A Brief Note on Lou Brock’s Relatively Low Career WAR Total
Later today, Jay Jaffe will give Lou Brock the longer look his career and place in history deserve, but I felt it was worth delving into a subject that comes up from time to time when baseball analysts discuss the St. Louis stalwart’s accomplishments. Since his passing on Sunday, I’m sure many a modern fan has looked up Brock’s stats page, found his 43.2 career WAR, and opted to either discount WAR as a stat or Brock as a player, along with the writers who voted him into the Hall of Fame on the first try. I would caution against either approach.
Wins Above Replacement is an incredibly useful framework for comparing players to their peers and across eras. It’s also impossible to quantify every aspect of a player’s game, and it gets harder the further we get from the present. Brock presents a rather unique case, one I’ve written about in the past. The outfielder was an above-average batter for a very long time (only 23 batters in the last 70 years have more than Brock’s 11,238 plate appearances and 109 wRC+). He wasn’t just a singles hitter either, as his power was roughly average during the run-starved 1960s, but he’s obviously more well-known for what he did with his legs. Metrics available at the time Brock played serve to diminish his WAR in a manner that left most of his peers unaffected.
To wit, of Brock’s 187 offensive runs above average, 75 were due to stolen bases. Unfortunately, this misses all of the other runs Brock created advancing on batted balls, which would likely give him somewhere between five to 10 wins above his current WAR. His peers have fewer potential issues in this regard because the majority of their runs come from hitting, meaning WAR misses little of their career production, while Brock gets uniquely penalized.
And while defensive metrics are still somewhat noisy today, the Total Zone metrics used for Brock’s era are further removed from the field. Those metrics would have you believe Brock was an above-average fielder in his 20s and then immediately became one of the worst left fielders in history in his 30s. Brock showed little sign of aging as a hitter and was a more prolific basestealer in his 30s, but he dropped nearly 10 wins in defensive WAR alone over his early 30s, two to three times as much as comparable left fielders. You don’t have to give Brock those five to 10 wins on defense, but we should at least accept the possibility that his value in the field is greater than he’s being given credit for in WAR.
Lastly, there’s Brock’s postseason performance. Brock played in three World Series that each went seven games, with the Cardinals winning two of those three championships. His .391/.424/.655 line resulted in a 213 wRC+ over 92 playoff plate appearances. He added 14 steals, getting caught just twice. In Game 1 of the 1967 World Series, Brock stole second base and then scored on two groundouts for the deciding run. In Game 7 of that same series, he stole second and third and scored on a sacrifice fly to increase the Cardinals’ lead. If Brock performed like he did in the World Series over a month of the regular season, it would be worth around two wins. Brock did it in the 21 most important games of his career in events across a five-year period.
Context is important. While we often focus on what a player or a statistic can’t do, there’s more value in understanding what a player or statistic can do. Much of what a player accomplishes on the field is captured by WAR, including much of the career of Lou Brock. The Cardinals’ great was a good offensive hitter, one of the greatest basestealers of all time, and one of the greatest players in World Series history. Much of that value is captured in WAR, but Brock’s unique skillset and postseason heroics are a bit of anomaly when compared to other Hall of Fame-level players. This post isn’t an attempt to sum up Brock’s career or his life — I’ll leave that task to Jay — but it is a reminder that every player deserves a deeper dive than a single number might provide, no matter if it is 43.2, 938, or 3,023.
Craig Edwards can be found on twitter @craigjedwards.
Thank you, Craig!
Yeah his WAR was low, but I think his induction is a lot to do with putting a dent in the modern steals record, and his overall presence on the Cardinals in the 60’s and 70’s
These are all very reasonable points. I’d never thought too much about Brock’s low WAR and just assumed he made the HOF due to steals and hits. However, this context certainly builds a strong case that he was a better player than his WAR suggests.
Now, if that weren’t the case, I’m sure he’d have still made the Hall because of the aforementioned numbers, but it is nice to know that there really is a strong case for him without considering those as the be-all-end-all.
Very useful reminder that WAR is not the be-al, end-all. I (and I’m assuming others here) fall in that trap sometimes too easily, and it’s always important to take the context of each player’s career into account above and beyond their raw WAR total.
WAR is ideally suited for comparing how “majorleaguerish” different players are in any given season. If you’re comparing how “majorleaguerish” players were across different seasons, you have to make additional assumptions and adjustments in using WAR. If you’re trying to compare players on “Hallworthiness”, you really shouldn’t build the bulk of your argument on WAR in the first place.
