MLB

Why Do Projections Always Hate the Royals?

Once again, virtually every pre-season projection has the Royals missing the playoffs. Why?

It's the Kansas City Royals versus the computers.

Seriously, for some reason, those electronic thinking boxes hate the Royals. Yes, if you looked at nothing but Baseball Prospectus' PECOTA rankings (out Tuesday) or FanGraphs' pre-season projections of 2016 records, you'd see the same thing. 

They don't like the two-time AL champs and defending World Series champions in 2016.

ProjectionWinsLossesPCTRdiffResult
PECOTA76860.487-445th
FanGraphs76860.487-194th


Both projections have the Royals going 76-86 this year, which by itself is not terribly eye-raising. Sure, they are the defending champs, but teams often struggle the year after winning a World Series. But it's the continuation of a trend where these projection systems seem to get it wrong on Kansas City.

As BP's own Sam Miller noted in his piece on the Royals out Tuesday, Kansas City has outperformed their PECOTA projections by 21 games over the previous three seasons. Only Baltimore, Pittsburgh and Oakland have done better. Last year, the Royals were only projected to win 72 games by PECOTA and 80 by FanGraphs.

Instead, they went 95-67 and did a little bit better in the postseason.

So why does this continue to happen? There are a few theories.

Bullpen Strength Hard to Predict

The Royals are short on starting pitching coming into every season. Last year was no exception. They are, instead, built around their bullpen, and relievers are notoriously inconsistent. 

A season's projections are based off of players statistics and what those players have done in previous seasons. Because of their volatility, projection systems like FanGraphs and PECOTA can sometimes get it wrong on key relievers, like Wade Davis. Miller notes that Davis in particular has been hard to project because he has such a short history of dominance. Prior to that he was a middling starting pitcher.

Contact, Contact, Contact

Players who walk a lot and hit for a lot of power, over time, are easier to predict. Heck, even guys who strike out a lot are easy to predict.

But that does not describe the Kansas City Royals offense.

The Royals make contact. A lot of it. Last year, they struck out 144 times fewer than any other team. Their 15.9% strikeout rate was far below the next closest team, the A's at 18.1%. 

They also had the worst walk percentage in baseball (6.3%) tied with the Marlins at the bottom of the heap. Their 139 homers were seventh-fewest in baseball, and their .144 isolated power (ISO) was tied for 10th-lowest.

With all their athleticism, the Royals use their speed and contact ability to get a lot of hits, finishing with the second-most in baseball last year, behind only the Detroit Tigers. And while they don't have homer-power, they hit a ton of doubles (300, third-most) and triples (42, sixth in MLB). 

It's hard to count on some of those things year in and year out.

Clutchness

There are some who don't believe that certain players in baseball are "clutch," that they do better in big situations. And while we love the numbers here, I believe some players (and teams) are more "clutch" than others, and that it is difficult to quantify.

FanGraphs has a team stat called "Clutch", which measures how well a player performed in high leverage situations. It looks at how much better or worse a player does in big situations, as compared to the league and in comparison to the player himself. So if a player is a .300 hitter all of the time, the fact he hits .300 with runners on base doesn't add much to his "Clutch" factor.

Last season, the Royals finished with the highest "Clutch" factor of any team in baseball, at 5.05. A "Clutch" for an individual player greater than 2.0 is considered excellent, 1.0 is considered great, 0.5 is considered above average and anything below 0.0 is considered either average or below average.

Eric Hosmer's clutch ratings were off the charts (2.17), while Lorenzo Cain and Omar Infante were also excellent in high leverage situations last year (1.86). Ben Zobrist and Kendrys Morales were also considered above average. 

As FanGraphs notes, Clutch is not meant to be predictive. However, it was the second straight year Kansas City led the league in Clutch, and in 2013, they had the fifth-highest Clutch score. So perhaps it isn't crazy to think this might be part of their DNA.

Late Season Additions

When PECOTA and FanGraphs put out their projections last year, the Royals had not acquired Zobrist or Johnny Cueto, two players who helped immensely during their stretch run. 

A projections system cannot include a half season worth of production from players not on the team at the start of the season.

Surprises

Projecting a 3.58 ERA and a 3.82 FIP for Edinson Volquez last year was tough. Even though he had a 3.04 ERA for Pittsburgh in 2014, his FIP was 4.15 and his ERAs the five seasons prior to that were 5.71, 4.14, 5.71, 4.31 and 4.35.

And did anyone expect Chris Young to put up a 3.06 ERA last season? Certainly not with a 4.52 FIP. And certainly not after posting a 5.02 FIP in 2014 and after missing the entire 2013 season due to injury.

But sometimes key players outperform expectations, and that happened with a number of Kansas City players last year.

Analysis

All in all, there are reasons why projection systems have a hard time with Kansas City. One of their other chief strengths -- defense -- is also hard to quantify and, therefore, project into a solid wins/losses number.

Still, there's no doubt Royals players will use this as more fuel to be added to the fire, as they look to repeat as champs in 2016.