# Deep Dive: Estimating Curtis Granderson’s aging pattern

In signing **Curtis Granderson** to a $60 million contract over the next four seasons, the Mets made a sizable bet on improving the team in the short term. The slightly front-loaded deal covers Granderson’s age 33-36 seasons, the years where players of his ilk — outfielders who strike out regularly, but rely on power for their offensive value — decline dramatically.

In this post, Toby Hyde will dig into the type of player Granderson has been, and the one he’ll likely be going forward. Finally, he will construct an estimate of Granderson’s value to the Mets…

Toby Hyde, Contributor

Granderson clearly fills a need for the Mets: a big league outfielder. In 2013, Mets outfielders produced 7 fWAR, 20th in baseball. The since-traded Marlon Byrd was responsible for half (3.5 fWAR) of that production. Before Granderson agreed to terms with the Mets, Eric Young Jr. and his career 77 wRC+ (where 100 is league average) was the placeholder in the corner not occupied by **Chris Young**.

In some ways, Granderson is a similar player to C. Young: they both hit home runs, strike out regularly and have played centerfield in the past. With **Juan Lagares** in center, Granderson and Young should complete a very fine defensive outfield.

The Mets farm system was not prepared to offer a starting everyday outfielder for Opening Day 2014. Cesar Puello, the best option in the top three levels of the farm system in 2013, finished his year hitting .326/.403/.547 with a career-high 16 home runs in 91 games in double-A, but missed the final month of the season because of a Biogeneis-related suspension. An aggressive hitter, some time in triple-A will serve him well to work on his strike zone discipline. Perhaps, after a few months in Las Vegas, he will be ready to help the Mets by the heat of summer. Promising 2011 first round pick Brandon Nimmo ended the 2013 season in a-ball with Savannah. He’s at least two years away from the big leagues.

In short, the Mets needed another outfielder.

The left-handed hitting outfielder hit free agency after an impressive four-year run with the Yankees, when he hit a combined .245/.335/.495 with 115 home runs. His 41 home runs in 2011 were, at the time, a new career high, one he surpassed a year later when he hit 43 in 2012.

Granderson’s 2013 was marred by injury when he was plunked by a pair of pitches. The first, in spring training, broke his foream. After he returned to action in May, eight games later, a broken pinkie sent him back to the disabled list.

When he did play, Granderson hit .229/.317/.407 in 61 games. By wRC+, again where 100 is league average, he earned a 97, making him slightly below league average. By bWAR, Granderson was worth 1.1 wins, while in Fangraphs’ accounting he was worth 1.4. Prorating his time out to say, 140 games, we can estimate his bWAR at 2.53 and his fWAR at 3.22. Fangraphs assigns a positive number to his defensive contribution because he was playing centerfield in a below average way, and right field in an above average way. Defensive numbers over partial seasons bounce around, so those estimates on his defense are rough. (Note: I use both fWAR and bWAR in this piece. There are various comparative searches that I find easier at one site or the other. For most purposes, now that there is a unified replacement-level baseline, they’re similar enough.)

**Power Spike**

Why did Curtis Granderson have back-to-back 40-home run seasons in 2011 and 2012, his age 30 and 31 seasons? Simple. He had a HR/FB spike. He hit line drives, ground balls and fly balls at rates that looked like his final seasons in Detroit (2007-2009) and his first with the Yankees (2010). However, in all of those seasons, his home run to flyball ratio hung between 10.9% and 14.5%. In 2011 and 2012 it climbed to 20.5% and 24.2% respectively. By 2013, it was back down to 11.3%, in line with his career norms.

**Walks and Strikeouts**

Granderson is a high-walk, high-strikeout hitter. He fanned in 28.5% of his plate appearances in 2012 and 28.2% in 2013, while walking 11% in both.

**Platoon Splits**

Over the course of his eight-plus professional seasons, Granderson has run a major platoon split, hitting righties much better than lefties. He’s hit .274/.357/.519 against righties with a 21.8% strikeout rate and a 11.1% walk rate. Meanwhile, against lefties, he has hit just .226/.295/.409 with a 26.7% strikeout rate and a 7.8% walk rate.

In 2013, his worst season by OPS+ or wRC+ since 2006, Granderson actually hit lefties (.242/.356/.435 in 73 PA) better than righties (.224/.300/.395 in 172 PA). Largely, this was a result of unusually good results on balls in play. He had a .371 BABIP against lefties and a .279 BABIP against righties. His underlying rate stats were better against righties (25.6% k rate/9.9% walk rate) compared to lefties (34.2% K/14% walk).

