Batter's Box Interactive Magazine Batter's Box Interactive Magazine Batter's Box Interactive Magazine

By the time a professional baseball player has played a few seasons of minor league ball, we're usually able to get a sense of how good a prospect he is. But to determine what kind of player he's likely to end up takes more than a cursory look at the numbers he's posted at each minor league stop. For more precise and fruitful analysis, we need to contextualize and aggregate the data. Once we do that, trends emerge.



Take the case of a journeyman minor leaguer like Simon Pond. Pond has had a varied career in 3 organisations, the last year and a half spent in the Toronto minor league system. Here is a list of stops he's made in his pro career to date:

   Montreal Expos


1994Gulf Coast Expos (R)
1995Gulf Coast Expos (R), Albany (A)
1996Vermont (A-)
1997Cape Fear (A)
1998Jupiter (A+), Harrisburg (AA/2 games)
1999Jupiter (A+)
2000Jupiter (A+)

   Cleveland Indians


2000Kinston (A+)
2001Kinston (A+), Akron (AA)

   Toronto Blue Jays

2002Dunedin (A+)
2003New Haven (AA), Syracuse (AAA)

My approach to analysing professional players is to focus on core skills. I believe there are 4 core hitting skills that may be approximated by looking at official batting stats. They are: 1) hitting for power [appoximated by the formula: (2B+3B+2*HR)/(AB-K+SF)]; 2) line-drive hitting [approximated by ball in play average - (H-HR)/(AB-HR-K+SF)]; 3) drawing walks [walks per opportunity less intentional walks - (W-IW)/(PA-IW-HBP)] and 4) avoiding strikeouts [K/(PA-IW)].

The next step is to contextualize the performance. Since park data for nearly every level of minor league baseball is hard to come by, I normalize to the average performance in a given year in the appropriate minor league. Excepting some cases like El Paso and Colorado Springs (which figure to have a large impact on batting stats in relation to the other parks in that league), normalizing to league works well because whatever effect the home park has relative to the other parks in the league is counterbalanced by the road parks.

The final step is to aggregate the data - to assemble it into larger, yet still meaningful, samples. Sample size is always an issue when evaluating a player, which is why every line of hitting data should include Plate Appearances.


ORGANISING DATA BY LEAGUE/LEVEL

In cases where a player spends parts of more than one season in a particular league or at a particular level, it may make sense to aggregate data by league and/or level.


LeagueAgeYearsPApowernorm$BIPnormwalksnormKnormobpslg
SouthAtlantic20.4 95,97 574.048-59.295-2.070-15.124+39.318.309
FloridaSt22.4 98-00 961.101-12.279-9.082-5.161+8.322.348
Carolina24.0 00-01 374.170+43.371 +24.084-5.164+18.388.500
Eastern25.5 01,03 700.165+23.325+9.089+13.149+20.367.469

Notes: Age is the average age, weighted by plate appearances; norm is the percentage above or below the league average performance [plus(+) always indicates a better than average performance, i.e. fewer strikeouts, more walks].

We can aggregate the data a bit more by grouping the Florida State and Carolina League data (since they are regarded as equivalent in terms of difficulty). The farther the data recedes into the past, the less relevant it is for current ability, so I will drop the South Atlantic League data from the following chart. It may be useful to sort out this season's performance, since Pond has hit much better than ever before.


LevelAgeYearsPApowernorm$BIPnormwalksnormK normobpslg
A+23.5 98-02 1792.134+16.307+2.081-7.162+10.345.414
AA (pre 2003)24.6 98,01 432.166+22.296 -1.068-9.166+14.319.440
AA/AAA (2003)26.7 2003 396.177+35.353 +18.102+24.116+33.414.522

Simon Pond seems to have greatly improved his strikezone judgement, displaying a large increase in his walk rate while striking out less often. He didn't do much of anything until he arrived in the Jays' organisation, which suggests that Toronto has had a hand in his dramatic improvement this year.

Nevertheless, the first and most likely hypothesis that presents itself is that Pond is playing over his head. If so, then he will ultimately be of lttle interest to the Blue Jays. If his performance in 2003 turns out to be legitimate, it suggests that Pond can be something like an average major league hitter (with defensive limitations, it should be noted). A September call-up would be a just reward for this minor league vet's perseverence.




