Is Prospecting Improving?

A true minor league system is something that separates baseball from most professional leagues, although other leagues have tried, baseball's system remains the best and most logical around. The National Basketball Association has the National Basketball Development League, but most casual fans can't even name their team's affiliate, and the professional team seemingly has little control over the geography of their team unless they purchase one themselves like the Los Angeles Lakers did. The National Football League had the National Football League Europe season for more than a decade before folding following the 2007 season, teams were granted roster exemptions for sending players to the league, but usually the players were practice squad bait at best.

The system that most closely is tied with baseball's system is the National Hockey League's - teams take full advantage of foreign leagues as well by stockpiling young Europeans in their homeland until they appear ready to take their game to North America. However the teams are no where as frequent or deep as most minor league baseball teams, and the fascination with the minor league youth doesn't seem as great in the hockey community as the baseball community.

Baseball America for example publishes annual top 10 lists for each team, a handbook with 30 players per team - totaling 900 players - and a top 100 list covering the best and brightest youth across the league. The publication also does a subscription service for year round coverage of the minors, college, and high school ball which draft lovers flock to.

Former ESPN minor league expert John Sickels self publishes his own annual handbook: The Baseball Prospect Book in which he gives his own view on over 1,000 players, assigning each a letter grade and focusing on that grade rather than the actual rank of the players. Along with the guide Sickels publishes a subscription based newsletter where he unveils his 50/50 prospect list - a combination of the top 50 pitching and hitting prospects. His site, Minorleagueball.com draws over 9,000 visitors daily and readers of the site have the ability to post their own lists and opinions on anything prospect related to which they take full advantage of - community prospect lists, top 100, 200, even upwards of 500 lists, and of course countless ads for fantasy keeper leagues and prospect oriented websites alike.

There are countless other websites and books that judge prospects but for our experiment we're going to use the Baseball America top 100 lists. A few reasons for the choice: the lists are completely free and available to the public, the publication features a large staff and therefore gives more variance in the opinions put out and less room for bias, and presumably the staff has more access to live reports with the time to digest more information.

Using data since 1990 we'll simply place the value of each player's career OPS+ or ERA+, two statistics which account for all-time performances to judge each player, after assigning the values they'll be combined and averaged out. Obviously not every player on the lists will have made the major leagues; those players will be assigned a zero in place of their value. The goal is to establish whether with improved means of judging prospects - through technology and innovative thinking about players' in general - is prevalent in accurately predicting 100 players who will at least develop into players of average value.

For space and time's sake only 1/4th of each list was included; we're only going to look at the top 25 players of each list through 2005. For each list we're looking to see which of the top 25 gets the closest to being at least 100, meaning completely average, everyone with a value lower was below statistical average, and over 100 means, obviously, above statistical norm.

Year    Average
1990    89.44
1991    96.16
1992    88.88
1993    88.96
1994    102.00
1995    99.96
1996    96.92   
1997    104.08
1998    89.20
1999    97.24
2000    79.40
2001    88.96
2002    89.16
2003    100.84
2004    94.60
2005    89.36

There seems to be no clear correlation of time and average player value raising, but a question one could follow up could be have the positions of the players gotten more accurate, in this case has Baseball America been accurate enough with the number one picks that they are the absolute best of the combined rankings?

Number    Average
1    111.06
2    99.38
3    106.69
4    99.31
5    101.63
6    90.31
7    96.44
8    86.19
9    82.56   
10    112.63
11    104.25
12    99.00   
13    111.19
14    86.34
15    97.31
16    95.31
17    88.75
18    99
19    98.50
20    75.75
21    74.06
22    84.63
23    74.69
24    79.81
25    81.25

Again there doesn't seem to be a high correlation of averages, although it's interesting to break the list down into fifths for a more concentrated look:

Group    Average
1-5    103.61
6-10    93.63
11-15    99.63
16-20    91.46
21-25    78.89

The higher ender prospects generally work out better than those at the end - as should be expected - the interesting thing would be the high workout rate of the middle prospects, what could be some of the causes for the seemingly high breakout rate?

One theory would have to be the pure randomness of the distribution of numbers - since career numbers were used players ranked in the middle range early in their careers like Miguel Cabrera, Scott Kazmir, Roy Oswalt, Billy Wagner, and Carlos Beltran. There also seems to be a safer rate of players making it to the majors in the 11-15 grouping, perhaps dealing with the likelihood of players who won't necessarily be stars, but solid players being ranked in that range rather than the top 10 where players are more apt to have high potential rather than the immediate results.

The statistics seem to indicate that it doesn't appear that prospecting is getting better, but it doesn't seem to do such a bad job for that matter, it's simply a game of guesswork using the best tools available, the variations shown are enough to put some weight in the rankings, but not enough to consider them an invaluable resource for exactly how good a player will be.

References:

BA Lists
B-Ref