Lefties have given Granderson fits, over the course of his career, but he has improved against southpaws from his younger days through his Yankee peak in 2012. For example, take a look at his on-base percentage against lefties, by year, moving forward from 2007 through 2012: .225, .310, .245, .292, .347, .304.

**Defense**

Once a strong center fielder, Granderson will play a corner outfield spot for the Mets, likely deferring to Juan Lagares’ prowess in center. UZR rated Granderson as a combined 27 runs above average in 2006 and 2007, but that was a long time ago. In his four years as a Yankee, he was a combined 14.2 runs *below* average. That’s a slightly controversial rating, as the other advanced systems do not rate him as poorly. Baseball Info Solutions’ Defensive Runs Saved, puts Granderson at seven runs below average for his Yankees’ tenure. Total Zone measures -8 for the same four-year span.

There’s a line of thinking that playing next to a very strong defender, as Granderson did when playing alongside Brett Gardner, will hurt his defensive ratings. If Granderson was playing next to say, **Lucas Duda**, he would have caught more balls, both within his zone, that is the area a center fielder is expected to cover, and outside of it. Both types of catches, in and outside his area, would have helped his zone numbers.

Lets assume moving forward that Granderson’s defense on the corners will help his value a little – say five-ish runs a year.

**Projection Moving Forward**

At this point, we should have a pretty good handle on what kind of player Granderson has been, but the question the Mets just spent $60 million to answer is what kind of player he will be in his next four years.

The Oliver projection system at Fangraphs sees Granderson as a 3.0 fWAR player in 2014, who experiences a slow, steady decline. Dan Szymborski’s ZiPS system does not see as big a bounceback in 2014, and then a similar if slightly steeper decline. Analysis that rests on these two projection systems which have Granderson contributing something like eight or nine WAR over the life of the contract turn generally positive. This includes Eno Sarris’ fine work here.

Many of the analyses of the contract assume a gentle aging curve for Granderson. However, most of his peers, once they started to go, went in a hurry.

I created my own specialized aging curve for Granderson. First, using Baseball Reference’s play index to find players like Granderson, I selected outfielders who put up between 15 and 25 WAR in their age 28-32 seasons who struck out at least 596 times. These are good, solid, above-average everyday outfielders. The bWAR cap of 25 removes six players – who are or approach Hall of Fame level (Alex Rodriguez, Sammy Sosa, Jim Thome, Reggie Jackson, Carlos Delgado and Bobby Abreu). The rest are comfortably Granderson’s peers: he put up a 18.3 bWAR over his last five years. (Note: I would have preferred a rate-based search around strikeouts or production, but neither Baseball Reference nor Fangraphs offers that publicly. My original cutoff was 600 strikeouts. However, moving down four whiffs adds Jim Edmonds into the list and, much more on him later. Edmonds is a fairly good match for Granderson’s peak by WAR value and strikeout rate.)

We’re dealing with 19 players, found here.

I removed the following four players who moonlighted in the outfield, but were not primarily outfielders in any sense of the word: Dan Uggla, Mo Vaughn, Carlos Pena and Jorge Posada.

I removed the following guys whose careers ended before 1980, because hey, it was a different game 33 years ago and players follow different workout regimes now, and can be expected to age differently. Bye-bye: Willie Stargell, Bobby Bonds, Tony Perez.

Again, to make sure we’re dealing with decent defensive outfielders, I mandated that all of the comps had played at least some center field. Goodbye, Frank Howard and Mikey Tettleton.

This left our final comparable list as: Alfonso Soriano, Dale Murphy, Gorman Thomas, Nick Swisher, Ray Lankford, Jayson Werth, Jeremy Burnitz, Mike Cameron and Jim Edmonds. Swisher, who is as old as Granderson, has not begun his age 33 season. His career offers no clue as to how Granderson will age in his 33+ years.

These are good offensive outfielders. Their combined age 28-32 seasons look like this.

Player | HR | WAR | SO | G | PA | BA | OBP | SLG | K% | CF |
---|---|---|---|---|---|---|---|---|---|---|

Alfonso Soriano | 172 | 16 | 639 | 704 | 3188 | .280 | .333 | .529 | .20 | 12 games |

Dale Murphy | 170 | 23.9 | 677 | 799 | 3459 | .275 | .368 | .512 | .20 | Yes |

Gorman Thomas | 165 | 16.3 | 721 | 731 | 3080 | .238 | .331 | .477 | .23 | Yes |

Curtis Granderson | 145 | 18.3 | 690 | 673 | 2858 | .246 | .333 | .484 | .24 | Yes |

