Lessons from Mr. Miyagi

Mr. Miyagi to Daniel:

First learn balance. Balance good, karate good, everything good. Balance bad, might as well pack up, go home.

Later on…..

Daniel: I'm just scared. You know, the tournament and everything...

Mr. Miyagi: You remember lesson about balance?

Daniel: Yeah.

Mr. Miyagi: Lesson not just karate only. Lesson for whole life. Whole life have a balance. Everything be better.

Later on…..

Daniel: You're the best friend I've ever had!

Mr. Miyagi: You... pretty OK, too. Go, find balance.

Appreciation of "El Come Dulce"

Unbeknownst to me, Bobby Abreu’s nickname was "El Come Dulce".  I’m having trouble finding the exact translation, but it seems to roughly translate as "He eats candy" or "the sweet eats", or something like that.  I’m not sure if it’s a reference to his eating habits or his sweet swing.  Anyway, it’s a cool nickname.

====================

Quick Sidebar: The All-Venezuela Team

In honor of Abreu, here’s my All-time Venezuela team.  It’s a good squad, very deep in quality shortstops: Luis Aparicio, Omar Vizquel, Dave Concepcion, Ozzie Guillen, Chico Carrasquel, Elvis Andrus, Alcides Escobar, Carlos Guillen. 

I might be a little premature in putting Altuve as the #1 second baseman or Carlos Gonzalez as the left fielder, but I’ll take my chances.  It’s a good squad:

Batting Order:

====================

Back to Abreu

Bobby Abreu was one of my favorite players.  I think it traces back to a connection I had with him dating back to my early fantasy baseball years.  I was already participating in one league where we held an auction from scratch every year, but I was invited into another league where they played what’s known as the "Ultra" version of Rotisserie….that is, you had a 40-man roster – 25 active players, and 15 on reserve, and the reserve roster could contain minor leaguers.  Then, from year to year, you could keep up to 15 players on your primary roster (although salaries would increase every year, so you couldn’t just keep players forever without paying the price) and up to 7 minor leaguers, so that your team was an ongoing concern that had continuity from year to year.  I really got hooked, because it made you study not just major leaguers, but also trained you how to look for good minor league prospects.

I joined the "Ultra" league in 1996, inheriting the team from someone that dropped out.  It wasn’t much of team, but one of his minor leaguers was Bobby Abreu, who at the time was in the Astros organization.  He had been in the organization since age 17.  By ’96, he was 22 and was Baseball America’s #29 prospect.  I was excited to have him.

He made a few appearances for the Astros in ’96 and ’97, but after the ’97 season it got pretty eventful for him.  One of the 2 new expansion teams (Tampa Bay) selected him with the 6^th pick in the expansion draft, but then immediately traded him (the same day) to Philadelphia in return for SS Kevin Stocker.  That turned out to be one of the most lopsided "one for one" trades in history.  Here are some of the most lopsided one-for-one exchanges that I found:

Christy Mathewson for Amos Rusie
Jeff Bagwell for Larry Andersen

Pedro Martinez for Delino DeShields

John Smoltz for Doyle Alexander

Bobby Abreu for Kevin Stocker

Once Abreu settled in Philadelphia, he was fantastic.  He spent 9 seasons with the Phillies as their primary right fielder, and he contributed across the board…..power, speed, average, getting on base, even winning a Gold Glove.

Abreu’s been out of the game for a year now (he retired after the 2014 season), and I’ve been thinking about him a lot lately.  He appeared briefly in our Hall of Fame Tournament project as an "At Large" candidate, but didn’t get out of that round. 

He was one of my favorites, but he’s not exactly what I would call a strong Hall of Fame candidate.  I think he was an outstanding player, and he was very valuable.  His rWAR is right around 60, which is a kind of unofficial threshold that many use for consideration of a good Hall of Fame candidate.

