Jarrod Dyson and Outfield Sprint Speed Aging

Based on media reports, the Orioles have been in communication with Jarrod Dyson, a fleet, left-handed outfielder who has excelled everywhere in the outfield.  The communication raised to a level that was serious enough for the Orioles to consult his medical files and find them to be supposedly unconcerning.  For several of us at the Depot, this is some good news.  Dyson was a major element of several of out plans for the offseason and a piece that I have thought fits the club well.  He provides speed, a good on base percentage, and a solid backup option to Adam Jones (and a potential solution in 2019 when the club might be Jones-free).

He has a couple major drawbacks.  First, he cannot hit southpaws.  This means that he can be a heavy component of a platoon (most pitchers are right-handed), but he is a liability to late inning relievers and is largely restricted to defensive replacement duty when a left hander starts on the mound.  He has been a very effective pinch hitter with a 134 wRC+ (over 24 PAs).  Second, he is getting into his twilight baseball years and will begin next season in his age 34 season.

While my preference would be a two year deal, I recently noted that I would be fine securing Dyson to a three year contract.  This resulted in a flourish of tweets telling me I was paying too little attention to his age and that my perspective was rooted too much into his defense, which was benefited by his speed.  To some extent, the narrative holds true when we look at his position based UZR/150s (runs saved):

Age	Sprint (ft/s)	LF	CF	RF
31	30	25.2	18.2	11.2
32	29.6	29	19.6	35.5
33	28.8	19	8.6

Dyson's defensive performance was otherworldly in his age 31 and 32 seasons.  However, he suffered a major decline in speed this past season.  The optimist would suggest that his hyperextended toe resulting from a July 22nd wall crashing and a sports hernia that was diagnosed August 19th led to his decreased speed and that we should expect a return to form with a healed hernia.  That optimistic view notes that speed tends to change mostly due to body morphology changes or accumulated hamstring and back injuries.  Toe injuries can be death, but it appears he has recovered from that and hernias tend to mend well.  Neither of these look like a chronic issue.

Edit: In the comments someone succinctly noted that I am not appropriately communicating the outcomes for sports hernias.  I will walk through the process. Before the article was posted, I did some research. Only player I found to have repeated sports was Scott Posednik who underwent five surgeries in the span of about three years.  It is uncertain to what extent that impacted him as he was still fairly fast, but less disruptive which may be due to a lingering groin issue or simple aging.  From there, I ask a person I went to school with who performs surgeries like this. He said that basically everyone is fine and able to compete in three months and that recurrence was rare, maybe 5 percent.  After getting the comment below, I sought him out again and he said that to suggest a sports hernia is a serious injury feels to him to be a rather outdated perspective and perhaps the individual is not current with modern therapies. However, he said that while he performs surgeries that he is not much involved in the rehabilitation process and follow up.  So...I contacted a friend of mine who works for an MLB team on their medical staff and is experienced with bringing players back from all sorts of injuries.  He stated that sports hernia surgeries are a very successful procedure with few recurrences.  He did not have a number on hand, but said that it has to be less than 10%.  After all of this, I have concluded that my original text is accurate and the commenter did not have a complete understanding of the outcomes for, at least, this demographic.

Overall how does speed change in outfielders as they age?

Using sprint speed data from 2015-2017 for outfielders, I was able to construct the following table.

Age	Sprint Change (ft/s)	Example	>28 change	Example
22		28.00
23	-0.03	27.97		30.00
24	-0.12	27.85	-0.05	29.95
25	-0.06	27.79	-0.02	29.92
26	-0.05	27.73	0.02	29.94
27	-0.09	27.65	0.01	29.95
28	-0.08	27.57	-0.03	29.92
29	-0.10	27.47	-0.04	29.88
30	-0.12	27.35	0.00	29.89
31	-0.10	27.25	0.10	29.98
32	-0.11	27.14	0.08	30.06
33	-0.11	27.03	0.03	30.09
34	-0.06	26.97	0.15	30.24
35	-0.14	26.83
36	-0.11	26.72

What the table shows are two situations.  The second and third columns explore what happens to the population in general in terms of how speed changes from one year to the next as well as what happens to an age 22 runner who sprints 28 ft/s as he ages.  The second take a subgroup out of that population to see what happens to plus runners (>28 ft/s) as they age with a value that is more like that value (30 ft/s) for a 23 year olds starting point.

