Take a walk, lose some average

In the 1986 Baseball Abstract, Bill James postulated that when a player experienced a sudden spike in walk rate one year, it would often be followed by a substantial drop in batting average the following year.

According to James, taking a base on balls is a “veteran player skill” because it involves an older ballplayer compensating for slower reflexes and bat speed. But, James reasoned, the increase in walks will result in a decrease in batting average because the pitchers will adjust and will begin to make the hitters hit good pitches.

Without doing much research, James mentioned Gary Matthews from the 1985 season and Willie Mays from 1971 as examples that best fit this theory.

But because James presented this theory in the player comments section for Toby Harrah, I call this the Harrah Hypothesis.

The 36-year-old Harrah was the Rangers third baseman in 1985 when he hit .270 and set a career high with a 22.2 percent walk rate, up eight percentage points from the previous year. Throughout his career, Harrah was well versed in the strike zone and he consistently walked more than he struck out. But with such a sharp (and sudden) increase in walk rate, James theorized, Harrah was ripe for a decline.

In this case, James nailed it. The following season, Harrah’s average tumbled to .218—a whopping 52-point decline. He was granted free agency by the Rangers after the season and never played again.

The Harrah Hypothesis worked for the players James named in his Abstract, but does it still apply to players today? Thanks to John Burnson, publisher of the 2008 Graphical Player, we’ve put together a graph that plots the walk rates of all hitters age 30 and older from 2005 and 2006 and their change in batting average the following season. The larger the circle, the greater the change in average from one season to the next.

This is a small sample size, generated to see if there was some merit in the Harrah Hypothesis in today’s game. Basically, we wanted to see if we were on the right track. The first thing that jumps out is the lack of data for players 35 and over—and those who experienced an increase in walk rate and followed that with a slight increase in batting average. But for players 30 to 35 who experienced an increase in walk rates, it looks like might have something. That will be our initial study group.

Here are the players aged 30 to 35 who saw at least a 2.5 percentage point increase in walk rate from 2006 to 2007:

Magglio Ordonez: +4.3 percentage points.
Brian Schneider: +3.6
Vladimir Guerrero: +3.4
Todd Helton: +3.0
Paul Konerko: +2.9
Jason Varitek: +2.8
Eric Byrnes: +2.6

If we subscribe to the Harrah Hypothesis, we would figure each of the six to experience a sharp drop in batting average in 2008. Let’s look at three to see just how primed they are for decline.

Magglio Ordonez
After beginning his career with an aversion to the base on balls, Ordonez slowly learned the value of plate discipline. By his fourth year in the league in 2000, he walked on 9.3 percent of his plate appearances, which was right at league average. He would hover around this territory until last season, when he set a career high walk rate of 11.3 percent—up over four percentage points from the previous season.

In 2007, Ordonez had the best season of his 11-year career. He set personal highs in batting average by 43 points (.363) and on-base percentage by 52 points (.434), and missed his career best slugging percentage by a mere two points (.595). His OPS+ of 167 was also a career high.

Ordonez was certainly good, but he also had a bit of luck on his side. His .385 BABIP was 62 points better than his previous career high.

The best comp for Ordonez in this situation isn’t Harrah, but it’s Matthews from the 1984 season. Like Ordonez, Matthews was 33 when he set a career high with a .410 OBP and a 129 OPS+ while hitting .291. Like Ordonez, Matthews was aided by a little luck: His .339 BABIP in 1984 was also a career high. It was a solid year and anyone who was watching WGN that summer will tell you it seemed as though the man they called Sarge could do no wrong.

But the following year was a disaster for Matthews. His walk rate plateaued, but his other numbers tumbled across the board. He lost 56 points off his average and hit .235/.362/.406—well below his career averages—while his OPS+ dropped to 107. Yikes.

Since Ordonez, like Matthews, benefitted from an above-average BABIP, it’s not a stretch to predict that he will lose more than a few points off his batting average. And considering Ordonez posted so many career high numbers in 2007, it’s not difficult to make the call that he will regress in 2008. The Hardball Times Season Preview thinks he will hit .314/.386/.508, which would be in line with his career averages of .312/.370/.522.

