Since the Wins Above Replacement (WAR) statistic was developed, folks seem to either detest it or love it, with very few landing somewhere in between. I’m one of the tweeners. While, overall, I think WAR is a really valuable tool to evaluate players, particularly players in the same era or playing the same position, I also think it is just another imperfect statistic that shouldn’t be viewed as gospel.
All baseball statistics are flawed to some degree. Hits are often judgment calls of the official scorekeeper. So are errors. That makes earned runs, and therefore earned run average, a bit subjective, too. The same for fielding percentage, since it’s based on the number of errors a player commits. And so on.
Some stats are flawed because they’re incomplete. For instance, sacrifice flies don’t count against a player’s batting average, but they weren’t tracked as an official statistic until 1954. We don’t have reliable data for the number of times player were caught trying to steal before 1920. Sliding Billy Hamilton has the 3rd-most steals in the history of the sport, but all of them took place from 1888 to 1901, so we don’t know if he was 914-for-914 or if he was thrown out more than he was successful. Something can be found that’s a bit off with most stats.
The same is true of WAR. It’s based on many of those flawed stats, so it’s flawed, too. On top of that, WAR is a complicated calculation that isn’t agreed upon by everyone. The folks at Baseball-Reference.com calculate it one way, and the folks Fangraphs calculate it another way, and, while they typically come up with very similar results, there can be wide differences between them, particularly regarding pitchers and catchers.
Still, regardless of the flaw and quirks, as long as you know them going in WAR is a really valuable measurement. It’s one of the few statistics that attempts to provide a player with an overall measurement for everything he does on the field. That’s impossible to measure precisely, of course, because completely accurate weights can’t be given for data that isn’t collected, or has subjectivity baked into it, but it’s still nice to have one measure that’s an approximation of a player’s overall career or seasonal value.
Where people get in trouble with WAR is in misapplying it. Here is an example of that:
TheWARMonger has an interesting account on Twitter. He posts comparisons like this, where all of the “black ink” offensive statistics favor one of the two players, but WAR favors the other, and then he makes snide comments about how WAR is broken. Often he’s right, for some of the reasons I’ve already stated, and there’s a lot of value in pointing out that WAR is imperfect and should not be used as an absolute measure of a player’s career. That happens too often with the group I mentioned who fell in love with WAR, including many baseball writers who vote for the Hall of Fame, and it’s a sloppy practice that should be criticized. WARMonger does that.
But…
Some of WARMonger’s examples aren’t great, and the one between Dusty Baker and Jim Fregosi is a good example of that. Part of that is in the graphic he uses in his posts, which comes from Stathead.com. It’s the database search tool for the various Sports Reference sites, including Baseball-Reference.com. It’s the same graphic tool I often use here when posting comparisons of players.
For this graphic, virtually all of the career data displayed is for offense only. It doesn’t display any defensive data, or full baserunning data, or positional adjustments, or era adjustments, or ballpark adjustments, and all of those are components of the WAR calculation. That leads to examples like this one, where all of the offense appears to favor Baker, but WAR favors Fregosi, and we’re left scratching our heads and shaking our fists at the folks who invented WAR, and sort of dismiss it all as a crock.
That would be really unfortunate in cases like this one, because this is one of the times that WAR did exactly what it was supposed to do. Namely, it allows us to see that Jim Fregosi was probably a better overall baseball player than Dusty Baker, something we wouldn’t necessarily see if we just looked at the back of their respective baseball cards.
I’ll demonstrate that, and I’ll use nothing but the same common sense most longtime baseball fans have and some different tools from Stathead to do that.
Let’s look at Baker and Fregosi side-by-side, season-by-season, and see which one we think provided more value. Luckily, each of them first reach the big leagues when they were nineteen years old, so we can compare their overall careers arcs pretty easily.