Things like each era’s coaching and hitting philosophies matter.
A lot of the “slap hitters” of the past didn’t even try to hit differently because they weren’t expected to. Before Ruth, homeruns were essentially accidents and not something hitters actively sought.
Walks were often seen as missed opportunities and sacrifice bunts were practically mandatory in many conditions.
Times change, the game evolves, and yesterday’s productive specialsts would be playing a different game if the came up today. It wasn’t that long ago that pitchers were taught to pound the lower hal of the strike zone and avoid the letters like the plague. Hitters ignored high pitches and umps called ball on pitches tbt might be homeruns if swung at.
Brock almost certainly lost a lot of PAs bunting and advancing runners that never scored, not because he wasn’t a good enough hitter but because that’s what he was told to do. Can’t ding him for being a product of his time.
Excellent point! I just read Jay’s tribute to Brock, and it’s interesting to recall that Brock hit for decent power in his twenties – sort of like Adam Jones with a bunch of extra SB. But right after he turned 30, Busch Stadium installed Astroturf, and Brock swapped power for more contact and ground balls (Baseball-Reference shows his ratio of groundouts to air outs going up noticably once St. Louis went to turf). I’d also have to think that his lower Total Zone defensive numbers had something to do with the turf….either his having to adjust to it or TZ having trouble accounting for it.
Very good point about the turf, and as The Duke’s comment suggested, I think that the trade of Flood may have played a part in the decline of Brock’s defensive numbers as well.
It’s not just that the grass was replaced with turf. The Cardinals moved into a cavernous new stadium in 1966 when Brock was 26 (while the turf was added in 1970 when he was 30). The walls weren’t significantly moved in until 1973 and then moved back in 1977 before permanently being moved in during the ’90’s, long after Brock’s retirement. Likewise, the turf wasn’t removed until 1996.
Technically, you only lose at bats with sacrifice bunts, not plate appearances.
“To wit, of Brock’s 187 offensive runs above average, 75 were due to stolen bases. Unfortunately, this misses all of the other runs Brock created advancing on batted balls, which would likely give him somewhere between five to 10 wins above his current WAR.”
I’m a little confused on this. What is it about older WAR calculations that causes this to be missed? I understand how the old defense metrics have huge error bars, but I’m less clear on how baserunning-runs-not-counting-steals are missed using old statistics.
I think the data aren’t robust enough when you go back that far. You can’t tell how often a player goes first-to-third on a single, for instance. At some point, all those data are included and thus you can include it in baserunning WAR. Prior to that point, it’s just SB and CS. That’s my understanding.
If you take a look at this leaderboard from 2002 on, it does wSB as well as runs on dp and ubr on batted balls, which aren’t included before 2002. Look at SB runs and then at BsR. Brock gets only SB, while most players have close to doubled their totals for BsR. https://www.fangraphs.com/leaders.aspx?pos=all&stats=bat&lg=all&qual=y&type=c,6,34,35,36,-1,23,37,38,39,-1,40,60,41,-1,201,205,200,-1,52,51,50,61,111&season=2020&month=0&season1=2002&ind=0&team=0&rost=0&age=0&filter=&players=0&startdate=2002-01-01&enddate=2020-12-31&sort=16,d
The actual gaps in the data used to calculate WAR in the past (and pre-2002 doesn’t seem like “the past” to a lot of people) aren’t widely known and are easily forgotten. This probably should be at least noted in any piece that dips into earlier eras.*
*(Get it, DIPS into earlier ERAs?)
The BsR statistic is contributions/demerits from SB/CS, advancing / not advancing on the basepaths (e.g. 1st to 3rd on a single, scoring from 2nd on a single, 1st to home on 2B), and grounding into double plays.
The problem is that data for the second of those contributions, advancing or not, is only available since 2002. So all the times Brock took an extra base don’t count toward his total. Brock is semi-special in this regard, as you need to be an elite baserunner for this missing contribution to be sizeable.
The effect of these missing credits can be seen, for example, in Johnny Damon having the same career BsR as Brock. It also helps explain (I believe) how Damon’s BsR became much better starting in 2002, and being higher in his 30s than his 20s.
This was my question too, thank you
Baseball Reference accounts for taking the extra base back to 1953. Not sure why people are saying the data aren’t available prior to 2002, it’s just play by play data from Retrosheet.
