we had 46 responses to the call for edmonds projections --- that's a great turnout, thanks to everybody who took a shot at this. the mean averages of those 46 projections line up almost exactly with the ZIPS projection posted by dan szymbroski at baseball think factory:
| avg | obp | slg | hr | rbi | r | |
|---|---|---|---|---|---|---|
| VEB community | .282 | .388 | .555 | 30 | 95 | 88 |
| szymbroski's ZIPS | .269 | .387 | .542 | 30 | 95 | 90 |
| bill james | .275 | .391 | .544 | 34 | 98 | 98 |
| ron shandler | .271 | .389 | .543 | 28 | 91 | 89 |
| PECOTA | .285 | .407 | .582 | 34 | 96 | 88 |
pretty amazing, no? those differences in batting avg and slugging are extremely small; about 5 singles over the course of the season. but we could hardly have gotten much closer in obp, hr, rbi, and runs. the VEB community even matched ZIPS to the thousandth of a point in projecting edmonds' isolated power for 2006; both "systems" (if you will) put his iso at .273. (for those unfamiliar with the term, isolated power = slugging avg minus batting avg.)
my calculations are slightly at variance with the ones DCRedbird posted this morning (see the comments to yesterday's thread); one of us probably made a typo entering the data. but the variance is minuscule and meaningless.
nate silver's PECOTA projections haven't come out yet, but if anybody has bill james' 2006 book, please put james' forecast for edmonds in the comments and i'll add it to the table. ditto ron shandler. i have rotoauthority's 2006 projections and will add them ASAP.
will be back later this morning with some additional information re our projections. i think this was a successful experiment and am inclined to repeat it with other players. if there's a specific player you'd like to see next on the list, post in the comments.
Update [2006-1-27 11:20:48 by lboros]: the bill james projections have been added to the table above; thank you, farley. and bschulte passed along ron shandler's totals (thank you sir) --- our figures are in extremely close agreement with shandler's, too. and tim from the roto authority was kind enough to provide PECOTA numbers --- see the table.
i've had a chance to look over the projection results, and one thing jumped out at me immediately: while 14 people projected edmonds to drive in 100+ runs, only 3 projected him to score more than 100, and nearly half of the respondents projected edmonds to score fewer than 90. those seem like low run-scoring expectations for a player who is projected, collectively, to have a .390 obp with 30 homers. as a group we are projecting that the players batting behind edmonds are going to drive him in fewer than 60 times in about 180 opportunities ? ie, fewer than 60 runs on non-homers, in approx 180 times on base via a non-homer. that?s roughly a 30 percent pickup rate; doesn?t sound very good. does that mean we have no confidence in rolen encarnacion bigbie and molina? here?s a quick comparison to previous years:
| year | non-HR times on base |
times driven in | pickup rate |
|---|---|---|---|
| 2005 | 189 | 57 | 30 pct |
| 2005 | 214 | 60 | 28 pct |
| 2003 | 165 | 50 | 30 pct |
| 2002 | 214 | 68 | 32 pct |
on the contrary, our run totals for edmonds evince a high degree of confidence in the 5-8 hitters. we are projecting that scottie juan et al will drive edmonds in at least as frequently as sanders, renteria, tino, jd drew, and other down-order hitters of years past.
here?s a further breakdown of the projections:
AVG
high: .300
low .260
mean: .282
median: .280
projections > .290 = 12 (27 pct)
projections < .280 = 18 (40 pct)
most popular: .280 (5), .278 (4)
cluster: 15 projections (1/3 of entries) between .282 and .278
only one projection of .300 or above; only one projection below .270
OBP
high: .412
low: .337
mean: .388
median: .390
projections > .400 = 10 (22 pct)
projections < .380 = 8 (17 pct)
most popular: .400, .390 (4 each)
SLG
high: .623
low: .475
mean: .555
median: .555
projections > .600 = 5 (11 pct)
projections < .500 = 2 (4 pct)
most popular: .552 (4), .600, .560, .540 (3 each)
cluster: 18 projections (39 pct) between .540 and .560
HR
high: 39
low: 22
mean: 30
median: 31.5
projections > 30 = 30 (65 pct)
projections < 30 = 16 (35 pct)
most popular: 32, 33 (7 apiece)
cluster: 27 projections (59 pct) between 28 and 33
RBI
high: 112
low: 68
mean: 95
median: 95.5
projections > 100 = 14 (30 pct)
projections < 90 = 8 (17 pct)
most popular: 91 (4), 109, 103, 97, 96, 94 (3 each)
RUNS
high: 104
low: 67
mean: 88
median: 90
projections > 100 = 3 (7 pct)
projections < 90 = 20 (43 pct)
most popular: 90 (8)
cluster: 18 projections (39 pct) between 87 and 91