Master and Commander?
What Strikeout-to-Walk Ratio Tells Us (and What it Doesn’t)
In the baseball blogging game, one never knows when inspiration will strike, or from where:
Crasnick’s long-ago tweet got us to thinking about K/BB ratio, a stat that seems to have gained currency beyond the analytics community, and what it can (or can’t) tell us about a pitcher. It’s become a supporting pillar of, among other things, the HOF credentials of Curt Schilling and Mike Mussina.
In very broad terms, K/BB ratio can function as a general proxy for pitcher command, the ability to not only hit the strike zone, but to throw to a specific part of the zone. Command differs from control, which is simply the ability to throw strikes. Major league pitchers can throw strikes if they want to; the problem confronting pitchers is that major league hitters can hit most strikes. Think about how many times you’ve heard a broadcaster say, after a ball has been stung to the deepest reaches of the park, that the pitcher “missed his target.” That, in a nutshell, is command: The ability to hit that target within the strike zone.
Not that he needed the help, but the table below underscores Crasnick’s beautifully dismissive wisdom. Pitchers with exceptional K/BB ratios tend to be… exceptional. Here are the best career K/BB, or “command” ratios of all-time (min 2500 IP):
These are some of the best pitchers to ever toe a slab (if you tell me you expected to see Javier Vasquez among these names, you’re either a wizard or a liar). You’ll doubtlessly note some conspicuous exclusions along the lines of Roger Clemens, Walter Johnson, Sandy Koufax, etc.[i] That’s because K/BB, while a fun stat, only tells us so much about pitcher command. After all, there are many times the target is outside the strike zone. As Greg Maddux, perhaps the greatest command artist of all, once said: “Make the balls look like strikes, and the strikes look like balls.”
So if pitchers who strike out a lot of batters while walking few tend to be very good, is the converse always true? Are pitchers with lousy K/BB ratios… lousy?
The Worst of the Worst? Sort of?
Here are the pitchers with the worst K/BB ratio in history (min 2500 IP):
As you might expect, no true immortals here. But no real duds, either. Most of these pitchers, in fact, provided significant value over the course of their careers. The most famous — and successful — of this lot is “Sunday Ted” Lyons, who has a plaque in Cooperstown in recognition of his efforts over two decades for some terrible White Sox teams.[ii]
K/BB ratios can be a little misleading: Lyons’ command looks awful by today’s standards, but he was just about league average for his time — it was a high-contact era, and hitters regularly walked more often than they struck-out (they did neither with much frequency). The same conditions hold true for all of the pitchers on this list: Despite “terrible” command, none of them were terrible pitchers (you don’t stick around the majors for more than 2500 innings if you can’t pitch). They were products of their time.
And here’s where that old bugaboo “context” comes in. As you’ll note, all the pitchers (with the exception of Juan Marichal and Dennis Eckersley) with the best lifetime K/BB ratios began their careers in the 1980s or later. It’s not a coincidence:
The trends are stark: The league K/BB ratio hovered around 1:1 from 1918 to 1953; the 1950s saw an upsurge in strikeouts (and a corresponding lift in HR), pushing the league K/BB ratio above 1.50; strikeout rates would continue to increase through the 1960s (peaking at 5.9 K/9 in 1968); pitchers first whiff twice as many batters as they walk until 1964, with K/BB rate peaking at 2.09 in 1968, the year of the pitcher (Juan Marichal posts a 4.73 K/BB mark to pace the majors; Denny McLain leads the American League at 4.44 — a mark that doesn’t crack the top-10 in 2016). MLB lowers the height of the mound in 1969, the league K/BB ratio drops below 2.00 — and stays there until 2001; after a dip below 2.00 the following season, the K/BB rate surges past 2.00 in 2003 — and it’s been setting new highs, on an almost annual basis, ever since. The seven highest K/BB ratios in MLB history were recorded over the last eight seasons:
Highest League K/BB Rates:
Putting this in perspective, ML pitchers have averaged 2.64 strikeouts for every walk over the last three seasons; Tom Seaver averaged 2.62 strikeouts for every walk over the course of his career.