The problem with Abreu is….he wasn’t a "star".  I don’t think many people think of him as a Hall of Famer.   He didn’t grab headlines. He had only 2 All-Star seasons.  He rarely led the league in anything.  His "Black Ink" score is only 5 (average Hall of Famer=27), and his "Grey Ink" score is only 88 (average Hall of Famer=144).  He had virtually no support in for any major awards (his highest MVP finish was 12^th place, he won just one Gold Glove and one Silver Slugger).  I suspect when many fans hear his name, the first thing that pops into their minds is that he won the Home Run Derby in 2005, when he hit a then-record 24 home runs in the first round.  You may also recall that, in that season, he had 18 HR’s by the All-Star break, won the derby, then only hit 6 more HR’s in the 2^nd half, and there was a lot of speculation that his performance in the derby affected him the rest of the year.

He was consistent, he had a broad base of skills, he was valuable…..but I don’t think he’s destined for the Hall of Fame.  Even when he does come up on the ballot in 2020, the focus will be on some other guy by the name of Jeter. 

So, this isn’t about making a case for Abreu for the Hall of Fame.  I don’t think he’ll get much support.  No, this is simply in appreciation of "El Come Dulce" for what he was…..one of the more balanced offensive players we’ve seen.  I wanted to find out who were the "best across the board" type of players.  He inspired me to do this study.

Finding Balance

I decided to approach this study by examining some of the more basic offensive categories, but I wanted to not just look at career totals, but also some "rate" stats.  I didn’t want to just look at players that played a long time and compiled large raw numbers, but rather wanted to see how strong they were on a per AB or per PA basis.  I wanted to capture power, speed, plate discipline, ability to hit for a high average, and ability to get on base.  I decided on the following 13 categories:

Career Counting Stats

Games Played (basic measure of longevity)

Home Runs

Hits

Non-Home Run Extra Base Hits (i.e., Doubles + Triples)

Walks

Stolen Bases

Rate (or Per AB/Per PA) Stats

Batting Average

On Base Percentage

Slugging Percentage

Home Runs per Plate Appearance

Non-Home Run Extra Base Hits per Plate Appearance

Walks per Plate Appearance

Stolen Bases per Plate Appearance

Why doubles plus triples?  Well, I wanted doubles, and I wanted triples, but triples by itself isn’t much of a stat anymore.  There just aren’t that many.  But, I didn’t want to lose the impact of them for those that have been good at hitting them, so I bundled them with doubles and called it "non-HR extra base hits".  I’ll use "XBH-NHR" for shorthand from here on out, but just think of it as "doubles plus triples"

I also considered runs and RBI (both career counting stats as well as per plate appearance), but I thought those would be too dependent on team and batting order position (for example, an outstanding leadoff hitter would be pretty disadvantaged when it comes to RBI opportunities).  So, I did 2 versions of this study….one including runs & RBI, and one without, although I’m going to focus here on the one without runs and RBI.

I downloaded data for 3,866 hitters from Fangraphs.com.  Next, took the 13 categories for each player and translated them into percentiles to gauge how well each player did, relatively speaking, within each category.  To use Pete Rose as an example, since he’s been in the news so much anyway, here’s how his percentiles come out:

As you would expect, Rose comes out much better on the "counting stats", where he accumulated large totals, than he does per plate appearance.  That’s no surprise.  Rose topped out in games and hits, where he’s the all-time leader (100^th percentile), and he’s near the top in things like XBH-NHR (his 881 doubles plus triples is 5^th behind Cobb, Speaker, Musial, and Wagner, good for 99.8^th percentile), and walks (he’s 15^th all time, 99.6 percentile). 

On the other hand, while his raw HR’s are good for the 88^th percentile, HR/PA are only good for a 43^rd percentile.  And even though he had stolen more bases than 90% of the players in the data set, on a per-PA basis, he drops to 52.7.  Again, no surprise, as Rose was the master accumulator, in overwhelming with the total weight of the numbers, rather than being a great "rate" player.

Composite Sketches

In order to get at a number that would measure who was the best "across the board", I came up with two summary approaches for the 13 different categories:

1)      Multiply the 13 percentiles to come up with a composite number ("Composite Product")

2)      Find the harmonic means of the 13 percentile figures ("Harmonic Mean")

In either case, a really low performance in a category would really hurt your score, but wouldn’t necessarily kill it.  This rewarded players who were outstanding across the span of the categories.

