Prospect Speed Analysis

Speed is a fascinating aspect of baseball. On top of the simple notion that watching blazing runners is fun, speed plays a key role in what position a player ends up at and the kind of base runner he is.

I've spent a good chunk of time trying to use statistics to quantify speed this offseason. My approach has largely been trial and error but I wound up with something that I later found out is not all that dissimilar from Bill James' speed score.

Like James, I looked at the frequency a player attempts a stolen base, how often he is successful, double plays per at-bat, and triples per at-bat. I later added some polish by 1) dividing steal attempts by roughly how often the player reached first base and 2) subtracting home runs and strikeouts from at-bats, an effort to isolate the outcomes I was after.

I now have my own speed data on 200 very good prospects. Let's put it to use:

 

The Fastest Prospects

Because the indicators I used to calculate speed become tougher to excel in as runners face better competition, I separated the groups below into upper and lower minors. Only players with more than 250 plate appearances were considered -- Michael Bourn would have ranked second on the upper minors list. H.L. stands for high level and represents the highest level where the player has registered 100 plate appearances. The colors in the speed column go with the rubric posted toward the end of this article.

Velez could receive regular at-bats with the Giants perhaps even beginning opening day. He attempted to steal approximately 60.5% of the time he reached base last season and was successful 77.3% of the time on his way to 58 stolen bases.

It's noteworthy that few of the hitters above have had much success hitting for power. So their contact abilities and patience at the plate -- along with their defense -- will likely decide how valuable they could be to a big league club.

 

While most of the players on this list are still a ways away from contributing in the bigs, Maybin figures to be the Marlins' opening day center fielder. And unlike many of the players on the list above, Maybin has shown flashes of serious power. He may have the most power/speed upside of any prospect in baseball.

 

Prospects Who May Change Position

Shortstop, second base, and center field are all positions that demand great speed and instincts. Here, a player without above-average speed will likely provide below-average defense. Listed in order of the likelihood they'll change positions, I've noted some elite prospects who may have to move due to a lack of range. I'm going to hold off on prospects who had fewer than 200 plate appearances last season -- like Matt LaPorta.

 

Fernando Martinez, CF NYM - He's currently below-average base runner who will need to make some major instinctive improvements to stick in center field. But he's young and the Mets may want to give him another year to prove himself.

Jed Lowrie, SS BOS - When I saw Lowrie play in college he was a second baseman. Now he's being tested as a shortstop in pro ball. If he continues to play short, Lowrie's lack of speed will likely make him a below-average defender.

German Duran, 2B TEX - A fringe-average runner by my calculations, Duran may need to rely on exceptional instincts to play second base in the big leagues. He grounded into a lot of double plays and attempted relatively few stolen bases last season.

Carlos Triunfel, SS SEA - Similar to Martinez, Triunfel's score may be limited by poor instincts more than lack of speed. But I have him as an average runner and there aren't many average runners who make good shortstops.

 

Other Worthwhile Notes

I broke hitters down into six speed groups: Crazy Fast, Burners, Above-Average, Average, Below-Average, and Base-Cloggers. Here's some information about each group:

• Half of the crazy fast guys play center field while the other half play second base.
• Every burner plays shortstop, second base, or center field.
• Just five of the 34 above-average runners in my population (14.7%) play positions besides SS, 2B, or CF -- the fastest of those runners being Cody Johnson (LF, ATL).
• Fifty percent of the average runners play SS, 2B, or CF.
• Less than a quarter of the below-average runners play SS, 2B, or CF.
• Four of the 48 base-cloggers are 2B (3), SS (1), or CF (0) -- that's 8.3%.

I also created a crude rubric on each speed category with stolen base estimations and positional expectations:

I created the groups based on deviations from the mean (14.26). The standard deviation was 16.83.

 

Methods

Like a lot of my work, this study was more of the back-of-the-envelope variety than a formal mathematical exercise. My formula isn't close to textbook worthy -- Speed = (100*SBr)+6*(SB%-CS%)+(300*3B/AB)-200*(GDP/AB) where SBr is stolen bases per times on first base.

Note that I did not include times reached on errors or fielder's choice -- don't have access to that kind of data. And I subtracted strikeouts and home runs from the "at-bats" I used to divided triples and ground into double plays by. Here's some more information about the variables I utilized in my formula:

• My average SBr is 16.5%. Three players with over 400 plate appearances topped 50.0% in this category: Eugenio Velez (60.5%) -- as mentioned above -- Eric Young Jr. (54.2%), and Manny Burriss (51.5%).
• My average SB% is 69.7%. Michael Bourn, who spent all of 2007 in the big leagues but didn't have enough at-bats to lose "prospect status", was successful in 94.7% of his steal attempts.
• My average CS% is 29.7%. Travis Snider was caught in 76.9% of his attempts.
• Tony Thomas, who only had 233 plate appearances, tripled in 6.8% of his at-bats. My average 3B/AB is 1.3%.
• And lastly, my average GDP/AB is 2.4%. Jesus Montero grounded into double plays in 8.1% of the "at-bats."

I chose not to include runs scored in this study. James did. And this is one of the main differences in our speed approaches. Runs are too team dependent for me.

 

Potential Problems

My sample size is pretty limited. I only looked at 200 prospects and that group is largely arbitrary.

I also only used one-year samples. I considered using two-year samples but many of the players I was looking at don't have two complete seasons worth of data. Also, quite a few of them had fewer than 300 plate appearances in 2007. Another problem with one-year samples is the variation in GDP/AB. A single GDP can have a large affect on that rate, as most of my population grounded into fewer than 10 double plays last season.

Injuries can also play a big role in a player's year-to-year speed. Drew Stubbs, for example, played through turf toe last season and it's fair to assume the injury had a significant impact on the indicators I tracked for my speed score.

Young players who are moved through the minors quickly may have the speed to be good base runners but lack the experience to take advantage of that speed. I've seen Jay Bruce run and I think he's an above-average runner, though my speed score merely rates him as average.

I rewarded players heavily for their stolen base success percentage. Cody Johnson wasn't caught stealing once last season, but he only attempted seven steals.

I did not make any adjustments for handedness. Though lefthanders should be able to reach first base faster than righthanders, they also tend to hit more ground balls to the right side of the infield.

 

Want my speed score for one of your favorite prospects? Email me at adamf@projectprospect.com.