The single-season K/BB record for a starting pitcher is held by Minnesota’s Phil Hughes, who scuttled 11.63 batters for every one he escorted to first base in 2014 (only 16 BB in 209 IP).[iii]
Best K/BB ratio, single season (min 162 IP)
Notice anything jarring about this chart (well, other than the inclusion of Carlos Silva)?[iv] Every one of these seasons took place in the last 23 years; seven out of 10 in the last 15. Prior to 1994, only two pitchers produced a K/BB ratio north of seven: Cy Young (1905) and Fergie Jenkins (1971). Since 1994, the feat has been achieved 20 times. Does this mean pitchers en masse are more dominant today than ever? Again, yes and no (more on that in moment).
The increase in K/BB rates has been driven by an enormous, league-wide increase in strikeouts. Pitchers have averaged 3.22 walks per game over the last century-plus (3.1 the last 25 years); single-season rates have never been lower than 2.5 BB/9, or higher than 4.1 BB/9. Strikeouts, however, have been on an upward trajectory for most of the last century — and in an uninterrupted bull market for the better part of a decade. As Michael Bauman wrote in 2016:
“In 2016, the MLB average K/9 is 8.1, the highest ever. In 1993, it was 5.9. The last time MLB didn’t set or tie the all-time record for average strikeouts per inning was 2007, when big league pitchers came together to post a collective 6.7 K/9… The 23 percent increase in strikeouts over the past decade represents an almost fundamental alteration of the way the game is played.”
Pitchers as a whole are striking out a greater percentage of batters than ever before, while keeping walk rates essentially flat. [v] So in one specific aspect of the game (strikeouts), they are more dominant than ever — that said, while strikeouts are at historic highs, runs scored per game are nowhere near historic lows, mostly because hitters continue to mash home runs like it’s 1999.[vi]
So how to truly determine the best single-season “command ratio?” Glad you asked. We call it adjusted Command Ratio (CR+), and if you’re familiar with ERA+ or OPS+, it works in a similar way. CR+ compares K/BB rates against the league average by taking a pitcher’s K/BB rate, dividing it by the league rate, and multiplying by 100. A CR+ of 100 is exactly League average; a CR+ of 200 means the pitcher’s K/BB ratio was twice as good as league average. Using this formula yields a different single-season list:
Best Single Season Adjusted Command Ratio (min 162 IP)
Silva and Scherzer fall out of the top-10 entirely, while Saberhagen’s 1994 season is buffed to an even brighter shine (his 564 CR+ is 11% better than the next best mark in history). A couple of so-so pitchers from the Deadball era enter the Top-10, while Pedro offers even more evidence for his 1999–2000 peak as perhaps the highest in the game’s history.[vii]
Again, it doesn’t paint a complete picture of how well (or poorly) a pitcher performed in a given season — but it’s a fun bit of ephemera. For the record, the owner of the worst adjusted Command Ratio for a full season is Ernie Wingard, who in 1924 struck out 23 batters while walking 85 for a CR+ of 32. Perhaps the oddest part of Wingard’s season? Despite walking nearly four batters for every one he struck out, Wingard somehow fashioned a winning record (13–12, for a St. Louis team that finished below .500) and pitched to an ERA 29% better than league average over his 218 innings (we hope he bought his defense a steak dinner as thanks).
Worst K/BB ratio, single-season
Ok. You’ve made it this far. How about some trivia sure to impress your baseball-geek buddies during your next round of boozy trivia?
Four pitchers with a minimum of 900 IP have compiled exactly the same number of career strikeouts and walks.
