community projection results: wells

before we look at the projection results, some day-after-presidents-day readings and musings:

• i got a few thoughts off my chest about ronnie belliard.
• sjoshi gets a few thoughts off his chest about rick ankiel, the creature that would not die. speaking of whom: he's featured in today's p-d.
• at Beyond the Boxscore, ryan vb (founder of Diaspora and writer of the Rams blog Turf Show Times) examines larry walker's hall-of-fame brief.
• in new york, they're still frozen by wagonmaker's bender, still watching it go by . . . .
• blez Q+As A's manager bob geren at Athletics Nation
• Gaslamp Ball recently concluded its Q+A with paul depodesta: read part 1, part 2, and part 3.

on to the day's main business. projecting pitcher performance is about as foolish an errand as any on earth; it's also a huge part of what distinguishes winning organizations from losing ones. the cards have made the nlcs 5 times in the last 7 years largely because they made accurate forecasts on darryl kile, woody williams, and chris carpenter. i'm oversimplifying, of course --- pujols and edmonds have had a little bit to do with their success --- and the cards were wrong on some guys too. but you get what i'm driving it: when you're throwing darts at the board of player performance, it's best for the darts with the pitchers' names on them to land closest to the bull's-eye. . . . and with that tortured metaphor, let's look at some numbers. my thanks and appreciation to Jonathan23 for compiling all the data and sending it over to me.

we had 51 entries in the wells projection. here's how our line stacks up against various others; they're listed in ascending era order:

			IP	H	BB	SO	HR	W-L	ERA	WHIP
CHONE			114	115	53	74	12	n/a	4.26	1.474
VEB			171	173	68	121	19	12-10	4.45	1.411
CCH			165	162	72	115	18	12-10	4.47	1.418
PECOTA			82	83	36	53	9	4-5	4.47	1.451
Shandler			145	n/a	n/a	99	15	7-10	4.59	1.500
Bill James			138	144	63	99	16	7-9	4.70	1.500
ZIPS			121	121	67	91	15	6-9	4.98	1.554
Marcel			100	113	46	69	12	5-8	5.31	1.590

the "CCH" line refers to Cardinals Clubhouse, which is taking VEB's cue and running its own set of community projections. fascinating, isn't it, that the two separate communities of cardinal fans arrived at nearly identical projections for kip wells? exact same won-loss record, near-exact matches in era and whip; even the walk and k rates and the hr/9 are an almost perfect fit. that's creepy. and it's a trend: compare our edmonds projection to CCH's:

			AB	R	H	2B	3B	HR	RBI	BB	|	AVG	OBP	SLG
VEB			425	79	116	21	1	25	86	69	|	.272	.373	.503
CCH			427	73	117	25	1	24	82	64	|	.274	.369	.506

i went over and checked out the CCH threads, wondering if perhaps the same guys posting numbers here are also posting over there; that would explain the similar outcomes. but it doesn't appear to be the case; it looks as if we have two discrete groups of forecasters, arriving at the same conclusions independently. this can mean only one thing --- and when i figure out what that is, i'll let you know. . . .

anyway, back to wells: if we lump all the projections together (i counted the twin cardinal projections as one, to avoid skewing the average), we come up with the following composite projection for wells:

			IP	H	BB	SO	HR	W-L	ERA	WHIP
composite			124	128	58	87	14	7-9	4.66	1.491

that's not quite what we VEBbers have in mind for the guy. the innings are probably too low; before last season, wells threw 133 innings or more 5 consecutive years, with a yearly average of 170 ip. but let's take the composite projection on its face: would that suffice? let's view it alongside the line compiled by the two departed pitchers he's being asked to replace:

			IP	H	BB	SO	HR	W-L	ERA	WHIP
wells			124	128	58	87	14	7-9	4.66	1.491
ponson/weaver			151	177	55	78	23	9-8	4.88	1.536

the second line elides ponson's 1-inning stint as a relief pitcher; those stats cover 28 starts (13 ponson, 15 weaver). so if wells --- this year's reclamation project --- can hit his rather modest composite projection, he'll more or less replace last year's reclamation jobs; fewer innings, but better results. that's not a particularly high bar to clear, but the comparison places things into some context. if wells simply approximates the ponson/weaver line, and carp approximates his line from last season, then the rotation improves if the remaining 3 starters include

1. one guy who approximates suppan's production (ie, an era in the low 4s), and
2. two guys who post ERAs below 6.00

those are not difficult requirements to meet; nobody has to pitch way beyond his abilities for it to happen. so i continue to think the rotation is going to get significantly better; the question is, by how much? it was so bad last season that the starters could improve by 5 games and the cardinals could still finish with 80something wins and miss the playoffs. in other words, the staff will probably be better than last year's but worse than those of'04-'05. . . . . gee, i've got a firm grasp on the obvious this morning.

a final note: both cardinal communities mirrored PECOTA's projection for wells, with one rather massive caveat: we hometowners expect wells to throw twice as many innings as PECOTA does. here's PECOTA times 2, alongside our projections --- the CCH guys were spot-on this one:

			IP	H	BB	SO	HR	W-L	ERA	WHIP
VEB			171	173	68	121	19	12-10	4.45	1.411
CCH			165	162	72	115	18	12-10	4.47	1.418
PECOTAx2			164	166	72	106	18	8-10	4.47	1.451

so if we just clone the guy, we've got ourselves a decent #3 starter.