clock menu more-arrow no yes

Filed under:

The Glasnow Effect: Can we tie command to “getting lucky”?

New, comments

Examining the correlation between command and ERA-FIP in light of Glasnow’s Rays debut

MLB: Los Angeles Angels at Tampa Bay Rays Kim Klement-USA TODAY Sports

Editor’s Note: This piece has been updated with a couple brief thoughts from after Glasnow’s Rays debut Wednesday night. They are at the end of the piece.

Tyler Glasnow is bound to face heightened expectations in Tampa Bay, not much unlike the player he (alongside Austin Meadows and Mr. PTBNL) was traded for. For Chris Archer it was the dreaded “ace” label hanging over his inflated ERA. For Glasnow, it will be that he is the arm selected to replace one of the most popular pitchers in Rays history.

Not helping Glasnow’s case, either, will be that he and Archer suffer a similar affliction — although it is in a far smaller sample for Glasnow than Archer, here is an interesting note:

Archer: 1,063.0 IP, 3.69 ERA, 3.48 FIP, 3.45 xFIP, 0.24 ERA-FIP
Glasnow: 141.1 IP, 5.79 ERA, 4.91 FIP, 4.60 xFIP, 1.19 ERA-FIP

Before the season even started, I looked at some of the reasons for Archer’s running trend of underperforming his FIP.

Over the past three and a half seasons, of the 64 pitchers to throw at least 500.0 IP, only three pitchers have under-performed their ERA by a bigger margin than Archer (sixth percentile). Over the past season and a half, of the 502 pitchers with at least 100.0 IP, only 19 pitchers have under-performed their ERA by a bigger margin than Glasnow (fourth percentile).

When thinking about what Glasnow and Archer have in common, poor command is the first idea that jumps to mind.

As noted in the aforementioned preseason article, here are Archer’s percentile ranks, per Baseball Prospectus’ command statistic, updated for 2018 and with Glasnow added in the mix (where not applicable, “XX”):

(Before we get to the charts, note that Baseball Prospectus’ command statistic looks at pitchers’ ability to locate pitches in certain zones within the strike zone consistently. It’s an imperfect measure, but it’s also the best measure of a fascinating aspect of the game.)

Archer and Glasnow Command Ranks

Year Archer CMD Percentile Glasnow CMD Percentile
Year Archer CMD Percentile Glasnow CMD Percentile
2013 4.8 XX
2014 5.9 XX
2015 3.4 XX
2016 6.0 XX
2017 26.7 5.9
2018 17.3 24.4

The improvement Glasnow has made in 2018 and whether that is due only to a move to the bullpen, rather than overall improvement, is a matter for another article. For now we can observe, clearly, that neither guy thrives in terms of command.

This got me thinking: Did we just trade for a pitcher doomed to fail in the eyes of Rays fans the exact way the man he was traded for did? What about if we take a look at this on a league-wide level? Is there a correlation between command and “getting lucky”?

It that last hypothesis were true, it would make some sense. Pitchers like Clayton Kershaw, Johnny Cueto, and Kyle Hendricks have some of the best command in the game. They also consistently post ERAs lower than their FIPs. I decided to start in 2013, and the original plan was to look at the last five years’ worth of data.

Needless to say, the first-year result deflated my theory quite a bit.

Far from having a strong correlation, command and ERA-FIP (note: I will be referring to ERA-FIP because that is the stat, but in this case, it was FIP-ERA in order to compare apples-to-apples rankings) didn’t even have a positive correlation in 2013! The exact figure was -0.09, a far cry from anything close to correlation.

The result was so negative, that I decided to skip a few years, wondering if BP hadn’t tinkered with their command formula a bit since the early days and maybe there would be new (and more bias-confirming) results in more recent years.

Not much better. At least there is a positive correlation this time, but it’s minor at best, and just random at worst (0.15 to be exact).

The next thought to cross my mind was: Maybe this is being muddled by the middle class, the examples used earlier (Glasnow, Archer, Cueto, Kershaw, Hendricks) were all towards the extremes in terms of command.

So, just taking the top and bottom 20 by command, is there anything closer to correlation?


Certainly not enough to make a strong case.

The 2018 Top and Bottom Tier chart comes the closest, but even then, we’re talking about less than halfway to statistically sound correlation. And that’s after all the manipulation of half-season and shrunken sample size.

Lessons Learned for Tyler Glasnow

Well, as is the case with many of these types of baseball studies, what we didn’t learn can still tell us something. Let’s not set Glasnow in stone as one of those pitchers who will always have a higher ERA than FIP. Looking over the two years chosen above, there were — and are — plenty of low-command pitchers who can beat their FIP (Al East combatant Rick Porcello did it in both of the sample seasons).

Plus, there’s the fact that Glasnow is less than 200.0 innings into his big-league career. It will be interesting to see how Kevin Cash, Kyle Snyder (who sports the same height as the 6’ 8” Glasnow, which has to mean something), and the rest of the Rays powers that be look to handle Glasnow in their efforts to extract every ounce of potential from this high-ceiling arm.

The good news is that Glasnow’s Rays debut couldn’t have gone much better than it did Wednesday night.

Glasnow got handed the start a day after coming over in the trade deadline, and while he was always going to be on a pitch count, he did as well as he could in his limited start, going 3.0 IP, allowing just one run on two hits and a walk, notching five strikeouts on some nasty — and controlled — stuff. Glasnow was pounding the strike zone, with 34 of his 48 total pitches coming in as strikes.

DRaysBay commenter BravesRay noted that Glasnow appeared to be more comfortable working up in the zone (the lone home run came on a curveball that didn’t quite get out of the zone), theorizing that because of his extreme height, this may be a more comfortable place for Glasnow to work in. It’s not a bad thought, and enough to make you wonder if Glasnow will be able to trade his 57% groundball rate for fewer walks by pitching up.

Glasnow’s fastball also seemed to have some nice natural cut, and just look at how many pitches caught the very end of either side of the zone in the Brooks Baseball graphic above. It appeared as though Cash, Snyder, and Glasnow already had a plan — and it was a good one.

Obviously this was one outing — three innings and fewer than 50 pitches — but the early returns have to assuage any worries at least a bit.