Should MLB Teams Try to Emulate Kansas City's Contact-Based Offense?
Hearing commentary after the Royals' World Series victory, it would be easy to think a paradigm shift is about to occur in baseball.
En route to their title, Kansas City thrived on speed, defense, and as youâ€™ve surely heard, making contact.
The Royals tied for the highest contact rate in baseball in 2015, at 81.9% (incidentally the same rate as the Aâ€™s). No team in the Majors struck out less frequently than Kansas City, who only went on down on strikes on 15.9% of their plate appearances.
No one is even close, actually, as the gap between Kansas City and the second-ranked Aâ€™s in K-rate is equivalent to gap between Oakland and 12th-ranked Cincinnati.
Thanks to their low strikeout rate and roughly league average batting average on balls in play (.301), the Royals hit .269 as team, behind only the Blue Jays.
Despite their high average, though, the Royals ranked 11th in both on-base percentage (.322) and slugging percentage (.412), thanks to the lowest walk rate in baseball (6.3%) and a .144 isolated power rate (tied for 19th).
These numbers translated to a 99 wRC+ after park-adjustments, meaning the Royals were a slightly below average offense in a context neutral setting.
The Royals did rank seventh in runs per game, averaging just under 4.5, thanks to success with men on. Only the Blue Jays drove in a higher percentage of runners on base, and itâ€™s not illogical to guess this has something to do with putting the ball in play (itâ€™s harder to drive in a run with a walk).
The current run environment in MLB, where strikeout totals are at all-time highs, could theoretically make a contact-based approach more appealing, but we should hold off before crediting Kansas City for reinventing the wheel.
The Relationship Between Scoring and Contact Rate
The logic behind why the Royalsâ€™ contact rate has made them successful is admittedly sound.
In each season from 2008 to 2014, a new all-time high strikeout rate was set. In 2015, the league-wide strikeout rate was 20.4%, which could only tie with 2014 for the highest ever.
Because teams are more dependent than ever on strikeouts, it would make sense than an offense that avoids striking out would have success.
This certainly worked in Kansas City...but the results in Oakland and Atlanta were less encouraging.
Oakland, as mentioned, tied with Kansas City for baseballâ€™s highest contact rate, while the Braves came in fifth with an 81.0% contact rate. The Aâ€™s ranked 14th in baseball in runs scored and 21st wOBA, while the Braves were last in baseball in both wOBA and scoring.
And despite playing in this environment where power pitchers dominate, the teams at the bottom of the contact rate list did just fine.
The Cubs (74.8%) and Astros (75.9%) had the lowest contact rates in the Majors, but the Moneyball staples of power and getting on base helped both to be among the highest-scoring teams in their respective leagues.
Houston was sixth in the Majors in runs scored and trailed only the Blue Jays in terms of wOBA, thanks to the Majorsâ€™ second-highest ISO. The Cubsâ€™ non-pitchers had baseballâ€™s second-best walk rate, and tied for the fifth best wOBA.
Looking at baseball as a whole last year, there was no statistical relationship between a teamâ€™s contact rate and its runs scored total (a correlation coefficient of -0.02; see graph) or its wOBA (-0.09)
Has Avoiding Strikeouts Become More Important?
Suppose we isolate strikeout rate, rather than look at overall contact rate. Avoiding strikeouts is something the Royals have excelled at recently, as they had the lowest strikeout rates in each of the last two seasons. In 2015, they were historically good, posting the lowest strikeout rate relative to league average since 1950, according to FanGraphs.
Hitting coach Dale Sveum told USA Today this approach gives his team an advantage:
It is a big difference, putting the ball in play and not striking out. I had somebody five years ago say, â€˜Whatâ€™s the difference between striking out and hitting a groundball to the shortstop?â€™ What kind of mindset is that? Iâ€™ve never seen a strikeout hit the outfield grass. Iâ€™ve never seen somebody make an error on a strikeout. Iâ€™ve never seen a strikeout go over the fence. Making contact is huge.
