The Good and Bad News Regarding Robinson Cano
If you just look selectively at some of the names, it doesn’t seem so bad. So far this year, through a little more than a quarter of the season, Robinson Cano has hit about as well as Matt Kemp. He’s hit a little better than Troy Tulowitzki, and a lot better than Carlos Gonzalez and Chase Utley. Those are all names of proven star players! The problem being, they’re star players who’ve sucked. On the one hand, it’s encouraging to see these names near the bottom of the wRC+ list — it’s a good reminder that struggles can be perfectly normal — but that doesn’t make the struggles themselves any easier to tolerate, and in the case of Cano, his problems are among the factors contributing to the Mariners being a disappointment. It’s not that they’ve wasted a month and a half of Nelson Cruz hitting for Barry Bonds power, but they’ve accomplished less than you’d think.
Because of the team expectations, there’s more pressure than there usually is. Because of the enormity of Cano’s contract, there are more eyes on him than there usually are. And because of Cano’s age, there’s a bit more fear than you’d usually figure. The good news is there’s potentially good news. The bad news is we don’t know what to make of it yet.
Cano, right now, has a 78 wRC+. We know that’s weird because last year he was much better, and for his career, he’s been much better. One also shouldn’t compare roughly two months to other full seasons, because over full seasons, events have time to balance out. So let’s plot Cano’s whole career, to see how he’s trended before and to see whether this slump is unprecedented. He’s posted this year’s numbers over 44 games. Here are his 44-game rolling averages for wRC+, since his big-league debut:
Even the most consistent players are inconsistent players. In part this is because we measure players by their results, and not by their talent, and results bounce around. Talent is what we always want to know, but we kind of have to beat around the bush. By results, the plot above is instructive. As with everyone, you see relative peaks and valleys. However, where Cano is now, he hasn’t been since the middle of 2009, when he was 26. For several years since, Cano’s kept himself above 100, over months and a half at a time. His slumps have been either less severe, or shorter in length. This is legitimately unusual, even if it’s not the first time it’s ever happened.
So the door’s open for further analysis. Cano has a fairly high groundball rate. But last year, he had the same groundball rate. Cano has a lower contact rate, but in 2012, he had basically the same contact rate. The plate-discipline stuff looks similar. Cano’s hard-hit rate is strong enough. He’s definitely gone to the opposite field less, but he’s gone up the middle more. How about some words from Cano himself?
“I don’t know,” he said. “I feel good. I’m seeing the ball. I would say I’m missing my pitches — pitches right down the middle that I’m hitting for foul balls. There are no excuses for that. I’m just struggling. Everybody goes through these things and I’m trying to stay positive. But it’s not clicking right now. There’s no excuses for that.”
Not the most helpful thing, but maybe there’s some substance in there. From the same article, there’s talk that Cano has reviewed video of his swing and spotted some flaws. So he’s working to correct those, and, baseball is a game of constant adjustments, and, so on and so forth. Maybe that’s it. What we really want to know: is Robinson Cano now a less-talented baseball player? With help from Baseball Savant, we can try to see what Statcast says.
I know a lot of people play around with average batted-ball velocity, but I also like to use cutoffs. And what could be a cleaner, more pleasing cutoff than 100 miles per hour? A batted ball struck at 100 miles per hour is a well-struck batted ball. That implies good things about bat speed and contact. And, it’s funny: to this point, Cano has hit 44 batted balls at 100+, that’ve been picked up by the trackers. That ties him with Hanley Ramirez for the most in baseball, two ahead of Josh Donaldson. On a rate basis, Cano’s at 40% 100+ batted balls, basically tying him for tenth. Of all the balls that’ve been recorded, we find Cano’s rate of balls hit 100+ above the rates of Paul Goldschmidt and Nelson Cruz.
What that seems to suggest is that the force is there. The ball has, pretty frequently, come off Cano’s bat with great haste. Now, not every single ball in play has been recorded. There are a few data errors. And we don’t have information for Cano’s seasons in the past. But you can’t do much better than ranking near the top of the league. Seems like Cano’s bat speed is okay. Seems like the quality of contact is there.
