Trade Deadline Target: Jake McGee

Hello, friends. The trade deadline is coming up again, and by virtue of being a contender for a play-off spot, the Cardinals are again looking to be buyers at the deadline. With the depth of the Cardinals, a position player upgrade will be tough even with Matt Carpenter and Brandon Moss temporarily out, especially in an extreme sellers market. And with the rotation currently trying to be the first since the 2003 Mariners to go the entire season with just five starters, there's no apparent opening in the rotation either. That leaves us with the bullpen.

The Cardinals like paying a low price for their deadline upgrades, which means they're also not usually acquiring the top end of the market. So while an upgrade like Andrew Miller is enticing, it would also be unlike the Cards to part with the future talent required to pry him from the Yankees. GM John Mozeliak's strategy instead, may be to try to buy-low, like he did in the all three deadline deals in 2016.

With that in mind, what relievers are there that (1) have been good in the past but have lowered their stock in 2016, (2) projects as better going forward, and (3) plays for a team that should be selling at the deadline? I can think of one such player, recently displaced Colorado Rockies closer Jake McGee. McGee was one of the best relievers in the game from 2012 to 2015, ranking 6th in fWAR with the only names ahead of him being a who's who of the games most well-known dominant closers: Aroldis Chapman, Craig Kimbrel, Kenley Jansen, Greg Holland, and David Robertson.

The Rays traded McGee to the Rockies this off-season for Corey Dickerson, with the teams also each swapping a prospect. McGee hasn't been the same dominant reliever in 2016 though. With a 2.33 FIP in 2015, the Rockies probably didn't expect that to continue at Coors, but they also likely expected him to be better than his 4.22 FIP he owns so far on the season. Crazy enough, at Coors that's still good for a 94 FIP- when adjusted to league and park. Regardless, that's up from 59 a year ago, so even when adjusting for park McGee has been much worse than anticipated. The projections see a change in talent, as his projected FIP is up from 3.13 at the start of the year to 3.49 now (remember, the projections know he plays for the Rockies, so a 3.49 is actually fairly decent).

Commonly, fans look to home/road splits for Colorado players, with the idea that their performance on the road is what should be expected if the player changed teams. This level of analysis hasn't been reliable though, likely in part to the short-term sickness people experience when they go from a high-altitude environment to a lower-altitude one.

If the team were to acquire a reliever, it would be optimal for him to be left-handed, like McGee. The team has two left-handers in the pen already, but one of them, Tyler Lyons, is used primarily as a long-reliever, and Matheny hasn't trusted him in close situations. The other,  Kevin Siegrist, has been strong at times, but exhibits a rare reverse platoon split acknowledged by Joe over the off-season. That reverse split has continued this season as he currently sports a 19.8% K-BB% against RHH and a 15.4% mark against LHH. Kevin's also in the middle of a down year, with his overall FIP up from 2.91 FIP in 2015 to 4.50 in 2015. Over his career, McGee has been death to lefties, while still posting strong results against righties. That has changed quite a bit in 2016:

You have to keep in mind he's pitching half of his games in a stadium like no other, but those marks against right-handers this year are still really bad. So what's the deal with McGee? Well first off, his fastball velocity is down, though it differs depending on the source. Brooksbaseball.net has him at 95.6 mph in 2015 to 94.2 mph in 2016. Fangraphs' pitch f/x supplied data has him down from 94.4 to 93.5.

How big of a deal is that? Well, it's definitely not good any time a pitcher experiences lower velocity. Looking at his fastball velocity by month, his velocity declined from 2014 to 2015 as well, and tailed off in late 2015. Maybe this a continuation of a trend of fastball velocity decline, maybe he's trying to pace himself through a full season better. It's certainly troubling though, and will require watching.

There also seems to be a problem with the curve ball.

