One day I wondered if Joe Nathan could make the Hall of Fame
So I researched. I conceived a word problem: What is the standard for a Hall of Fame closer and can Joe Nathan meet that standard? I became pregnant with possible solutions: tables that show Nathan’s place in the history of closers, graphs that demonstrate his dominance and analysis of his future prospects. Now, after about two days of hard labor, I give birth to a beautiful post, brimming with potential.
This is Joe Nathan pitching at some point this season, as closer for the Texas Rangers. The Rangers acquired the 37-year-old on the cheap (two years, $14.5 million) last November, hoping he would rebound from an ineffective and injury-laden 2011. 2011 was itself a season in which Nathan tried to prove himself after an injury; he underwent Tommy John surgery in March 2010 that sidelined him for the rest of the year. That was the second time he needed arm surgery in his career—that and his age gave the Minnesota Twins enough pause to buyout his contract for $2 million after the 2011 season.
Nathan’s recent struggles with staying healthy brought an abrupt end to an impressive run of dominance. From 2004 to 2009, Nathan was the second-best closer in the game, behind only the greatest ever, Mariano Rivera. In those six seasons, he threw 418.2 innings, with a 1.87 ERA, .934 WHIP and a 4.32 strikeouts for every walk issued. The steadiness he brought to the back end of the bullpen helped the cash-strapped Twins win the AL Central three times, though they never advanced past the first round of the playoffs. Nathan was, in fact, historically great during his stint with Minnesota.
Using baseball-reference’s Play Index, I set out to compare Nathan to other prolific closers in history. Since 1961, there have been 27 pitchers with at least 700 innings pitched, 250 saves and 80 percent of their appearances in relief. Below is a table, which may be embiggened, that shows these pitchers and several key statistics. As you can see from the nifty color gradient I added, Nathan, despite two poor seasons as a starting pitcher in San Francisco, compares favorably in ERA+, oppenents’ batting average, opponents’ OPS+ (by which the table is sorted) and strikeout-to-walk ratio. His HR% (HR/BF) is below average. Maybe take some time to admire the stats and how my skills with color gradients have improved dramatically.*
So Nathan, by virtue of those six years, looks like one of the best of all time. But six years of stellar performance does not warrant entry to the Hall of Fame, especially for closers. A quick search of baseball-reference’s database shows that only five men who can be classified as closers have been inducted. They are: Dennis Eckersley, Rollie Fingers, Rich “Goose” Gossage, Bruce Sutter and Hoyt Wilhelm. We may safely assume that Rivera will be elected once he is eligible, and if he ever stops playing. Recently retired Trevor Hoffman also seems like a good bet for the Hall. But Hoffman and Rivera are the only two closers to amass more than 600 saves (Nathan sits at 253), and the five who already made it played under much different conditions.
The role and utilization of pitchers has constantly evolved throughout the history of baseball. In the early 1900s, starting pitchers generally threw every other day and stayed in the whole game. The modern starting rotation of five pitchers did not develop until quite recently. This reduction in pitcher use and corresponding increase in number of pitchers can be traced to two causes: greater anatomical knowledge afforded by advances in medicine and increases in player salary. As owners poured more money into individual players, ensuring their health became more important to teams and science, for the most part, stated that pitchers ought to be taken out of games sooner and given ample time to rest between outings. Bullpens expanded to include more pitchers; specific roles like closer developed for ease of managing all the arms.
In the 1970s and early 1980s, closers like Fingers and Gossage pitched more than one inning every time they were out on the mound. It wasn’t until around 1987 that the modern one-inning closer arrived. Tony La Russa (henceforth referred to as Tony the Russian in this blog), then-manager of the Oakland A’s, started using Eckersley for only one inning at a time, playing all kinds of matchups—based on handedness, pitcher-batter history, etc.—in the innings preceding. Since then, relievers have all been some kind of specialist: LOOGY (lefty specialist), long relief, groundball, closer, what have you.
My point is, it’s impossible to compare Nathan to any of the five closers currently in the Hall of Fame. Eckersley was a modern closer for 10 years, yes, but he was a multiple-inning closer before that and a damn good starter even before that; he was inducted for his entire body of work. Fingers and Gossage pitched in the later days of multiple-inning guys, Sutter and Wilhelm before them. Lee Smith, who earned 50% of the vote on last year’s Hall of Fame ballot, straddled the two eras, but was mostly past his prime as a one-inning guy, so even he isn’t a fair match. The best comparison we can draw is between Nathan, Rivera, Hoffman and Billy Wagner, another closer on the above table who is a borderline candidate.