For example, in the composite product measure, if you had 3 categories and you had the 80^th percentile in all 3, that would score better than two 90^th percentiles and a 60^th.  Your "average would be the same in both cases (80), but in the first example, your "composite" score would be .80*.80*.80 = .512, where as the second example would yield .9*.9*.6 = .486.  I then multiply by 100 to get a more readable figure. 

If you do this for Rose, you get a figure of 6.5.  Where does this place him by this measure?  Well, it’s not exactly at the top of the list.  He comes in ranked #160, in a knot with Dustin Pedroia, Shin-Soo Choo, and Alan Trammell.

The second measure uses harmonic mean.  You’re probably familiar with Bill’s "Power-Speed" number, which is the harmonic mean of two figures: Home Runs and Stolen Bases.  That uses two figures, but you can use harmonic means with any number of figures.  The general formula for "n" number of things you’re measuring is:

Harmonic Mean = n / (1/ x₁ + 1/ x₂ + 1/ x_3+……1/ x_n)

Where, n = Total number of numbers or terms.
x₁, x₂, x₃, .... x_n = Individual terms or individual values. 

So, we can do it even with 13 different figures.  In our case, n would be 13, and the denominator would be the sum of the "inverses" of the percentiles. 

In Rose’s case, he comes up with a harmonic percentile figure of 78.1.  Where does that put him?  About the same as the first measure….he ranks 166^th in this metric, basically tied with Harlond Clift and Leon Durham.

Again, this isn’t a measure of value or Hall of Fame worthiness or anything of the sort.  It’s intended to reward and classify players by how well they did across the span of the categories being measured.

Finally, I ranked players by those 2 metrics ("composite" and "harmonic"), and then averaged the two ranks to get an overall ranking.  Then, finally, I’ll show the final percentile ("Balance Percentile") to show how they compare to the full list.

Results

Let’s look at the top 50.  I’m not going to show all 13 categories for each player, because that would be a bit of an eye-chart, so this will just summarize everything without all the details:

The next 10 who just missed the cut are Shawn Green, Harry Stovey, Johnny Damon, Kiki Cuyler, Reggie Smith, Moises Alou, Harry Heilmann, Joe Morgan, Albert Belle, and Derrek Lee.

By the way, it’s a happy little coincidence that Jackie Robinson came out as #42 in the ranking (actually tied for 42).  I promise I didn’t do anything special to make that happen.

Also, I don’t know if Babe Ruth is considered a surprise on this list or not.  Obviously, he’s high 90’s in most hitting categories, but even in stolen bases he’s in the 82^nd percentile with 123 career steals.  His one weak category is SB per PA, where he’s about in the 50^th percentile, but that wasn’t enough to make too much of a dent in his overall showing.

As you can see, you don’t have to be a Hall of Famer to do well on this list.  About half of the names on the list are Hall of Famers.  Leaving aside cases like Bonds, who’s obviously qualified and not in for other reasons, and players like Chipper and Griffey Jr. who will certainly get in but just haven’t had their votes completed yet, there are several top performers who aren’t Hall of Famers that do well by this measuring stick:  Larry Walker, Jeff Bagwell (although he may get in soon), Brian Giles, Ray Lankford, Ellis Burks, Dick Allen, Bernie Williams, Bob Johnson, Minnie Minoso, Paul O’Neill, and others rate well by this approach.

Some of the top names are likely not a surprise.  Barry Bonds as #1 is probably one that many would have guessed (Barry’s dad, Bobby Bonds, by the way, comes in at #107…..still in the top 3% in the study).  Willie Mays, Hank Aaron, Babe Ruth, Rogers Hornsby, Frank Robinson, A-Rod….they all rank very highly.

A few surprised me.  I didn’t expect Larry Walker to be at #2.  Now, I should mention, that nothing in here is adjusted for time or place, so we need to take Walker’s performance with a grain of salt.  But, I’m not sure I much I’d knock him down.  Walker was a good all-around player even before he went to Colorado.  He flashed a lot of that promise with Montreal – good power, good speed, hitting for average.  He ended up in the 90^th percentile or better in every measured category except for Walks per PA (86.9 percentile) and Stolen Bases per PA (78.6).   Anyway, I didn’t include any adjustments for time or place.