It's lost a tick of velocity, and a little more than two and a half inches of drop. He's throwing it more often, likely because of the velocity loss on the fastball. There are numerous things that could be causing the decrease in whiffs per swing and grounders: the curve is worse in terms of velocity and both types of movement. He's throwing it more often, so hitters get more looks at it. The decline in fastball velocity means hitters are a little less geared up for the fastball, making it easier to wait on the curve.

McGee's curve was so good last year no one even hit it in the air, whereas this year it has a lower than average GB%. There's certainly a problem here. McGee is still locating the curve the same as last year: here's a heatmap of curveball location in 2015, and here's 2016. His vertical release point between the fastball and curve has widened since last year, but only by a very small amount, and it's a difference he's maintained in years prior to 2015 without a problem.

At this point you might shrug and say, "Well, the stuff is getting worse, so maybe that's the new normal for him", to which I say, "Maybe not!". When people talk about playing in Colorado, they often talk about how the high altitude affects curve-balls. Alan M. Nathan at The Physics of Baseball claims that the physics involved dictate that curves will break only 82% as much at Colorado.

With that in mind, I wanted to see how pitchers have fared in general when joining the Rockies. Pitch f/x stats go back to 2007. On fangraphs' leaderboards, I selected for all pitcher seasons, starter or reliever, where the player threw 30 or more innings. I then trimmed these down to pitchers who threw at least 5% curve-balls. I then found every instance in which a player threw 30 innings for a team that was not the Rockies in one season, and then threw at least 30 innings in a Rockies uniform the following year. That was only 14 pitchers, and here's how the vertical movement on their curve changed for each when joining the Rockies:

Only two pitchers have lost more drop on their curve than McGee when moving to Colorado during the Pitch f/x era. Strangely, some players have added vertical drop to their curve-balls when moving to Coors, which just shows you how wonderfully unpredictable baseball is. The spread of this chart shows that a lot more can happen to a curve-ball than just changing home parks, but on average this group lost about 80% of an inch in curve-ball drop. That's a neat find, I guess, but also not much compared to McGee's two and a half inch decrease in drop. It's safe to say from this analysis that the curve-ball probably has more going wrong for it than just the altitude difference.

I also did seasons in which a player pitched for the Rockies, then pitched for another team the following year:

In 2011, Ublado Jimenez threw over 30 innings for the Rockies, and then was traded to the Indians and threw more than 30 innings for them, so I just used 2011 for him rather than back to back seasons. Interestingly, this group gained almost 80% of an inch of vertical drop when leaving the Rockies, so maybe we're on to something here.

So, I do think that McGee can be better than this, and perhaps is hurt by pitching in Coors more than the average pitcher, as his secondary pitch is a curve-ball, which seem to be tough to throw well at high altitude. That velocity drop is a big warning sign though. The decline limits the bounce-back potential, and thus limits how much in value the Cards should be willing to pay.

The Rockies are clearly a bad team at present time and should be highly motivated to trade McGee, who is making $3.5M in his second to last year of arbitration eligibility. But the Rockies are also one of the toughest franchises to read, thanks to never appearing to have a cohesive strategy. They're a team that's not built to win now or in the future, and it's been that way for a while now. Also, the Rockies just traded for McGee last year, they likely value him higher than most, and might want to hope for him to bounce back for them rather than trade him at his lowest point. The analysis here at least points to a bounce-back not appearing very likely though. The curve is worse, and he's throwing it more often because the fastball is also worse. That's opposed to say, Trevor Rosenthal, who is having problems with command, both with the change-up and with the fastball, but has mostly the same stuff. Regaining command is easier than regaining stuff, so there's hope for Trevor that isn't there for Jake.

If the Rockies are panicking over McGee's decline and happen to like some Cardinals prospects that the Cards don't particularly like, perhaps a match can be made. It seems unlikely for those things to be true though. While I was hoping to find a bounce-back candidate in McGee, he looks more like one of the countless relievers who had several strong seasons followed by simply losing their stuff.  I think we can cross Jake McGee off the list of potential acquisitions.