What the data appears to indicate is that the population of outfielders in general see their speed degrade with time.  Someone in Dyson's situation would be expected to lose about a third a foot per second as he aged over a three year deal when fit to that model.  If we assume that 28.8 ft/s is his real talent level, then that would be a 28.5 ft/s.  That would have made him tied with the fastest Oriole on the roster in 2017 (Craig Gentry).

For the faster group, the data does not go as far.  I did not feel comfortable extrapolating out to where I did not have five or more data points to consider, so the results above would only consider Dyson's first year.  What I find interesting about the results though is that elite runners tend to remain elite runners.  In other words, if you have a body type that is fast then you tend to remain fast.  We see this pretty often that while world class sprinters tend to peak in their mid to late 20s, but they retain that speed into their late 30s.  True, they rarely challenge for medals in their late 30s, but they only lose a stride or two over several years.  That means a great deal on the track, but perhaps not as much out on the ball field.

So, sure, maybe offering a three year deal to a speedster entering his mid thirties is something that should be avoided, but I still think the gamble is fine.

Trey Mancini is as Fast as Adam Jones

Baseball is a rather wonderful sport with a variety of ways to appreciate it.  It can range from a pure reactionary level of enjoyment with bias expectations at every turn, the pure notion of fanaticism, to heavily entrenched awareness of historical outcomes and uncertainties related to those potential outcomes.  One elusive area for many who appreciate the game has really been the scouting side.

Scouting to many is a black box.  A scout watches a player, tries to often qualitative evaluate ability and future ability.  That then is conveyed down a path to the public at large.  It can be frustrating to try to understand the process.  It is also curious how the public sphere of scouting resembles an echo box of scouting writers hitting and missing on the same exact players where one might consider there to be more variation.  Largely, the public industry has become rankings and enough of a description to be able to tweet out or participate on a message board with some manner of scouty credibility.

What has improved over the year is the technology and the ability to access that technology.  We can easily acquire fastball velocity, movement of pitches, and flight paths of balls.  We can see how able a player is at covering ground and making catches in the outfield.  We can see a player's coverage at the plate and exit velocity of batted balls.  We also can now see maxed out running on the basepaths to give us a good idea on player speed.

In the past, we have largely relied on play by play derived measures to develop speed metrics.  We look at stolen bases and caught stealing, we look at defensive range in the outfield, we compare ability to advance on batted balls, and how well a player stretches out hits.  Now, we have another tool in our tool belt and it does something grand: it measures a player's top speed on the basepaths.

Statcast measures sprint speed, the top speed a player achieves when maxed out running the basepaths.  For reporting, MLB requires ten events of maxed out running and calculates top speed by the fastest second.  What this can miss is acceleration.  Players may be able to reach top speed more quickly than others in the course of a 4 second or so run, but this metric gives us a good idea of top end running ability.

By looking at all qualified players this year, we can devise a frame of mind for a 20-80 scale.  My scale is based on the assumption that the average qualified baserunner is a 50 score.  I also assumed a more traditional take that every increase in score of 10 is equivalent to a standard deviation of the population.  We we wind up having is an average speed for a MLB player of 27.1 ft/s and a standard deviation of about 1.2 f/s.  This gives us the following tool grade table:

Grade	Speed (ft/s)
80	30.7
75	30.1
70	29.5
65	28.9
60	28.3
55	27.7
50	27.1
45	26.5
40	25.9
35	25.3
30	24.8
25	24.2
20	23.6

Byron Buxton has recorded the highest average sprint speed with 30.2 ft/s, so this scale would see him as a 75.  As you would expect the possibility of a player being three standard deviations from league average would be pretty astounding, so even a player like Buxton is unlikely to be rated an 80 based on this methodology.