And our man, Bobby Abreu?  He checks in at #3 on our list.  He really doesn’t have a bad category.  He’s basically 93^rd or above in every category except batting average (88^th), SB Rate (88^th), and HR per PA (79^th).

In fact, I thought I’d see, out of all players in the data, who had the "best" lowest category.  In other words, if you take everyone’s worst category of the 13 being measured, who had the highest of those?  Here’s who fared the best:

So, our man Abreu had the "best" worst category, or the "highest floor", if you will, which was one of the reasons he rates so well.   If your "weakest" category is essentially in the 80^th percentile, you really don’t have a weak category at all.  You could actually make an argument that these players exhibited the best balance, or at least had the greatest "absence of weakness" across the categories measured.

It was interesting to see Cliff Floyd do so well.  I don’t usually think of him having much of a career, because it always seemed like he was hurt, but he did do very well across the board in both career and rate stats.  He came in at #73.  And, Jackie Robinson I’m sure would have finished further up the list than #42 if he had had a longer career.

So who didn’t do particularly well?  Out of players with long careers, here are a few selected notable players outside the top 50:

Cal Ripken

Cal got killed by a few categories – 51^st percentile in career steals, and 13^th percentile in SB per PA.  In addition, his percentiles for Batting Average OBP, and BB rate were all in the 60’s and 70’s.  And, even though he’s in the 99^th percentile in career XBH-NHR, he’s only in the 59^th percentile in that category on a per PA basis.  So, he’s well down the list, all the way down at #712.

Mickey Mantle

Mantle comes in at #117, which really isn’t that bad, although I thought maybe he’d be higher.  As with some of the others here, his XBH-NHR Rate (doubles and triples per PA) was really low, 29^th percentile.  He just didn’t hit many doubles.

I can’t remember if this has been published before, but here’s a little freak-show stat for you.  In my data set, I’m showing that there are about twice as many doubles as HR’s.  However, there are some players that hit significantly more HR’s than doubles.  Here are the highest HR/2B ratios among players with 5,000 or more PA’s.  McGwire dominates this one…..

Rickey Henderson

Well, Rickey did make the top 100 at #99, which isn’t bad, but I thought he would rate higher.  One category that hurt him a lot was the XBH-NHR Rate category, where he was only in the 34^th percentile.  Which, if you look at his record, makes sense.  He really didn’t hit that many doubles or triples.  That wasn’t his game.  He had an awful lot of years where he didn’t even reach 20 doubles, and he averaged about 2-3 triples a year.  His HR rate, Slug Pct, and Batting Average percentiles were all only in the 70’s.

Brooks Robinson

Well, we didn’t really expect him to do well on this, did we?  The career stats are pretty high up except for steals, but the rate stats did him in.  Poor OBP, poor average, poor speed.  He’s way down at #1219.

Eddie Collins

Collins does pretty well on most categories, but the power (especially the HR’s) killed his chances.  He’s at #282.

Honus Wagner

Wagner came in at #83, which really isn’t bad.  Some of that is surely era-related, as Wagner played in an era of few HR’s, so he doesn’t fare well in those categories, and he probably would have done better if I tried to adjust for that.  His Walks per PAwas only in the 55^th percentile.

Andre Dawson

Dawson was certainly a good all-around player, but he only comes in at #408 here, as low OBP and BB rates hurt his overall score.  His Walks per PA of 5.5% was only in the 18^th percentile.

Well, I could go on forever, but this seems like a good time to wrap it up. 

Curious about others?  Just post your query in a comment below, and I’ll respond.

Wrapping it Up

In any event, our man Bobby Abreu does quite well by this approach.  I doubt there’s a Hall of Fame plaque in his immediate future, but he should take comfort in knowing that he rates well here.  It’s a tough standard to do well across all 13 categories.  Even great players tend to have a few categories in which they are less than stellar.

Congratulations, Bobby – you achieved balance.  Mr. Miyagi would be proud.  Now you just have to work on that "crane kick"…..