A Minor League Career Analysed | 9 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
_Chuck Van Den C - Tuesday, July 29 2003 @ 08:15 PM EDT (#65465) #
I hope that Paul DePodesta writes a book one day, revealing all the stuff he's worked on over the years in Oakland.

I fear that the passage for the past two weeks would read something like: "Have temporarily lost interest in baseball. Have found Minesweeper especially addictive. Am leaving research and contract negotations of present personnel, specifically Scot Hatteberg, to interns. Trust all will go well."
Coach - Tuesday, July 29 2003 @ 09:13 PM EDT (#65466) #
John, there's no guarantee that a hitter with lousy plate discipline is going to be coachable. Ricciardi's player development is about minimizing risk and expense, so they draft guys who already "get it," rather than hoping they can fix someone who might never understand.

It's a pleasant surprise that some of the inherited "tools" guys have been able to make adjustments, but no matter how good the teaching, there's no way of knowing in advance who the good students will be, and J.P. is always trying to work the percentages in his favour.

No need to provide translations, by the way -- long before the advent of English Mass, I was an altar boy, and for some reason, I took three years of Latin at St. Mike's.
_John Neary - Wednesday, July 30 2003 @ 11:26 AM EDT (#65467) #
By the way, Pond's MjEqAs so far this year are .270 in AA (271 PA) and .263 in AAA (127 PA).

These translations need to be taken with several moles of NaCl. Nevertheless, here's a list of some Jays and their current EqAs:

Wilson .276
Johnson .276
Pond (translated) .268
Phelps .267
Hinske .264
Catalanotto .263
Woodward .256
Hudson .255
Bordick .250
Berg .249

Please note that I in no way mean to suggest that Pond is as valuable a player as the other guys on that list, nor that he has proven that he can hit as well as them. I present the list merely as a statistical curiosity, and you can interpret it at your own risk.

I will, however, say that I'd rather have Simon Pond on my bench than Tanyon Sturtze in my bullpen.
_Jordan - Wednesday, July 30 2003 @ 12:45 PM EDT (#65468) #
I'd rather have David Hasselhoff on my bench than Tanyon Sturtze in my bullpen. But I agree, John.
_Lurch - Wednesday, July 30 2003 @ 01:30 PM EDT (#65469) #
Hasselhoff could sing to keep up clubhouse morale.
_Brian - Sunday, March 14 2004 @ 11:04 PM EST (#65470) #
Moles of NaCl - glad I took grade 13 Chem or else I might not be able to understand you guys.

Pond's spring has been great so far, so maybe the Jays have a career minor leaguer who finally has what it takes to play with the big boys. As a Jays fan, it's nice to see so much promise in the minor leaguers.

BTW: If you're going to pick a Baywatch star to be on your bench, why not make it Pamela Anderson or something? Just curious.
_Jurgen - Saturday, May 15 2004 @ 02:03 AM EDT (#65471) #
I got into a bit of an argument with Aaron about this, but why is avoiding strikeouts an important tool for analyzing performance if the growing evidence suggests that the K is just another out? Intuitively, I'd only be alarmed by rising K rates if a player's power and/or walks were dropping, too.

Am I completely off base?
Mike Green - Saturday, May 15 2004 @ 04:30 PM EDT (#65472) #
Jurgen, here's the way I think about it. A strikeout is not like any other out (popouts excluded), in the same way that a home run is not like any other hit. The result of balls in play is effected by the hitter (line drive/ground out/fly ball), the defence and blind luck.

I'm inclined to treat variations in $BIP over seasons with suspicion that they result from blind luck whereas I'm more inclined to see changes in $HR and $K as indicative of development (good or bad) because these outcomes are more directly within the control of the hitter.
_Jurgen - Sunday, May 16 2004 @ 12:51 AM EDT (#65473) #
OK, but...

Look at Jim Thome. Thome strikes out nearly 170 times a season. But he walks like a crazy mofo, and when he makes contact, holy crap, he does good.

Would Thome necessarily be a better hitter if he cut down on his strikeouts? I'm not sure the answer is simple. I'd want to know whether there was any change in his patience and power before I answered that question. It seems like strikeouts are only important in relation to the three other skills Robert mentions, whereas those other skills have more intrinsic value. Yes, they should all be considered in relation to one another... but strikeout rates more so than the rest.

That's how I see it.

But I'm willing to be unconvinced.
A Minor League Career Analysed | 9 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.