The data available doesn’t confirm Craig or others assumptions. Accessing baseball dot tomthress dot com, or Tom Thress Win-Loss Records, which uses said Retrosheet PBP data. he was an excellent 6 wins above average in basestealing, but just 1.3 in other facets of running, his baserunning advancement was a net zero with his baserunning outs.
I’m also not seeing hidden value in his defensive game, he was a capable if not excellent fielder through 1968 / age 29, then is average or poor for the remainder of his career. Defense peaks early, so seeing someone decline in the 30s is pretty natural. DRA, TZ, T-T, FRAA all are in agreement on this front.
Lou can be in the HOF for his SB record, 3000 hits, and his post-season exploits, but his value numbers suggest he’s a HOF mistake.
Well, a natural defensive decline is still significantly more valuable than the way this article says he supposedly fell off a defensive cliff and before long ultimately wound up becoming a terrible defender. Such a thing shouldn’t happen to someone who was originally a good defender unless there’s a chronic leg or foot issue at play (like with Albert Pujols), and if that were the case it should’ve also affected his baserunning.
Good points generally, but I’m confused about some of the specifics. You write: “Unfortunately, this misses all of the other runs Brock created advancing on batted balls, which would likely give him somewhere between five to 10 wins above his current WAR. ”
Both Fangraphs and Baseball Reference include baserunning in WAR. Does this number not account for taking extra bases? And if it doesn’t, how did you arrive at 5 to 10 wins? Not disagreeing, just wondering.
“Those metrics would have you believe Brock was an above-average fielder in his 20s and then immediately became one of the worst left fielders in history in his 30s. ”
He was worth about -7 defensive runs in LF from age 31 on. That’s not the worst in history and although a little worse than you’d expect for someone who was a good OF, does it really qualify as one of the worst in history? Your point about his baserunning being excellent is a good one and makes me question the data. Also the fact that he was so bad in 71 and 72, and that every single year had him negative. I wonder how people at the time viewed his defense.
In his 20s he had curt flood covering cf. in his 30s he didn’t, so it’s possible Flood may have allowed him to play a better LF.
I thought he was one one of the worst full time defensive left fielders I have seen ( and I’m a huge Lou Brock fan). To me it’s more likely the early defensive numbers are too good, not that the later ones are too bad.
Brock was a menace on the base paths and I’m sure his presence at 1B helped the hitters behind him. I’m not sure if that can be quantified
Before 2002, the data for baserunning is all SB. And sometimes doesn’t include CS.
The 5-10 WAR is arrived at because modern players get about half their BsR value from steals, about half from the stuff that wasn’t measured in Brock’s day.
I think the real problem is that he didn’t get on base at a league-leading rate. The guy wasn’t even on the level of Bobby Grich or Lou Whitaker, much less Joe Morgan or Rod Carew.
The baserunning thing is interesting and important (less convinced about the defensive metrics part) but the OBP is what enables the baserunning to matter. That’s the real reason his career WAR is so low–5-10 extra wins still doesn’t really do it for him.
Okay, all that said, I do understand why fans feel really passionately about him. Any guy who steals that many bases must be huge fun to watch. I still remember Vince Coleman, who was essentially a version of Lou Brock who wasn’t good, but he was incredible every time he got on the basepaths.
This is a great and timely article. I definitely fell into the category of people who presumed Brock was overrated because of his relatively modest WAR totals. I figured he was in the Hall of Fame based largely on his astonishing one-skill prowess and World Series performance, and that perhaps he would have been more valuable if players in his time had better understood how to deploy small skills and take high-percentage risks (the kind of play that makes Mike Trout, Joe Morgan, and 1997 Craig Biggio look much better than the sum of their parts).
It really helps to know what modern metrics miss when looking at past performance.
Well said!
Albeit another cardinals puff piece
I think this makes a good case that Brock is undervalued by WAR. I also think it’s not necessary to make that case to resolve the conflict.
It could be that Brock was, in terms of wins added, not as valuable as the typical hall-of-famer. However, I’m totally fine with him being rewarded for excelling at what, at the time, was considered very valuable. We give medals to olympic athletes who, sometimes, display skills that have no practical purpose anymore.
Brock clearly “feels like” a hall-of-famer, and given that the hall is subjective anyway, why not vote in a guy who held the all-time record in a stat a lot of people care about?