Age 19
Baker got into six games with the Braves in 1968 and got two hits in five plate appearances. Fregosi got into eleven games for the Angels in 1961, and got six hits in 27 plate appearances. Baker was mostly used as a pinch-hitter, but also played some center field, while Fregosi was strictly a shortstop. Not much value added by either, let’s call it a wash. What does WAR say?
Fregosi: 0.0 WAR
Baker: 0.0 WAR
So far, so good.
Age 20
Baker barely played in the big leagues in 1969. He got into three games and had seven plate appearances. He was hitless and struck out three times. In the field, he played both center field and left field. Fregosi, on the other hand, got a fair amount of playing time for the Angels in 1962. In 58 games, almost entirely as a shortstop, he batted .291 in nearly 200 plate appearances. His OPS was .761, which was better than the league average of .719. He struggled a bit defensively, with -6 Fielding Runs. Between the two of them, you’d have to say that Fregosi was worth more to the Angels than Baker was to the Braves, right?
Fregosi: 0.3 WAR
Baker: -0.2 WAR
That’s fits, too. Baker was hitless in his brief time in the big leagues, so he’s very slightly negative in terms of the “wins” he brought to the team. Fregosi, though, brought the Angels some value. That was particularly true offensively, given the fact that only two major league shortstops. Tom Tresh and Woodie Held, had better OPS marks than Fregosi in 1962. Getting offense from a position where it’s rare carries a lot of extra value.
Age 21
This was Fregosi’s first full season in the big leagues, and he played really well. He batted .287, had an on-base percentage of .325, and slugged .422 (also known as a slash line of .287/.325/.422). The league overall hit just .247/.312/.380. This was even more rare among shortstops. The only regular shortstops in the majors with a better OPS than Fregosi were Dick Groat and Eddie Bressoud, but Bressoud did that while playing half of his games in Fenway Park, a much easier park to hit in than Dodger Stadium, which is where the Angels played in 1963. That means Bressoud and Fregosi had the same adjusted OPS score (or OPS+), 114, despite Bressoud’s unadjusted OPS being higher because OPS+ accounts for home ballparks. Additionally, Fregosi’s defense improved, and he was slightly into positive numbers (+1) in Fielding Runs.
Baker, meanwhile, still barely played for the Braves. He got into just 13 games and came to the plate just 27 times. He batted .292, but all of his hits were singles, so his batting line was just .292/.333/.292, good for a below average OPS+ of just 66. Obviously, baseball fans would view Fregosi as having been far more valuable to the Angels than Baker was the the Braves. The writers agree with us, as Fregosi finished 23rd in MVP voting. Baker, of course, didn’t get any votes. WAR agrees with us, too
Fregosi: 4.3 WAR
Baker: -0.1 WAR
Age 22
Baker still couldn’t crack Atlanta’s roster for most of the year. He had only 64 plate appearances and hit terribly, just .226/.238/.258. Fregosi, though, blossomed into a star. He made his first All-Star team, and batted .277/.369/.463 at a time when the league was hitting .247/.315/.382. He did this, again, in Dodger Stadium, a horrible park for hitters. He set new career highs in runs, homers, and walks, and cut his strikeout total by over 15%. He also played exceptional defense, posting 16 Fielding Runs, and finished 13th in MVP voting. Clearly, Fregosi was more valuable, and WAR demonstrates this:
Fregosi: 7.9 WAR
Baker: -0.7 WAR
Age 23
Finally we’re going to have a somewhat fair fight, because Baker managed to crack the Braves lineup regularly in 1972, when he was twenty-three. Playing mostly center field, he batted a robust .321/.383/.504 at a time when the league batted .248/.315/.365. His OPS+ was an exceptional 142, and it would prove to be the highest mark of his career. His defense was respectable as well. Fregosi had another very solid year. His hitting regressed a bit to .277/.337/.407, but that was still a 114 OPS+ at a time when only three other shortstops were hitting like that. Likewise, very few center fielders could touch Baker’s 142 OPS+ mark, also just three others, in fact.