But, of Cano’s balls hit at least 100 miles per hour, most have been grounders. Just 41% of them have been hit in the air. Goldschmidt, for example, is at 90%. Cruz, 74%. The average is in the vicinity of 60%, for players with decent high-velocity batted-ball skills. Again, we don’t know where Cano has been before, but it stands to reason well-struck grounders are less dangerous than well-struck non-grounders. Well-struck grounders can turn into outs, especially when they’re pulled by a left-handed hitter. Well-struck air balls fly toward or beyond the fences. Cano’s bat-to-ball force is fine; it’s the alignment that seems a little off.
And I’m afraid I don’t know what that means. This data is too new to me to know what to expect. Maybe this is a part of how players decline. Or maybe this is what a harmless slump looks like, where the difference between slump numbers and regular numbers is a few millimeters on the bat. Maybe Cano really has been just missing, for no particularly negative reason, and that will sort itself out. Or maybe this is bat control going before bat speed goes, and Cano’s getting worse quick. For whatever it’s worth, Christian Yelich has an even higher rate of 100mph+ batted balls than Cano, and a lower rate of such balls being hit in the air. Yelich, presumably, isn’t experiencing age-related decline. Add new information and everything’s still mysterious.
This is all kind of a funny turn, given that one of the reasons provided for the Mariners signing Nelson Cruz was so that he could in some way protect Cano in the lineup. Cano, from a few weeks back:
“I love having him here. I’m seeing more strikes and fastballs than I’ve seen in years. He makes such a difference to everyone in the lineup.”
With Cruz behind him, Cano’s hit worse. Where last year he was issued 20 intentional walks, this year he’s at one. Cano’s at a career-high rate of first-pitch strikes, and he’s at a career-worst rate of pitches seen while behind in the count. This is a partial explanation for his struggles: Cano’s fallen behind a little more, so he’s had to expand a little more, and sit on pitches a little less. Cano definitely isn’t being pitched around. But he’s not making the pitchers pay for it. It’s another example of how protection can alter behavior, but not really improve the end result.
And for Cano, having Cruz behind him means he needs to punish the hittable pitches. He’s going to have fewer walks available, so he’s going to need to build his numbers around hits and extra bases. So far, no dice. Encouragingly, the good contact is there. It’s just not leaving the bat at the proper angle. What that means going forward is anyone’s guess. It feels like some of those well-hit grounders will turn into well-hit liners, but I don’t like making assumptions with data I’ve never used. Which is why we’re left to play the waiting game the Mariners hope ends any minute.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
If there was an interactive way to plot any given player’s rolling wRC+ (or any stat) over the course of a selected sample size of games, I might never stop playing with it… Does anyone know if such a site exists?
Also, thanks for another great one, Jeff!
I’ve seen Fangraphs embed Highcharts in articles before; plus that’s what they use for probably every interactive chart on the site (like win expectancy, PITCHf/x game charts, etc.)
Jason Heyward is having a similar issue, only his ground-ball rate has increased much more dramatically.
This has ALWAYS been jason heyward’s issue.
Heyward has a career GB/FB of 1.48. This season he’s at 2.61.
Last year it was bad though. NOw it is worse
I wonder if Cano’s decrease in HRs last year was an initial sign of some decline? His drop last season was interesting and a bit extreme. Sure, there was an expectation he’d lose some moving to a park with a less-kind RF, yet of Cano’s 27 HRs in 2013, 11 were hit at Yankee Stadium, 16 on the road. Looking further, his power outage actually started during second half of 2013 when he hit only 6 in his final 65 games, after hitting 21 bombs prior. While his home/road HR splits were not the norm for his Yankees career, his power numbers on the road were consistent. He hit 11 HRs on the road in 2012, 12 in 2011, 13 in 2010, etc. I thought 2014 might have just been a fluke, but maybe it’s just where Cano is right now. Also, his hitting coach on the Yankees, Kevin Long, used to run cage drills with him designed to get him to pull more balls with loft. If he’s hitting more ground balls with authority and less in the air, perhaps that’s contributing to it.