Nick Swisher | 127 | 15.3 | 669 | 743 | 3135 | .263 | .362 | .470 | .21 | Yes |

Ray Lankford | 123 | 23.9 | 629 | 690 | 2856 | .288 | .382 | .523 | .22 | Yes |

Jayson Werth | 115 | 17.1 | 655 | 693 | 2763 | .270 | .368 | .478 | .24 | 108 games |

Jeromy Burnitz | 163 | 15.7 | 664 | 759 | 3185 | .259 | .363 | .511 | .21 | yes |

Mike Cameron | 110 | 17.6 | 696 | 671 | 2790 | .251 | .340 | .460 | .25 | yes |

Jim Edmonds | 130 | 24.3 | 596 | 655 | 2719 | .299 | .396 | .540 | .22 | yes |

How did this group age from 33-36? Not well. The following chart shows each player’s production by bWAR, in their age-33 season and beyond.

Player | 5 yr WAR avg 28-32 | 33 | 34 | 35 | 36 | 37+ |
---|---|---|---|---|---|---|

Alfonso Soriano | 3.2 | -1.6 | 0.5 | -0.2 | 2.0 | 2.4 |

Dale Murphy | 4.78 | 1.6 | 0.9 | 0.8 | -0.7 | -0.9 |

Gorman Thomas | 3.26 | -0.8 | 1.7 | -0.3 | -0.5 | |

Curtis Granderson | 3.66 | |||||

Jayson Werth | 3.42 | 0.7 | 4.8 | NA | NA | |

Jeromy Burnitz | 3.14 | 0.0 | 0.2 | 1.3 | 0.2 | -0.6 |

Mike Cameron | 3.52 | 4.3 | 3.2 | 3.1 | 3.0 | -0.1 |

Jim Edmonds | 4.86 | 6.0 | 7.1 | 4.4 | 1.3 | 2.2 |

Ray Lankford | 4.78 | 0.9 | 1.1 | -0.4 | 0 | 0.4 |

Graphically, their production looks like this: I normalized all of the players’ age 28-32 peak to one. Each year thereafter the player’s production by bWAR is illustrated as a single point as a fraction of their peak years.

After their age-32 seasons, only three players had a single season better than their average of their peak seasons of 28-32: Jayson Werth, Jim Edmonds and Mike Cameron. Edmonds was the only player to have as many as two seasons at age 33 and older that equalled his peak years.

What’s remarkable to me is how essentially similar all of the aging patters are, outside of those three. Moving from age 28-32, every other player lost over half of his value. Expressed as percent of their age 28-32 value by age 33 this group was worth 40% of their peak, by age 34 – 65%, age 35 – 33% and age 36 – 20%.

Fangraphs author Eno Sarris suggested a slightly different list of comparable players who put up 10+ fWAR between age 30 and 32 and struck out at least 24% of the time. This list includes three outfielders, all of whom are in our original study: Ray Lankford, Jim Edmonds and Mike Cameron. Edmonds and Cameron are the two most successful agers in the group. Lankford’s decline was steep from age 33 onward.

Now, we create specialized aging curves for Granderson using his comparable players. I simply took an average of each group’s production (bWAR) relative to their peak for their age 33 through 36 season. Then, we apply this set of parameters to Granderson’s starting level.

The following chart compares these aging curves with publicly available projections: ZiPS and Oliver.

Granderson Projections | 33 | 34 | 35 | 36 | Total |
---|---|---|---|---|---|

Full Set | 1.3 | 2.0 | 1.0 | 0.6 | 5.0 |

Sarris Comps | 3.1 | 3.2 | 2.0 | 1.2 | 9.5 |

Oliver | 3.0 | 2.6 | 2.1 | 1.5 | 9.2 |

ZiPS | 2.4 | 2.2 | 2.0 | 1.3 | 7.9 |

Graphically, Granderson’s projections for the next four years look like this.

What’s remarkable here is not the difference between the systems, but the convergence. All four estimators suggest that if Granderson ages like his peers, he will be a 2.0 bWAR player and below for both of the final two years of his contract with the New York Mets. It’s also notable that comparing Granderson to the broader range of outfield candidates my “Full Set” pushes his expected aging path down.

Is there a reason to think Granderson will age unusually well? This is your time commenters. Suggest it, and I will investigate it.

**The Decline/The GrandyMan’s Strikeout Problem**

In fact, there is reason to think Granderson will not age particularly well as is already showing signs of decline. Remember, his strikeout rates in his four years as a Yankee have been in order: 22, 24.5, 28.5, 28.2. There are three reasons Granderson struck out more in the last two years:

1. He swung at more pitches outside of the strike zone, he made contact with fewer pitches both

2. Outside the strike zone and 3. inside the strike zone.

If that 28% strikeout rate sounds high, you’re right. Since 1980, there have only been 21 outfielder seasons in which a player qualified for a batting title with a strikeout rate as high as Granderson’s 28.5% in 2012, (when he was healthy!). For these seasons, the average wRC+ was 110, while the mean fWAR for those seasons was **1.62 **and the median was 1.5. There’s certainly some selection bias here: this group of hitters averaged 27.5 homeruns per year. Put another way, teams do not tolerate that many strikeouts unless a batter is putting the ball over the wall regularly.

