Matt Kemp has "improved his physical conditioning, having lost 40 pounds," read a story here on MLB.com when Kemp reported to Dodgers Spring Training in February.
You're forgiven if that didn't exactly fire you up about Kemp's 2018 potential. After all, a player claiming he's shown up at camp "in the best shape of his life" has become something of a running joke in baseball circles. It's become such a meme that actual studies have been done on the phrase, showing very little on-field impact for hitters claiming to have improved their conditioning.
And don't forget this: Kemp showed up at Atlanta camp in 2017 with a version of the exact same story. He followed it with his worst full season in the Majors. You hear this talk every spring. It rarely matters.
Now, with all that said, in Kemp's case, early in the season, there's maybe at least something to this. Kemp looks better. He's hitting better (.340/.382/.580 in 55 plate appearances entering Saturday). He's moving better -- and, according to our recent rollout of 2018 Sprint Speed leaderboards, the data backs it up, too. Of nearly 300 players who qualified in both 2017 and '18, no one has increased his speed more than Kemp.
Here's what we mean by that. The way Sprint Speed works, it's measured in feet per second on what we consider to be "qualified runs." (You can read more about it here.) The Major League average is 27 feet per second, so Kemp's 26.6 ft/sec isn't exactly speedy, but he's gone from being basically the slowest left fielder in baseball in 2017 (24.9 ft/sec) to being only slightly below the positional average in '18. It's a sizable jump, the biggest of any qualifier in baseball.

Now, that list should tell you two things. First, with rare exceptions, most of these guys had below-average speed last season, giving them room to improve. Someone like Byron Buxton (30.7 ft/sec in 2017) can't really get faster, he can only hold steady (30.5 ft/sec in '18) or go down.
Second, you can also see that a lot of the players on this list who were slower last season had some pretty compelling health-related reasons. We should be careful to note that it's still early and that we're somewhat fitting a narrative to an outcome here, yet there's enough signal through the noise to believe that there's some evidence even before the first month is done.
For example, Mike Moustakas worked hard this winter to improve his conditioning after a disappointing free-agent experience, and he is also another year removed from his 2016 knee surgery.
Jose Pujols worked hard, as well, knowing the arrival of Shohei Ohtani would force him to play more first base. (He's already played three more games at first than he did in all of 2017.)
So, of course, did Kemp.

There's also simply the matter of health, or lack of it. Adrian Gonzalez (back), Jarrod Dyson (foot, hernia), Luis Valbuena (hamstring), and Jean Segura (ankle) all had lower-body concerns in 2017 that presumably haven't been issues in '18. (Cheslor Cuthbert also missed time, but it's hard to pin a wrist injury to foot speed.)
At the other end of the scale, one of the biggest speed decliners is Adam Eaton, who was running at a very strong 28.7 ft/sec in 2017 before missing most of the season after undergoing surgery on his left knee and landing on the disabled list again this year with a bruised left ankle. His Sprint Speed early on has fallen more than two ft/sec, to 26.5 ft/sec.
That importance of health applies to Kemp, too, since he landed on the DL twice in 2017, in April for a left hamstring injury and in July for a right hamstring injury. Those were just the latest in a litany of lower-body injuries, including left ankle surgery and a right hamstring strain in '13, and pulls to both hamstrings in '12. The speed that once allowed him to steal 34 bases or more in three separate seasons is long gone.
Since Statcast™ came online only in 2015, we can't compare the most recent versions of Kemp to his earlier heyday. But with the data we do have, it's clear to see that whether due to age, injury or conditioning, his Sprint Speed declined from '15 to '16, and again from '16 to '17.
Matt Kemp Sprint Speed, 2015-18
2015 -- 25.9 ft/sec
2016 -- 25.3 ft/sec
2017 -- 24.9 ft/sec
2018 -- 26.6 ft/sec
Major League average -- 27 ft/sec

The downward trend is clear, though it's interesting to note that in two of his previous three seasons, Kemp's fastest months have been April. We know that injuries were to come in previous years, though there may also be something to be said about feeling fresh early before the weight of the long daily grind takes its toll.
That said, we've tracked more than 1,000 Kemp runs over the last three-plus seasons, and two of his fastest five individual runs have come in the last two weeks.
Matt Kemp fastest individual Sprint Speed runs, 2015-18
30.9 ft/sec -- April 12, 2015 (triple)
29.8 ft/sec -- April 12, 2015 (groundout)
29.7 ft/sec -- April 29, 2015 (double, to third on the throw)
29.2 ft/sec -- April 16, 2018 (single)
29.1 ft/sec -- April 10, 2018 (double)

In theory, improved speed should help on defense, too, though it's a little early to draw any solid conclusions. What we can say is based on the balls hit to Kemp this season, an average outfielder would have been expected to make the play 94 percent of the time, and he's made the play 95 percent of the time, so about as expected. Last year, he underperformed by 10 points (expected to catch the ball 86 percent, made the catch 76 percent) and the year before, he underperformed by 7 points (expected 84, actual 77).
It's still so early for Kemp, and the possibility of injury or fatigue is a very real concern. But if he's not in the "best" shape of his life, it's perhaps at least better compared with the last few years. For a player generally seen as well past his prime, acquired in what was more an accounting move than anything, Kemp is off to a good start. He's been shockingly useful, based on the eye test. The early data says it might be legit.
For now.