Overall, this looks like it’s a very close call, doesn’t it? Both players hit at a level only three of their positional peers could match. Fregosi did it at the harder position, but center field is still a critical defensive spot. MVP voters viewed them similarly, voting Baker 22nd in 1972 and Fregosi 23rd in 1965. What does WAR say?
Fregosi: 5.3 WAR
Baker: 5.1 WAR
Well, well, well. That result aligns pretty much perfectly with what our baseball common sense would tell us, doesn’t it? It matched the MVP voting pretty closely, too.
Age 24
Fregosi kept chugging along in 1966 with a 108 OPS+ mark that only two shortstops exceeded. He still played solid defense, too. Meanwhile, Baker regressed in 1973 from his performance the previous year. His OPS+ was still solid at 117, but eight center fielders did better than that. Also, his defense regressed slightly. Overall, Fregosi’s performance as a shortstop was more rare, and therefore probably more valuable, than Baker’s as a center fielder. The voters agreed, with Fregosi finishing 28th in the AL MVP voting while Baker received no votes. Fregosi also made the All-Star team again. Baker did not. WAR agrees with all of this.
Fregosi: 5.6 WAR
Baker: 4.0 WAR
Age 25
Fregosi had a strong year offensively, with a 125 OPS+ that was the best in baseball among shortstops, by far. Only two other shortstops even cracked a league-average 100 OPS+ mark. He also made another All-Star team and won a Gold Glove, though it was largely based on prior performance as his defensive total was a good-but-not-notable +2 Fielding Runs. He finished 7th in MVP voting. Baker’s defense had slipped enough that he was moved to right field in 1974. He hitting was still good, but slipped again to a 108 OPS+, a mark 17 other right fielders exceeded. He got no MVP votes and didn’t make the All-Star team. A pretty clear win for Fregosi. WAR backs this up.
Fregosi: 5.4 WAR
Baker: 3.5 WAR
Age 26
Another All-Star team for Fregosi, plus he led the league in triples while posting a 111 OPS+ in 1968 that only Bert Campaneris exceeded among shortstops. He didn’t play as well in the field, but still finished 15th in MVP voting. Baker’s season was also largely unchanged from the year before. He had a 110 OPS+; Sixteen other right fielders were better. His defense in right field slipped as well. No All-Star team for him, and no MVP votes either. Once again, WAR mirrors what we’d expect.
Fregosi: 3.9 WAR
Baker: 1.7 WAR
Age 27
More of the same for Fregosi, only with better defense than the year before. His OPS+ of 114 was beaten only by Rio Petrocelli and Denis Menke among shortstops. He made another All-Star team and was 20th in MVP voting. Baker had been traded to the Dodgers after the 1975 season, and had a pretty bad year in 1976. Moved back to center field, he did pretty well there defensively, but his offense was really bad. His 74 OPS+ was bettered by 25 other center fielders. No surprise, his WAR total tanked.
Fregosi: 4.9 WAR
Baker: 0.1 WAR
Age 28
This was Fregosi’s final healthy year. It was also his last good year. His 127 OPS+ was tied with Menke as the best mark among shortstops in baseball, and he had an excellent year defensively, totaling +14 Fielding Runs. He made his final All-Star team and finished 12th in MVP voting. Baker, now moved to left field, rebounded from his terrible 1976 season and was excellent in 1977. His OPS+ was 134, fifth-best in baseball among left fielders. He didn’t make the All-Star team or receive any MVP votes, but it was still a good year, just nowhere close to as good as Fregosi’s. We see that in WAR.
Fregosi: 7.7 WAR
Baker: 3.5 WAR
Okay, before we get into the flip side of their careers, when Fregosi was hurt or struggling and Baker was playing some of his best ball, let’s summarize where they stood to this point in their careers.