Still better than Harold Baines…
Out of curiosity, do any other players come to immediately to mind when it comes to a lowish career WAR total not fully painting the picture? Not the obvious ones like players whose peaks were cut short by injury, and probably not catchers either (since that seems like pretty well worn territory). I wonder if there are other Brock-types with a ton of WAR that is likely hidden by limitations of the past.
It seems to me that many analyst are trying harder than is perhaps necessary to avoid context as a relevant issue in using WAR. Yet there it is staring us in the face.
How do you turn context into valid numbers?
There’s only so much data to work with and some is suspect.
As they say, “The past is a different country.”
Using WAR in a historic context is a joke. It should never be used to compare players from differing eras to each other. Its only worth is comparing similar players from the same era to each other. The bias from the current WAR formula affects players from previous areas disproportionately. You are essentially asking, how much does this historic player fit the model of how we value a player at this moment in history? This is a pretty textbook definition of the term “arbitrary”. Different equations will value him differently. In a few years he could either more or less valuable. He is literally dead and his value is changing. His defensive value may as well be a randomly generated number.
Here’s my even briefer note: Defense (aside from catcher) is significantly overrated by WAR. If the weights between Offensive and Defensive WAR were properly adjusted, then Brock’s career WAR total would look much more like you’d expect.
This. Moreover, WAR is not the same as “doing things that are useful for winning baseball games”. I love WAR and use it all the time, but it has a number of blind spots. For example, in a low run-scoring environment SBs (and basically all baserunning plays) are more valuable for winning actual baseball games than they are in a high run-scoring environment, but BSR metrics don’t adjust the weights accordingly.
Check out the full categorical breakdown at Tom – Thress – Win – Loss – Records, the defensive spectrum way be too wide as suggested, but his offensive contributions weren’t enough to make him valuewise HOF worthy.
Others have touched on it, but I’ll say it more bluntly. Brock’s low WAR is due to his terrible defense. How could someone with all that speed and playing LF (not CF) be so terrible defensively? He should be getting to every ball hit to LF and fighting with the CF for balls all the time. How is that possible?
Except his defense wasn’t rated as terrible, it was good to excellent in his youth and below average in his 30s, average for a career. He had great range when young, then average in his 30s, he did average in baserunner/advancement kills. He’s hurt by WAR by the outs he produced at the plate and on the bases.
Brock was not a good outfielder. He couldn’t throw, didn’t have great instincts off the ball, and didn’t play enough in the minors to develop a feel for any outfield position. To make things worse, the Cubs thought he was a center fielder merely because he was fast. The Cardinals, to their credit, knew that Brock needed to focus on what he COULD do, and let him do it, already having the great Curt Flood in center. But let’s not rewrite things–Brock was NOT good in the outfield. Ask anyone who saw him.
So another methodology question, and it isn’t just about Brock; also about all the other speed burners of his general time (Wills, Joe Morgan, Bert Campaneris, and so on). Is a stolen base during an era of grossly depressed slugging, as it was for much of Brock’s career, treated equally in terms of contributions to WAR to a stolen base in a more homeriffic era, like, say, today’s? And if so, isn’t that kinda, well, dumb?
This is something I’ve wondered about for years, based on analogies between the process of run creation and the way chemical reactions occur (I’m a chemist). Run creation in a power-happy time is, to a large although not overpowering degree, what a chemist would recognize as a two-step process: guy gets on base, another guy drives him in with a long ball. Each of these things has, to adequate precision, a 30% chance of happening. By contrast, if there are no extra-base hits, run production is a three-step process: guy reaches first, next guy advances him into scoring position, third guy drives him in. It doesn’t take much erudition to figure out that if each of these steps has a 30% chance of happening, run production will be suppressed relative to the two-step process. BUT, if the process of moving into scoring position can happen with an 80% probability, as it can with an elite base stealer with a green light, it’s back almost to the two-step “reaction”.
Obviously this is an oversimplification. Not all runs (by any means) in the modern era are due to someone getting an extra-base hit; not all runs in the 1960s and 1970s came in the station-to-station way. Clearly, though, the relative proportion of three-step runs compared to two-step runs was higher then than it is now. With that in mind, it seems equally clear to me that looking at an elite base stealer from that time through today’s lens is likely to undervalue what he was doing. Or is this adjustment being made, in ways not obvious from the descriptions of BsR, WAR, etc.?