Ned Garver was the best of the bunch. Pitching for some terrible teams, Garver finished with a career record of 129–157, despite an adjusted ERA that was 12% better than the league average (Garver, in fact, makes our Unluckiest Staff of All-Time, where he keeps some excellent company). Garver pulled off quite the magic trick in 1951, compiling a 20–12 record for a pathetic St. Louis Browns club that went 52–102. Garver became the second pitcher to win 20 games for a team that lost more than 100.[viii] He struck out 84 while walking 96.
Atley “Swampy” Donald wasn’t without his moments — he won an AL rookie-record 12 straight starts in 1939 — and he had impeccable timing, catching on with perhaps the greatest dynasty of them all: the late-30s — early-40s Yankees (explaining his career .633 W%). Swampy played on four World Series champions, and later had a long, successful career as a major league scout (signing, among others, Ron Guidry).
We’ve looked at the best of all-time, the worst of all-time, and the most symmetrical of all-time. So where does this leave us when it comes to K/BB ratio?
We can say with ironclad certainty that command ratio tells us a lot about pitcher effectiveness — except when it doesn’t. It tells us a lot about pitcher command — except when it (sometimes) doesn’t. Pitchers who produce lots of strikeouts as compared to walks allowed are usually very good, and pitchers who have poor K/BB ratios are usually very bad — except when they’re very good. Command ratio means more today than it did even 20 years ago, but it’s probably useless when measuring some of the great pitchers (like Lefty Grove, with his career 1.91 K/BB) who worked in very high-contact eras.
As with most things baseball, it’s all very tidy.
— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —
[i] Walter Johnson and Christy Mathewson share the record (nine) for most seasons leading the league in K/BB ratio. Dazzy Vance led the league for eight consecutive seasons beginning, at age 31, in 1924.
[ii] So named for the genteel schedule he kept in his pitching dotage. After 2600 innings (and more than 200 CG) over his first 10 seasons, Lyons’ shoulder by 1934 was no longer able to bear the ravages of a full season. Beginning in 1935, White Sox Manager Jimmy Dykes designated Lyons as his “Sunday Pitcher,” starting the stocky right-hander on six days’ rest. Lyons responded well to the reduced workload, arguably pitching the best ball of his long career after the age of 35. From 1935–1946, Lyons completed 139 of 181 starts, with an ERA+ of 131. His 1942 season — at the age of 41 — is a sensational study in statistical symmetry: 20 GS, 20 CG, 180.1 IP. Lyons went 14–6, leading the league in ERA (2.10) and ERA+ (171)
[iii] Clayton Kershaw was on pace for the best Command ratio of all-time for a starter before being sidelined with a back injury in 2016. After 15 starts, Kershaw had struck out 141 batters while walking… seven. That’s not a misprint: His K/BB ratio stood at 20:1 (he’d finish the season with almost 16 strikeouts for every walk over 149 innings).
[iv] Silva in 2005 established the single-season record for fewest walks per nine innings (0.43).
[v] There are a number of factors contributing to this rise. Hitting philosophy has changed: Strikeouts, rather than signifying a particularly shameful kind of failure, are just another out (which seems awfully accommodating to pitchers, since they’ve taken the approach that strikeouts are the best kind of out); pitchers are throwing harder than ever (bullpens are stocked with fire-breathing robot dragons designed to throw 101 MPH), and pitches are sharper than ever (cutters, sliders and splits have replaced the old-fashioned curveball as the wrinkle of choice); and teams have all but abandoned contact-based strategic elements like sacrifice plays. In response to the disturbing strikeout trends, MLB has proposed a tightening of the strike zone.
[vi] HR/G: 1999: 1.14; 2016: 1.16
[vii] The lowest K/BB ratio to ever lead the league was Walter Johnson, at 1.38 K/BB in 1925. Johnson’s CR+ mark of 179 was still 79% better than the league average of .77 K/BB. Had Clayton Kershaw been able to muster 13 addition innings in 2016, he tops the CR+ list at an absurd 618.
[viii] Irv Young was the first, for the 1905 Boston Beaneaters. Steve Carlton, as you know, was the last, for the 1972 Philadelphia Phillies.