Kevin Kiermaier and the Oliver projection system

Brian Blanco

Sometimes, things come as a complete surprise to all. Sometimes, they come as a complete surprise to all but a few.

The following is a question and response from a Baseball Prospectus chat back in February:

"Kevin Kiermaier is projected as a 5 win player by one system. Say what?" -- Dee Rae from Bay Area

"When we have driverless cars, there will be far fewer accidents, less traffic, more efficient use of the streets and the cars themselves, but we'll still have to sit semi-alert in the driver's seat." -- Sam Miller

Sam is one of my favorite baseball writers, and -- of course -- he's right on the money here. He found a clever way to say that while projections are a terrific tool, they're a system that does not have a dynamic human component. And using a projection tool means having to occasionally be ready to pass your own judgement, in case of an emergency.

Those who build a system to project baseball player performance don't run their projection systems -- the one they've spent years honing and perfecting -- then see something weird and then say "whoops, better change that!" You have to let the outliers and the unexpected results ride, because while a system may have faults, the projection system -- if it's a good one -- is based on data, science, and reason.

So, this was a classic example of when a projection system has limited data about a minor league player, and yet the system extrapolates good or excellent major league performance from that limited data set.

The projection system in question is called Oliver, and was developed by Brian Cartwright and originally published at The Hardball Times in 2010. Oliver is actually one of the top-performing projection systems (it's pretty good at predicting wOBA, according to Geoff at RotoValue), so we're not talking about something that's put together in a half-hearted way, here. We're talking about a well-respected system of forecasting.

Oliver, which tends to create projections based on 600 plate appearances, even for minor leaguers who aren't expected to sniff that level of playing time, does occasionally give us pretty solid numbers for minor-leaguers, based on minor-league equivalencies, and due to the fact it includes a fielding component, and a full FanGraphs Wins Above Replacement (fWAR) projection.

And that brings us to Mr. Kiermaier. Kevin Kiermaier rated a full Oliver projection, one that put him at 4.9 fWAR based in large part on the strength of his defense, as well as a solid projected wOBA of .306.* For the record, a single season fWAR of 4.9 puts a player at about an All-Star level of performance. In 2013, Jayson Werth and Starling Marte had fWARs of 4.6, and they each had terrific seasons.

* - For comparison's sake, Oliver projected a .310 wOBA for teammate David DeJesus, and a .309 wOBA for teammate Yunel Escobar.

But even this by itself wouldn't exactly be a shocker. Pittsburgh Pirates uber-prospect Gregory Polanco, who has played quite well since being called up this season, had a projected 4.6 fWAR by Oliver. Polanco was generally considered a consensus top prospect, and a guy who could have an immediate impact in the league.

The thing that made Kiermaier's projection so strange is that he's certainly not a top prospect. Kevin Kiermaier is a former 31st round draft pick, who -- after a pretty-good-but-not-great 2013 -- vaulted up prospect lists to be Chris St. John's consensus No. 11 prospect ... in the Rays' system. And while that does qualify him as a prospect, even a guy to watch, if scouts thought he could be a five-win player in the majors, he'd be the No. 11 prospect in all of baseball, not in just the Rays' system.

Of course, Kiermaier has played quite a bit for the Rays this year, thanks to Wil Myers' injury and a lack of overall quality performance in the Tampa outfield. And if Kevin Kiermaier keeps up his current level of performance (spoiler alert: he won't), he'll be worth somewhere near 4.5 fWAR by the end of 2014, which really isn't far from what that "one system" projected him as. Kevin Kiermaier has already been worth about 1.5 fWAR.

Generally speaking, while it's not insane for a rookie outfielder to put up a gaudy WAR in a partial first season (see Yasiel Puig), most rookies simply don't perform that well out of the gate -- especially when they're not a Trout- or Harper-level prospect. But Kiermaier has hit for massive power, gotten on base at a good clip, and recorded solid defensive numbers ... in an incredibly small 86-plate-appearance sample.

Though he's hitting for more power (more everything, really) than anyone expected, the defense he's shown isn't that much of a surprise. And this is the part that Oliver absolutely nailed.* Kiermaier has had a reputation for excellent defense in the minors. Baseball America had Kiermaier as the best defensive outfielder in the Rays system in 2012. In 2013, BP quoted a scout as saying Kiermaier had a "Brett Gardner-like toolset", which is a great way to give your team tons of under-the-radar value. Oliver was able to give Kiermaier the appropriate credit for his stellar defense.

* - Thanks to Christopher Long, who pointed out on Twitter I may not have been giving Oliver enough credit for being able to see Kiermaier's defensive value.

But we see Kevin Kiermaier's performance come out of nowhere (because no one was talking about him before the season aside from Dee Rae) and we say "well, no one saw that coming," and one might even say that's half true. But Oliver sort of did see it coming -- as it saw a minor leaguer whose bat could play up to something close to league average at the highest level, while providing solid defense. And while the bat (and glove, likely) have outperformed to this point, it behooves us to remember that sometimes these crazy things that a system spits out are actually done for very good reason.

This also, kind of, ties in to a recent piece by Mitchel Lichtman (aka MGL) over at his blog where he shows the math as to why we should count on pre-season projections, and not the last two months of performance, when projecting the rest of a player's season.

Projections might be even more important than we already think. And while MGL's research cited Steamer (a very good, perhaps best-of-breed projection system) instead of Oliver, it behooves us to respect, and to carefully consider what -- and why -- the projection systems are saying.

. . .

All statistics courtesy of FanGraphs and Baseball-Reference.

Bryan Grosnick is the Managing Editor of Beyond the Box Score. You can follow him on Twitter at @bgrosnick.

X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9351_tracker