However, that is not true for 20 grade speed.  We actually have four who quality as 20 grade speed when rounded: Miguel Montero (23.8 ft/s), Juan Graterol (23.4 ft/s), Brian McCann (23.3 ft/s), and Albert Pujols (23.0 ft/s).  Graterol, McCann, and Pujols actually are recorded below the 20 rating, which shows that an extreme lack of speed can be made up for with other qualities and context.  Largely, being a designated hitter or catcher where speed is not completely required with some element that past success and a big contract can keep you on a roster for awhile.

Where do the Orioles stack up?

	Speed (ft/s)	Grade
Gentry, Craig	28.5	60
Beckham, Tim	27.5	55
Rickard, Joey	27.5	55
Jones, Adam	27.1	50
Tejada, Ruben	27.0	50
Mancini, Trey	27.0	50
Schoop, Jonathan	27.0	50
Machado, Manny	26.9	50
Smith, Seth	26.4	45
Trumbo, Mark	26.3	45
Hardy, J.J.	26.2	40
Joseph, Caleb	25.8	40
Castillo, Welington	25.1	35
Davis, Chris	25.1	35

I think a couple things jump out to me.  Trey Mancini may be able to unlock some potential in left field if he can figure it out.  I have been getting negative reports on his fielding ability and that his athleticism is decreasing.  However, if he can maintain a 27 ft/s sprint speed for a few years and improve through experience in left field, then he can be an average defender out there instead of the mess (with infrequent very nice plays) that we have witnessed this year.

The second thing that popped out to me is that the Orioles are not exactly a club that values team speed.

For the most part, the above graph does not show that speed means success.  Speed often means being terrible because speed is a young man's tool.  Fast teams are typically young teams and teams who employ young players tend to be terrible teams.  Still, one might be concerned with how uniquely slow the Orioles are.

Thinking long-term, I wondered how the Orioles have faired over the past few years.  Particularly, I wondered how established future Orioles have measured up.

	2015	2016	2017
Tim Beckham	27.6	27.7	27.5
Adam Jones	27.7	27.6	27.1
Jonathan Schoop	26.4	27.1	27.0
Manny Machado	27.7	26.6	26.9
Mark Trumbo	27.1	26.7	26.3
Chris Davis	26.1	26.2	25.1

The sprint speeds tend to follow what we would expect.

• Beckham shows solid average speed for a shortstop or second baseman, wherever he may finally wind up.  He does not shows any decreasing trend.  
• Adam Jones went from about what one would consider a 55 rating to a 50 rating, which is a sizable drop.  He comes off as the fifth slowest CF in baseball for 2017.
• Schoop's depressed 2015 was likely due to his knee injury.  On the slower end of second basemen, but not exceptionally so.  He also looks to have below average, but acceptable speed for a corner outfield position.
• Machado is what we generally expected. He has bulked up and he is not the guy who stole 20 bases a few years ago.  He is settling into the 45/50 range, which he should maintain through his prime years.
• Trumbo's numbers are what we have seen in the outfield this year.  As a RF, he would be tied for second slowest (with Nick Markakis) and just a tad slower than Seth Smith.  Although he seems to hit better when playing in the field, it seems the corner outfield will not be an acceptable place for him.
• Davis dropped from slow to dreadfully slow this year.  That has shown up in his 2B:HR rates and his decreased movement around first base.  That level of speed erosion is troubling for a player who is in the second year of a very long contract.  The hope here is that Davis has a lower body injury that will heal this off season, but I have heard nothing about that.

Statcast's Sprint Speed will be something we should take a note of in the years to come to get a better handle on a player's physical abilities as well as maybe being an indication for injury.