Skip to content

Daniel Murphy Tops 2016 Baseball Most Undervalued

By Patrick Buzzard

The overall principles for this case study were very straightforward. The task was to pick any sport and any of the past three seasons and determine who was the most underrated player that season. For this study, I chose from the onset to focus on baseball and the 2016 season. At this time, it was a brand new dataset so I was interested in exploring it. I also chose baseball because data for it is widely available on sites such as fangraphs.com and baseball-reference.com.

I first went about it by going onto fangraphs.com and taking all players with over 300 plate appearances. I chose 300 because I determined that it was a good number to encapsulate players such as Jarod Dyson. As an occassional role player, the fact that he produced 3.1 Wins Above Replacement (WAR) in the 2016 was particularly surprising.

I then sorted my fangraphs table by including wRC+, Fld, BsR, OBP, SLG, and WAR. I felt these statistics conveyed the overall five tools of a baseball player. WAR is all encompassing, wRC+ conveys hitting ability, Fld conveys overall fielding ability and takes into account throwing, BsR tells of a player’s ability on the bases, OBP tells of a player’s ability to make it onto the bases, and SLG tells of power. This dataset looked like this when it was all said and done: i

Just thinking that these stats work though is not enough. After running a regression on the WAR of the players I found a .848 r squared and all the variables to have P-Values below .05 meaning everything was significant. The results of that regression can be seen below:

SUMMARY OUTPUT  
   
Regression Statistics
Multiple R 0.920842438
R Square 0.847950795
Adjusted R Square 0.845049092
Standard Error 0.758210981
Observations 268
  Coefficients Standard Error t Stat P-value
Intercept -7.932621873 1.141967065 -6.94645 2.9552E-11
wRC+ 0.043252006 0.010204469 4.238536 3.12031E-05
BsR 0.135442741 0.013599125 9.959666 5.03581E-20
Fld 0.114894284 0.006599335 17.40998 1.24321E-45
OBP 10.21008736 3.899950137 2.618005 0.00935923
SLG 4.813614418 2.35313547 2.045617 0.041792698

Now that I found that the variables were significant I calculated their z-score for each individual player. I then combined them and since it was a rather larger number, took a z-score of the total z-score to come up with a total value above mean statistic.

Since value in baseball ultimately comes from how much less you paying for a player based upon what they are worth I then took the difference between what each player was worth for the 2016 season and then subtracted their actual salary to create their surplus value. I then multiplied their surplus value by the total value z-score to measure how much a player utilized the surplus they were giving their team. The result of this calculation is illustrated in the table below:

Name WAR Z-Score Dol Sal Surplus   Value
Mike Trout 9.4 3.64060347 74.9 15.25 59.65   217.162
Kris Bryant 8.4 3.142083091 67.4 0.652 66.748   209.7278
Mookie Betts 7.8 3.019933198 62.3 0.566 61.734   186.4326
Corey Seager 7.5 2.173301043 59.9 0.51 59.39   129.0723
Josh Donaldson 7.6 2.446488269 60.7 11.65 49.05   120.0002
Jose Altuve 6.7 1.881991099 53.4 3.5 49.9   93.91136
Adam Eaton 6 1.908122481 48.1 2.75 45.35   86.53335
Freddie Freeman 6.1 2.336699051 49 12 37   86.45786
Francisco Lindor 6.3 1.611157431 50.6 0.5403 50.0597   80.65406
Manny Machado 6.5 1.709573528 51.8 5 46.8   80.00804
Daniel Murphy 5.5 2.071007653 44.4 8 36.4   75.38468
Brian Dozier 5.9 1.700145077 47.2 3 44.2   75.14641
Paul Goldschmidt 4.8 1.942141111 38.2 5.875 32.325   62.77971
Brandon Crawford 5.8 1.534029113 46 5.8 40.2   61.66797
Anthony Rizzo 5.2 1.693623103 41.2 5 36.2   61.30916

This led to players like Mike Trout, Kris Bryant, and Mookie Betts coming away with crazy numbers. This is a problem because they aren’t undervalued. They are both MVP’s. They are just under cost control. So, I removed from my dataset all those making under $1 million or over $20 million because those players all have been seen for the value they are worth or are under cost control. Then I removed MVP’s and other players who are of All-Star quality and haven’t hit the market. The end result of this can be seen in the table below:

Name WAR Z-Score Dol Sal Surplus   Value
Daniel Murphy 5.5 2.071007653 44.4 8 36.4   75.38468
Brian Dozier 5.9 1.700145077 47.2 3 44.2   75.14641
Brandon Crawford 5.8 1.534029113 46 5.8 40.2   61.66797
Justin Turner 5.6 1.453453638 44.5 5.1 39.4   57.26607
Jean Segura 5 1.380738099 39.9 2.6 37.3   51.50153
Ian Kinsler 5.8 1.550339044 46.3 14 32.3   50.07595
Dexter Fowler 4.7 1.599685549 37.6 8 29.6   47.35069
Christian Yelich 4.4 1.284550568 34.9 1 33.9   43.54626
Kyle Seager 5.5 1.209799634 43.7 8 35.7   43.18985
Starling Marte 4 1.451126196 31.8 3 28.8   41.79243
Anthony Rendon 4.7 1.191142521 37.5 2.8 34.7   41.33265
DJ LeMahieu 4.2 1.28599573 33.4 3 30.4   39.09427
Brandon Belt 4.4 1.293696168 35.1 6.2 28.9   37.38782
Charlie Blackmon 3.9 1.221646353 31.5 3.5 28   34.2061
Kole Calhoun 4 0.886229377 31.8 3.4 28.4   25.16891
Jarrod Dyson 3.1 0.836456342 24.4 1.725 22.675   18.96665
Neil Walker 3.7 0.8914835 29.9 10.55 19.35   17.25021
Ryan Schimpf 2.4 0.897774521 19.6 1 18.6   16.69861
Ian Desmond 3.3 0.586834353 26.8 8 18.8   11.03249
Steve Pearce 2 0.913750757 16.3 4.75 11.55   10.55382

After accounting for these variables, I found Daniel Murphy to be the most undervalued player of the 2016 season. He produced a surplus value of $31.9 million based upon the $12.5 million average annual value of his backloaded contract. Multiplying his surplus value by the 2.07 z-score he produces results in a score of 66.065. While certainly not the highest it is more than two standard deviations above the median score.

Daniel Murphy was able to accomplish this by vastly overperforming expectations of him going into his contract with the Nationals. Prior to the 2016 season Murphy had never reached a .900 OPS, only breaking the pivotal .800 OPS barrier in two seasons, though both of were not full seasons for him. He’d also never reached more than 3.1 WAR in his entire career. This past season he was worth 5.5 WAR. His wOBA of .408 was 83 points higher than any other full season of his career, 27 points higher than his best previous offensive season, his 49 game rookie campaign.

It remains to be seen whether or not Murphy can keep this production up. He didn’t gain any more patience, keeping a similar K% to his previous seasons along with a similar OBP-AVG, meaning that his higher OBP was largely AVG fueled. His XBH and hits in general did go up though, so it could have been that he added more power to his swing.

This could be the beginning of a new Daniel Murphy that the Nationals were able to capitalize on, or it could be a one season fluke. Either way the Nationals were able to get the most value for what they spent on Daniel Murphy out of every player on a free-market contract.