So I did that. I compared those four closers in four categories: fWAR (Fangraph’s Wins Above Replacement), bWAR (baseball-reference’s WAR—obviously they have different formulas), WPA (Win Probability Added) and WPA/LI (WPA over Leverage Index).
WAR is an all-in-one stat designed to identify a player’s worth in marginal wins compared to a replacement player who can be acquired easily and cheaply. It is generally calculated of runs produced for hitters and runs saved or prevented for pitchers.
Win Probability added is more complex but helpful for closers. Since tens (hundreds?) of thousands of games have been logged by official and unofficial scorekeepers, it is possible to determine how likely a team is of winning the game at any given situation. This is done by finding all of the games that shared that exact same situation (e.g. bottom of the ninth, away team up one, home team has a man on first with one out) and seeing how often the teams in that situation went on to win that game. The percentage of times those teams won is the Win Expectancy and is expressed in decimal form. A player adds Win Probability by doing something that leaves the team in a better position to win the game; he can also subtract it by fucking up. At the end of the season, a player’s WPA is calculated by summing up the WPA of all his plays, positive and negative.
Leverage Index (LI) is a stat that measures the importance of the events a player took part in. Obviously pitching when your team is down 10 runs isn’t as important as pitching when they are down only one. LI quantifies that difference. Different players, through no choice of their own, will be placed in situations with different levels of importance. By dividing WPA by LI, we can neutralize the effects high-leverage situations can have on a player’s WPA. Closers generally enter the game in a narrow set of situations: ninth inning, their team is up either one, two or three runs. Therefore the LI for all of them should be at around the same level. I included this stat only because it would be interesting to compare it to plain WPA. I don’t think I’m smart enough to draw any conclusions about possible discrepancies between the two.
Without further ado, here are the graphs. The x-axis is the player’s nth best season (as in, 1 would be their best season, 2 their second, etc.) and the y-axis is one of the four things I just went over. I limited myself to Nathan’s six full years as a closer.
Nathan’s best three seasons are about even with Rivera’s. His best five seasons are better than Hoffman’s. Billy Wagner had one historically great year, but was mostly worse than Nathan.
Baseball-reference’s formula better captures Rivera’s greatness, I think. And it doesn’t even include his dominant postseason play (neither formula does). Wow, he’s good. Anyway, Nathan’s six years compare favorably to both Hoffman and Wagner.
Nathan is closer to the pack in terms of WPA. But considering it’s a pack comprised (composed? HELP ME, GRAMMAR POLICE) entirely of alpha wolves, that’s not so bad. Perhaps the similarity can be partially explained by the similarity of the situations in which they all pitched.
WPA/LI is the least kind statistic to Nathan. Since LI is in the denominator, that may mean he had a higher LI than the other pitchers those seasons. In other words, he could have pitched in more important situations, which for a closer often means a one-run lead in the ninth compared to a two-run lead in the ninth. Now his disadvantage looks like an advantage; I can spin anything!
So basically, Nathan compared well with this class of Hall of Fame candidates, except his resume isn’t as long as theirs. How much longer would be enough to at least warrant consideration? I’d say four good years, as that would get him to 10 and people love round numbers, because fingers and whatnot.
Four good years is a tough task, though, considering Nathan is already 37. He has started this year very well, though, and Texas is a good team that should provide a lot of ninth-inning leads for him, like an Eastern European town providing small children for the giant who lives in the adjacent hills. So far in 2012, Nathan has a SwStr% (I’m not explaining it again) of 15.4 percent (!), his best since 2005. Tommy John surgery is no longer a big deal, pitchers come back from it all the time, sometimes stronger than before.
As long as Nathan can remain effective into his early 40s and stay healthy, he has a legitimate shot. The healthy part might be difficult with his repertoire of pitches. Nathan has always relied on an excellent slider, a pitch that wears on the elbow more than others and puts pitchers at an increased risk of injury. This year he has thrown it 33.6 percent of the time. Pitchers do not fare so well after two Tommy John surgeries, I read that today in the San Francisco Chronicle, so it might behoove Nathan to rely on other pitches more.
It won’t be easy, but I think if Nathan can get to 10 years of great closing, he deserves to be in the Hall. Maybe I will have an official vote by the time he’s on the ballot. (I hope I’m not just a baseball writer though.) If so, he will have my vote.
What do you think, Joe?
*ERA+ = ERA compared to league average and adjusted for park factors, calculated so that 100 is average and higher is better (like an IQ score); OPS+ = opponents’ OPS compared to league average and adjusted for park factors, calculated so that 100 is average, and lower is better for pitchers; HR/BF = home runs divided by total batters faced.