Cano may be past peak now, but he’ll need to rebound. If not, this will be a very long nine years for the Mariners.
this is mostly reflective of the decline in his skillset: http://www.fangraphs.com/community/robinson-canos-replacement-level-floor/
the ISO should pop back up to the .120-.130 range–he’s been unlucky on his fb’s–but the rest of it is mostly real. he’s hitting those hard hit ground balls into the shift, and they are going for outs. he did the same thing last year but got lucky on his balls in play. last year’s xbabip was .299, which is exactly where his babip is now.
Here’s a tool I put together that uses StatCast data to calculate expected AVG/OBP/SLG lines for players. Feel free to dig through it and offer any suggestions you may have.
https://docs.google.com/spreadsheets/d/1-XohbJlWIceDS2Rc8_7-rOxv9avU3IwMCecPkUNxlYU/edit?usp=sharing
His plate discipline numbers indicate he’s chasing. He’s always chased, but now he’s not making quite as much contact on pitches both in and out of the zone. Impossible to know if this is age related or not, but there’s a good chance it is. As a Mariner fan, I’d like to see him become more selective.
Cano is owed 9 years/$216m
Pujols is owed 7 years/$189m
Which of the 10/$240m contracts is more of an albatross at this point?
Yes
Good question, Pujols contract has less to go, his value is obviously less tied to position and I would hazard a guess the Angels are willing to roll with a higher payroll then the Mariners so although there is a higher chance Cano returns to his old form and he’s a couple years younger I think his contract is going to be albatross for his team
In the rolling average plot, the peaks seem to have an unpredictable amplitude but the period looks regular. Is this just a feature of plotting a rolling average or is there something else going on here?
The oscillations have a period significantly longer than the window, so they’re “real”. They probably aren’t quite regular enough that you could hand-pick a window to squash them completely… but the period in places looks kind of close to a year (maybe), so it might be interesting to look at his career splits by month to see if peaks and valleys tend to occur around the same time of year. Or drop a year-long rectangular window on the graph, which would squash any exactly yearly oscillation.
The 44-day window completely squashes any oscillation with an exactly 44-day period, and also any oscillation with a period that divides evenly into that (22-day, 14-2/3-day, 11-day, etc.), with some goofy effects between these points that mostly don’t matter unless there are really big, persistent oscillations in this frequency range. It’s a lousy approximation of a low-pass filter, but the raw data doesn’t contain clean enough patterns that you could notice filter artifacts at a glance. For smoothing away short-term noise you ideally want to use something like a Gaussian window instead of a rectangular one. That doesn’t tell you, “Has Cano ever had a worse 44-day stretch than this one,” though.
As Cano ages, his skills will diminish. If you truly/honestly assess this pattern, you ought to be a bit concerned. This is why the Yankees were unwilling to go 10 years on Cano. Smart for Yankees, good luck for Mariners!!
Big Cano fan from his NY days. He’s always been able to hit just about anything, which is part of the problem; he can swing at bad pitches. One of the biggest differences between this year and his career stats is the out of zone contact rate. It’s been 9 years since he was so low. Interestingly, he’s seeing more pitches, 3.66 vs. 3.42, so, with more 1st pitch strikes and fewer pitches seen when behind in the count, he may be getting anxious when behind and then swinging at bad pitches. Getting anxious and swinging at bad pitches was his downfall in the playoffs with NY.
Do good players that end their careers badly have a final descent into a valley from which they never recover, or do the alternating peaks and valleys overall slowly decline? I’d guess probably more of the latter.
Cano should rebound. The main question is whether each new peak will be less than those before and by how much. Alternatively, Cano could change his approach and derive value in other ways, as he did last year with less dingers but similar value.