This line of thinking cuts against sabermetric orthodoxy, which generally maintains that striking out is not a terrible outcome for a batter.
Are things different now that pitchers are striking out batters at historically high rates?
No, as the numbers still backup the statheads, rather than Sveum. Strikeouts, while great for pitchers, are neutral events for batters, not inherently good or bad. K% has a correlation coefficient with both wOBA and runs scored of -0.1, meaning there is a very weak relationship here.
Other factors are simply more important to offense than avoiding strikeouts, even in the current offensive climate. Some teams are like the Brewers, who had the eighth-highest strikeout rate in the Majors and also ranked 25th in walk rate and are tied for 22nd in ISO. As a result, they finished 25th in wOBA and 22nd in runs scored.
The Astros though, had the second-highest strikeout rate in baseball, but as mentioned, were one of the highest scoring offenses in the game, thanks a top-10 walk rate and the Majorsâ€™ second-highest ISO.
Hitting for power and drawing walks are both more important to scoring runs than avoiding strikeouts.
ISO correlates with runs scored at the very high rate of 0.75, while walk rate comes in at 0.36.
Before moving on, Sveum made two other points that should be addressed. He mentioned that nobody has made an error on a strikeout. Itâ€™s not technically true (a catcher could airmail a throw after a dropped third strike), but we know what he means: putting the ball in play theoretically puts pressure on the defense, which will eventually lead to errors.
After the World Series, where the Mets made a number of mistakes in the field at inopportune times, this argument seems compelling.
Over the course of the season, though, teams that made a lot of contact were no more likely to reach on an error than any other team. Reaches on errors correlated with strikeout rate and contact rate at a rate of 0.00 and 0.02, respectively. Thereâ€™s nothing here.
The Royals reached on an error 58 times, only twice more than the average team and four more times than Houston.
Sveumâ€™s more erroneous claim is the argument more contact leads to more home runs. Technically speaking, he is correct this time: no one has ever hit a home run on a strikeout. You do need to make contact to hit the ball out of the park.
However, the teams that hit home runs most often tend to be the ones that make less contact. Contact rate and home runs correlate at -0.45, meaning as contact rate decreases, home run rate increases (not at the same rate, but there is a relationship here).
Home run rate also sees an increase as strikeout rate increases (r = 0.36), so Sveumâ€™s comments seem a bit misleading.
Contact and Clutch
Before closing the book on this, itâ€™s worth noting that the 2015 Royals will be remembered for their contact hitting and high-leverage performance. Perhaps surprisingly, there is some connection between the two.
The Royals tied for 10th in baseball in wOBA (.318) but scored more runs than all but six teams, and performance with men on explains much of the difference.
With men-on base, their wOBA jumped to .333, which is third in the league. They were equally good when the game was on the line, again ranking third in the league in wOBA in high-leverage situations, according to FanGraphs.
The top two offenses in â€œ clutch scoreâ€ (win probability added minus context-neutral WPA) over the last two years belong to the 2015 Royals and 2014 Royals.
Whether you want to chalk this up to random sequencing variation or some intangible skill, this is one of the main reasons why Kansas City outperformed its third-order winning percentage by almost nine games, behind only St. Louis and Minnesota.
While watching the World Series could have convinced you otherwise, this is not a sustainable means of success, as clutch performance is not consistent on the team level.
That said, all other things being equal, teams that make a lot of contact do to tend to perform slightly better in big spots.
Strikeout rate and clutch score correlated at -0.35 in 2015, so there is some connection here. Jeff Sackman of the Hardball Times and Jonah Keri in Baseball Between the Numbers also found a slight correlation between not striking out and clutch performance.
If you square the correlation coefficient, you get a coefficient of determination of 12.3%, meaning that 12.3% of the variation in clutch score can be explained by strikeout rate.
As Keri said, the correlation is â€œnot overwhelming but enough to be interesting.â€
Overall though, it does not seem like contact rate is the new market inefficiency, and it is unlikely we will see teams going out of their way to emulate Kansas Cityâ€™s offense.