddy Sanchez, Pit: 0.604

Luis Castillo, NYM: 0.591
Eugenio Velez, SF: 0.573
Tadahito Iguchi, Phi/SD: 0.516


Some observations:

• Surprisingly, second base seems to be a bigger problem for the Twins in 2008, than SS


• Punto was an above average SS, ranking 15th out of 34 players, surpassing players like Michael Young and Miguel Tejada, while Harris at 23 was 3 spots below MLB average


• Punto was the highest ranking 2B at #28 but still well-below the MLB-average. Castillo (despite the fact that he is projected as a lock for the position in 2009, in most people's minds) was ranked at #31 and Harris at #35.


• Tolbert did not have enough innings or ABs to qualify for ranking in any position, while Everett was ranked close to the bottom in the SS rankings

How does this change my previous assessment of the middle infield needs? Not much. I did propose changes in both 2B and SS, and the numbers reinforce the need for change in 2B. I also think that Punto with his performance last year, plus the high contract offers for middle infielders (see: Ellis, Mark) has priced his way out of the Twins' organization. Steve Tolleson with a spectacular ALF performance (.426/.463/.590, 2HR, 16 RBI in 61 AB as of this post) and a solid minor league season will be in the mix and potentially fight with Tolbert for a position in the 25-men roster.