Because their high strikeout rates keep these players batting averages down, they must contribute offensive value through secondary skills. Among this high-strikeout group of 21 seasons, there are **two** above three fWAR: Adam Dunn’s 2004 (4.9) and Drew Stubbs’ 2010 (3.6). Dunn hit 46 homeruns with a 16% walk rate while Stubbs supplemented his offensive and defensive value by being worth 10 runs, or roughly one win, on the bases. Granderson has never had a 46 homer season, a walk rate above 12.3% or a season worth more than 6.3 runs on the bases since 2009.

Lets go back to those four seasons by our full outfield group age age 33+ that matched or exceeded their production from their peak years: Jayson Werth’s 2013, Jim Edmonds’ 2003 & 2004 and Mike Cameron’s 2006. If those guys could do it, could Granderson?

All four had strikeout rates significantly below Granderson’s in both 2012 and 2013.

Last year, Jayson Werth hit .318/.398/.532 in 129 games for the Nationals with 25 home runs. The 25 homers seems like a reasonable enough estimate for Granderson’s production if he can stay healthy in 2014. Werth walked in 11.3% of his plate appearances, a rate nearly identical to Granderson’s recent seasons. However, Werth struck out in only 19% of his plate appearances. That’s over 9% less than Granderson’s recent seasons. Granderson has one full big league season with a strikeout rate at 19% or lower — 2008 — when he hit .280/.365/.494 with 22 home runs for Detroit.

Jim Edmonds was great in 2003 and 2004. In ’03, he hit .275/.385/.617 with 39 homers, a 14.5% walk rate and a 23.9% strikeout rate and was worth nearly a full win on defense. In ’04, he poked .301/.418/.643 (!) with 42 homers, a 16.5% walk rate and a 24.5% strikeout rate. Those strikeout and walk rate numbers are just not close to Granderson’s 2012 and 2013.

Mike Cameron’s story is the same in 2006, when he hit .268/.355/.482 with 22 homers, an 11% walk rate and a 22.4% strikeout rate for the Padres.

Note that all of these seasons came with strikerate rates well below Granderson’s 2012 but similar to his 2010-2011 period. Can he return to that form?

**Conclusion**

Even with a strikeout rate in the upper 20s, if Granderson can stay healthy, bop 20+ home runs in the next two years and play average- to plus-defense on the corner — all extremely reasonable expectations — he will be roughly a 2-to-2.5 win player, and help the Mets in the next two years. If Granderson can cut his strikeout rate back to 22-24%, his batting average and on-base percentage will rise, and he could be a 3+ win player.

However, my expectation, is that by the **third year** of this contract, Granderson will be a sub-two-win player – a below average regular. The concern here is not strictly a dollar/win calculation. It’s simply the likelihood that Granderson just should not be starting everyday for a winning team.

Perhaps, Granderson, who has long struggled with lefties, will become the long half of platoon for the Mets with a right-handed hitter. The Mets actually have a pair of right-handed hitting outfielders in the upper minors, Puello and **Cory Vaughn, **who have hit lefties hard in their minor league tenure. Both Puello and Vaughn would start at the Major League minimum.

Granderson’s contract potentially says something important about future Mets’ payrolls. The Mets now have $36 and $35 million committed to **David Wright** and Granderson combined in 2016 and 2017. This is inconsistent with running a payroll below $100 million, as the Mets have done the last two seasons. They will need 23 other active players on the roster. Some of them, ideally, will be good. Some very good. Some of them will be expensive or at least, not earning the Major League minimum. In order for the Granderson contract to not hamstring the Mets’ roster construction past even 2014, the team’s payroll will have to rise comfortably above $100 million again so the team can continue to address other areas of weakness. This signing suggests, or even demands, that once again, the Mets will be willing to spend at a level well above their recent seasons.

In some ways this is an odd deal for the Mets. It makes the team better in 2014 by putting a league average player, who can provide insurance for centerfield, out in a corner. It could make them better in 2015. It is extremely unlikely to add a league average player to a corner in 2016 and 2017. There is downside risk all over.

The Mets identified a need in the outfield. They spent to fix it. However, the patch they chose looks like a short-term repair, and one that is already showing signs of ripping.