Even great players see their skills and production decline as they approach the end of their careers. How much of a decline is more open to debate and may depend on how productivity is measured. Ernie Banks’ 1969 season provides a good case in point. Traditional stats, especially RBIs suggested he was still quite productive. More modern, supposedly advanced stats, like Wins Above Replacement (WAR) tell a quite different story. A closer look at the numbers and their context suggests limitations to both approaches.
Ernie Banks was 38 in 1969 and was clearly past his prime. But he still put up more than respectable numbers with 23 home runs. He hit only .253, but that was still above the league average. His 106 RBIs were good for fifth in the league. He finished 12th in MVP voting as the Cubs had their winningest season since 1945.
Banks’ WAR, on the other hand, was -0.7 according to Baseballreference.com. He comes out a little better on FanGraphs (+0.1), but not much. It’s worth remembering that a zero WAR doesn’t mean an average player but rather a borderline major leaguer. Consider the expansion Montreal Expos, which lost 110 games that season. The team had a net WAR of +13.8. The San Diego Padres also lost 110 games, and their WAR was still positive at +5.9. Even the 2024 Chicago White Sox , which set a Major League record with 121 losses, had a net WAR of 7.0. Overall, Banks’ WAR that year placed him in a four-way tie for 402nd of 422 National Leaguers that year. A contrast between Banks’ traditional stats and WAR is shown below:
Why was Banks’ WAR so Low?
Players with good Triple Crown statistics usually also have a high WAR. But not always. WAR consider other elements of a player’s performance that the Triple Crown statistics fail to capture. It also makes various adjustments to facilitate comparisons across teams, leagues, and even eras. Many of these adjustments would disfavor a player like Banks at that stage of his career.
Teams expect a lot of production out of first basemen. Banks started his career at a shortstop where you often expect a light hitter who plays good defense. He put up some amazing offensive numbers for a shortstop, averaging 41 HR from 1955 through 1960. Even hitting 30 homers was, and still is, rare for a shortstop. He converted to first base later in his career, where you expect a power hitter. This was especially true in the days before the Designated Hitter, when teams most often stuck hard-hitting defensive liabilities at first base.
Playing in Wrigley Field boosted Banks’ offensive numbers – by a lot. Banks hit .278 with 15 HRs at home but only .226 with 8 homers on the road. There was a similar, though less extreme difference over the course of his career, as Banks hit 31 points higher at home. Wrigley Field is a good hitter’s park.
Banks didn’t get a lot of walks. He had only 42 walks over 155 games. WAR weighs a walk the same as a single, with on-base percentage more important than batting average. Banks’ OBP was only .309, below the team average of .323 and the league average of .319.
Why WAR Might be Underrating Banks
While the use of WAR and other advanced statistics can provide a more complete picture of a player’s performance, it has its limitations and blind spots. That’s especially true, if like WAR, it tries to summarize a player’s entire performance into a single number.
RBIs are important after all. RBIs don’t have the cache they once did. But the object of the offense is to drive in and score runs. That makes Banks’ impressive RBI totals hard to square with his negative WAR. The RBI totals look at least as good in context than in isolation. Batting stats weren’t inflated in 1969, certainly not to the extent we saw at the height of the steroid era. The Cubs were a good hitting team, but none of the starters hit above .300. They had little team speed with only 30 total stolen bases across the entire team. While Banks only scored 60 runs, his Runs Produced (Runs Scored + RBIs – HRs) still placed him third on the team, behind only Ron Santo and Billy Williams, future Hall of Famers at the height of their careers.
A hit and a walk aren’t equivalent. A walk can be just as good as a single, but not necessarily. If you a have a runner on second or third, he can score from a single and not from a walk. Drawing walks can be great, especially for those near the top of the order, with strong hitters behind them. Banks batted fifth, making walks less useful.
Batting order matters. It’s not merely a matter that batting fifth is treated as an RBI position in the lineup. Consider who was hitting behind Banks in 1969. Randy Hundley, Jim Hickman, and Don Young batted behind Banks and hit .255, .237, and .239 respectively. The Cubs had a great pitching staff that year but none of the starters were decent hitters. Given all that, there was more of an incentive to put the ball in play.
Defensive WAR isn’t very intuitive for first basemen. Banks had an offensive WAR of -0.1 and a defensive WAR of -1.7. (While both elements contribute to the WAR, the overall number is not an arithmetic average of the two.) Banks’ dWAR suggests he was a huge defensive liability in 1969 but that isn’t at all obvious from his fielding stats. He led NL first basemen in fielding percentage and finished second in range factor per 9 innings. He led in putouts and finished fourth in assists and double plays.
Defensive WAR can be hard to unpack. The top dWARs all-time are about what you’d expect. Ozzie Smith is #1 and Brooks Robinson is #3. It’s a lot less obvious for first basemen. Keith Hernandez may be the best defensive first baseman ever and certainly the best I ever saw. But his career dWAR was only +1.3. Gil Hodges, Steve Garvey, and Don Mattingly won 16 Gold Gloves between them, but each has negative dWARs.
The dWAR calculation has a position adjustment that penalizes first basemen. The NL’s 12 starting first basemen in 1969 had dWARs ranging from -2.1 to 0.4, with an average of -0.9. It was like Lake Wobegon in reverse, where (nearly) all the first basemen were below average fielders. In addition, dWAR compares a player’s performance to the average first basemen, not replacement level. But their relative dWARs were also confounding. Orlando Cepeda had the highest dWAR that year, even though persistent knee problems for much of his career meant his natural defensive position was DH. Meanwhile, slick fielding Wes Parker, who won the Gold Glove every year from 1967 through 1972 had a dWAR of -0.7.
It’s unlikely that a 38-year-old converted shortstop like Banks was the league’s best first baseman. But the -1.6 dWAR doesn’t make a lot of sense. When Cepeda became a full-time DH in 1973, his dWAR was -1.3. It’s hard to argue that a full-time DH contributed more defensively than a full-time first baseman who led the league in fielding percentage.
Clutch performance is important. Banks performed exceptionally well in clutch situations in 1969. While his overall average was .253, he batted .311 with runners in scoring position, and an amazing .340 when there were also two outs. He hit .310 in “late and close” games. There was also a marked difference in his performance in “high” vs. “low” leverage games. Leverage in this sense means how likely a particular at bat will affect the game’s overall outcome. He hit .297 in high leverage situations and only .213 in low leverage situations.
WAR incorporates numerous offensive and defensive statistics but doesn’t capture everything. This is a case of what statisticians refer to as “omitted variable bias.” WAR doesn’t pick up these clutch stats. I’m not saying it should since with any measure there’s a tradeoff between completeness and complexity and it’s best to avoid undue complexity. But the omission explains how a player with a negative WAR can still drive in 106 runs.
Conclusions
Both individual and composite statistics can provide a useful picture of a player’s performance. However, relying on any one metric has some serious pitfalls. I’m pretty sure that Ernie Banks wasn’t the fifth best player in the National League in 1969. I doubt whether he was the twelfth best player that year, either. But I’m also quite certain he was a lot better than the 402nd best.
Leave a Reply