Prospects from 1996, Revisited – Part 5

by Geoff Young on Jul 07, 2005

5. Win Shares by Letter Grade

[This is a continuing series that will run each Thursday throughout the 2005 season. Start here for full details.]

One question we might like to answer is how many win shares were earned by players in each grade level during years 1996-2004. John Sickels assigned a grade to each prospect on a scale from A to C-. The charts that follow give the number of players for each grade, the total number of win shares garnered by players for each grade, the average number of win shares per player for each grade, and the number of players in various groupings of win shares for each grade. We’ll look first at all players, then at hitters only, and finally at pitchers only.

All Players
No Grd WS WS/player WS range
0 1-50 51-100 101-150 151-200 201+
12 A 1333 111.08 0 5 1 1 2 3
14 A- 1064 76.00 2 4 3 2 3 0
32 B+ 1485 46.41 6 12 9 5 0 0
59 B 1482 25.12 19 30 7 2 1 0
76 B- 1530 20.13 30 36 7 3 0 0
95 C+ 1353 14.24 49 36 7 3 0 0
129 C 1670 12.95 72 46 6 4 1 0
98 C- 932 9.51 64 28 5 1 0 0
515 tot 10849 21.07 242 197 45 21 7 3
Key: No, number of players assigned a particular letter grade; Grd, letter grade assigned by John Sickels in his 1996 Minor League Scouting Notebook; WS, total number of win shares for all players within a certain grade specification; WS/player, average number of win shares earned through 2004 per player; WS range, total number of players within a specified range of win shares.

And here is a graphical representation of the number of players per WS range by grade. The number of players is given as a percentage (e.g., 65% of players assigned a grade of C- have not earned a single win share).

Grade vs Win Shares (All Players)

No real surprises here. Pretty much a linear progression; the most highly regarded prospects earned the most win shares, and so on down the line. Also, the higher the grade, the better the chances of making it big (101+ win shares, an admittedly arbitrary mark). Of the 26 players graded at A or A-, 11 (42.3%) have amassed 101 or more win shares. That’s the exact same number that have accumulated 101 or more win shares among 167 players graded at B+, B, or B- (just 6.6%). And the 322 players graded at C+, C, or C-? Only 9 (2.8%) of them have 101+ win shares.

Hitters
No Grd WS WS/player WS range
0 1-50 51-100 101-150 151-200 201+
8 A 1064 133.00 0 3 0 0 2 3
11 A- 961 87.36 2 2 2 2 3 0
20 B+ 996 49.80 2 9 5 4 0 0
38 B 1160 30.53 11 18 6 2 1 0
38 B- 1157 30.45 11 18 6 3 0 0
46 C+ 893 19.41 21 19 3 3 0 0
70 C 1061 15.16 40 22 4 3 1 0
55 C- 682 12.40 36 14 4 1 0 0
286 tot 7974 27.88 123 105 30 18 7 3
Key: No, number of players assigned a particular letter grade; Grd, letter grade assigned by John Sickels in his 1996 Minor League Scouting Notebook; WS, total number of win shares for all players within a certain grade specification; WS/player, average number of win shares earned through 2004 per player; WS range, total number of players within a specified range of win shares.

And the pretty picture:

Grade vs Win Shares (Hitters)

When we looked at this after the 2001 season, the Grade B- hitters had done slightly better than the Grade B hitters, with exactly the same number of players. Now there is basically no difference between Grade B and Grade B- hitters. Even the distributions are nearly identical.

Pitchers

Note that the ranges for pitchers are a little different. This is due to the fact that pitchers can’t accumulate nearly as many win shares as hitters.

No Grd WS WS/player WS range
0 1-25 26-50 51-75 76-100 101+
4 A 269 67.25 0 0 2 0 1 1
3 A- 103 34.33 0 2 0 1 0 0
12 B+ 489 40.75 4 2 1 1 3 1
21 B 322 15.33 8 6 6 1 0 0
38 B- 373 9.82 19 12 6 1 0 0
49 C+ 460 9.39 28 12 5 4 0 0
59 C 609 10.32 32 19 5 2 0 1
43 C- 250 5.81 28 12 2 1 0 0
229 tot 2875 12.55 119 65 27 11 4 3
Key: No, number of players assigned a particular letter grade; Grd, letter grade assigned by John Sickels in his 1996 Minor League Scouting Notebook; WS, total number of win shares for all players within a certain grade specification; WS/player, average number of win shares earned through 2004 per player; WS range, total number of players within a specified range of win shares.

And again, with the picture:

Grade vs Win Shares (Pitchers)

The B+ pitchers did much better than the A- pitchers. The A- pitchers (Alan Benes, Rocky Coppinger, Jeff Suppan) turned out to be pretty weak, whereas the B+ pitchers comprised five (Bartolo Colon, #1; Ugueth Urbina, #5; Chan Ho Park, #6; Danny Graves, #7; Dustin Hermanson, #10) of the top ten pitchers in terms of win shares.

There were small fluctuations among the B-, C+, and C groups but nothing too startling.

Summary

There’s actually a lot you can get from this stuff. Here are what I find to be some of the more interesting items:

  • For the most part, the more highly regarded prospects turned into the more successful players and vice versa.
  • Pitchers, as a group, were less predictable than hitters in terms of development.
  • The flameout rate (zero win shares) was higher for pitchers (52.0%) than for hitters (43.0%); perhaps surprisingly, the gap narrowed with lower grades (B- or below): 56.6% for pitchers, 51.7% for hitters.

As I said in the original study, this isn’t earth-shattering, but it does confirm some suspicions. Two and a half years later, I stand by that assessment.

Next week, we’ll look at yet another aspect of the wackiness that is prognosticating prospects. I would tell you what that aspect is, but frankly, I’ve forgotten. Memo to self: work on teasers…

Comments are closed.