Fregosi had been to 6 All-Star Games, and received MVP votes in 6 seasons. He’d won a Gold Glove, and led the league in triples once. Baker had one really strong year when he got MVP votes, but otherwise was shut out of that voting, never made an All-Star team, was up and down both offensively and defensively, changed positions a couple of times, and got traded. No one in baseball would have considered Baker to be anywhere near Fregosi’s level. If the Braves or Dodgers had offered the Angels a straight-up trade of Baker for Fregosi, they’d have been laughed at.
The WAR totals match everything the All-Star voters (a.k.a. baseball fans or managers) and MVP voters (a.k.a. beat writers covering the sport for a living) were seeing at the time. Through their age-28 seasons, their respective WAR totals were 45.2 for Fregosi and 16.9 for Baker. This doesn’t look out of line in any way, does it? We can quibble about positional weights or park adjustments and play with the math if we feel like it, but after all or part of ten seasons for each of them, there was a gap of nearly 3 wins per year in Fregosi’s favor that absolutely matched public perception, common sense, and available statistics. No matter what happens from this point forward, closing a gap of nearly 30 WAR while moving into his thirties as a ballplayer was going to be a really tough task for Baker.
And yet Baker came pretty close to closing it anyway. Partly that’s because Fregosi had a tumor removed from his foot early in 1971 and was never the same. It was a damn shame, because he was tracking toward a potential Hall of Fame career to that point. Within a year of the surgery he was moved off shortstop. His OPS+ from 1971 to his retirement in 1978 was exactly league average at 100. That was in part-time duty at first base or DH for the most part. His cumulative WAR total for those years was only 3.6. Just a brutal turn to a great career.
Meanwhile, Dusty Baker still had a lot of really good baseball to play. Settled in left field finally, he played pretty well there and even won a Gold Glove in 1981. He made a couple of All-Star teams, and got MVP votes in both 1980 and 1981. In six seasons, from 1978 to 1983, he nearly doubled his career WAR total, adding 16.5 WAR to give him 33.4 for his career. Though he was done as a full-time player when he was only thirty-four, he still added a bit of value in part-time roles for three seasons after that, tacking on another 3.7 WAR in nearly 1,000 plate appearances as he closed out his career.
It wasn’t enough to catch Fregosi, though, and it’s not a mystery or a conspiracy or a problem with the calculation of the WAR stat that tells us that. By the time Baker reached his stride with the Dodgers and Fregosi was declining, he was largely a one-dimensional corner outfielder. He hit well but not exceptionally, he played one of the easier positions and generally not well by that point, and those overall skills weren’t that unique. Left fielders who can post an OPS that’s 16 or 17 percent above the league average simply aren’t that rare. In fact, as we saw in the year-by-year breakdowns above, that’s kind of average for corner outfielders. Here’s how he compared in his final six full years:
1978: Baker has a 97 OPS+, but 20 left fielders were better.
1979: Baker has a 117 OPS+, but 9 left fielders were better.
1980: Baker has a 134 OPS+, but 4 left fielders were better.
1981: Baker has a 132 OPS+, but 6 left fielders were better.
1982: Baker has a 130 OPS+, but 5 left fielders were better.
1983: Baker has a 107 OPS+, but 21 left fielders were better.
On balance, Baker was in the neighborhood of being the 10th- or 12th-best-hitting left fielder in baseball at a time when there were 26 teams. Being slightly better than the midpoint of your peers offensively while being just sort of okay defensively simply isn’t as valuable as being a top-3 or 4 player and solid defender at a more difficult position like Fregosi was. Barring remarkable longevity in his career, which he didn’t have, Baker really didn’t have any hope of eliminating that big gap between himself and Fregosi.
So, WARMonger asked who had the better playing career. Obviously Fregosi did. It wasn’t as long as Baker’s, but it was better.
I like WARMonger’s account, and I agree with many of his examples of WAR having funky results that make us question some of its features. There’s value in reminding people of that.
I just hope he picks better examples to illustrate that in the future.
Excellent analysis!