If Ordonez declines as expected, and because he had such an outstanding season last year with a sky-high BABIP, it will be difficult to make the correlation of the Harrah Hypothesis.

Todd Helton
With a career walk rate of 14.8 percent, Helton always has understood the value of the base on balls. Part of that hasn’t been by choice—he’s been among the league leaders in intentional walks for the better part of this milleneum. The intentional pass is certainly part of it, but Helton does have a solid understanding of the strike zone. Otherwise, he wouldn’t be so far above league average when it comes to walk rate.

As you can see from the graph, Helton has proven the Harrah Hypothesis before. In his age 30 season in 2004, his rate increased by 2.8 percentage points and the following year he lost 27 points off his average. All his key rate stats continued in decline until last season, when he rebounded to hit .320/.434/.494 with an OPS+ of 133. Those numbers still don’t come close to his peak from the earlier part of this decade, but they were good enough to reverse a five year downward trend.

But can he hold off his decline for a second consecutive year?

According to the Harrah Hypothesis, that’s unlikely. Helton’s walk rate leaped from 14.3 percent in 2006 to 17.2 percent last season, representing his largest increase since his first two years in the league. Furthermore, Helton fits the profile of a player like Harrah or Matthews in that he has posted a solid season in the midst of a decline, which makes him an ideal player to watch under this theory. Like Harrah, Helton has a strong knowledge of the strike zone. Perhaps most important, since he’s fallen into this trap before, it seems even more likely that he will do it again.

The Hardball Times Season Preview projects Helton will resume his decline in the coming season, hitting .303/.420/.465. That would be a decline of 17 points off his batting average. But since he’s already been claimed as a victim of the Harrah Hypothesis once, it wouldn’t be surprising if he falls below those rates. If this theory holds true, Helton could dip below .300 for the first time in his career.

Vladimir Guerrero
Trying to read anything into Guerrero’s walk rate is like trying to drive with your eyes closed—it’s dangerous and potentially reckless. A notorious bad ball hitter, his chart representing his base on balls percentage looks more like an electrocardiogram.

Twice while he was in his 20s, Guerrero saw spikes in his walk rate, a 2.9 percentage point jump in 2002 and a 2.7 percentage point increase in 2005. His seasonal age will be 32 this year, so this is the first time since he entered his 30s that he’s eligible for examination under the Harrah Hypothesis.

Guerrero has long been one of the most dominant hitters in the game. With 365 career home runs and a line of .325/.391/.579, he has earned his place among the elite.

But, as mentioned before, he’s a free swinger. Last year, Guerrero saw only about 3.2 pitches per plate appearance. That ranked him third from the bottom among all American League hitters. It’s an approach that continues to confound the proponents of plate discipline. But what Guerrero lacks in discipline, he makes up for in pure ability.

If you combine his exceptional skills as a hitter and his swing-at-everything mentality, that means his walk rate has been very dependent on the other team issuing intentional walks. Like Helton, Guerrero is often among the league leaders in intentional walks. But where Helton has taken 100 unintentional walks in a season, Guerrero has topped 50 only once. Because of the smaller total of base on balls, it’s not surprising that in each of the three years in which Guerrero’s walk rate has increased, his total number of intentional walks has increased as well.

Because he takes so few unintentional walks, the intentional walk plays a large role in determining his walk rate. And because the opposition is walking Guerrero on purpose—almost 40 percent of his walks were intentional last season—he’s not a strong candidate to study under the Harrah Hypothesis.

Conclusions

While there’s enough evidence to support further study of the Harrah Hypothesis, it’s far from a slam dunk. There are so many other variables at play (career years and intentional walks to name two), it’s difficult to pinpoint two single, separate events and label them as cause and effect. And some players, because of their style of play, simply won’t ever fit into this study. In the end, we cannot look only at walk rate as a determining factor for an eventual decline in batting average.

But walk rate does appear to be a useful jumping off point when trying to identify players who are at a certain point in their careers where we can expect a decline in performance. While it’s not airtight, it deserves a closer look.