rolen community projection results

we got 57 responses in the rolen community projection. results:

oy. looks like i should have made playing time a variable after all. PECOTA and ZIPS both expect rolen to miss significant time in 2006, and a number of you expressed doubt that rolen would amass the 650 plate appearances i postulated for this exercise. i'll grant that 650 plate appearances may be unrealistically high -- but the numbers offered by ZIPS and PECOTA strike me as unrealistically low. in 9 big league seasons, rolen has failed to reach 500 plate appearances only once -- last year. his career average over those 9 seasons, including 2005, is 578 plate appearances; exclude last season and he has averaged 622 pa / year for his career. (the foregoing calculations exclude the 1996 season, when rolen wasn't called up until august 1.) for the four seasons prior to last year (ie, 2001 thru 2004) rolen averaged 642 plate appearances.

there's no reason he can't return to that level if the shoulder is sound; he's still only 31 years old. and if he's healthy enough to post a .906 ops, as ZIPS projects -- 16 points higher than rolen's career ops -- then it makes no sense to me that he would be limited to 500 plate appearances. a .906 ops means the shoulder has completely healed, in which case scottie will be out there every day. PECOTA's projection is more internally consistent in that regard, as it postulates a diminished level of ability -- .833 ops, 57 points below his career avg -- along with the diminished playing time. if the shoulder is preventing rolen from being rolen, that would explain why he might miss about a third of the season.

because the projected playing times are so disparate among these three models, i'm going to bring them into synch by pro-rating each batting line to 550 plate appearances. apples to apples and all that. here we go:

that's more like it. the VEB projection no longer looms as pie-in-the-sky optimism; on the contrary, our numbers come in remarkably close to ZIPS's (just as spot-on as we were with edmonds) and only slightly higher than PECOTA's. even those of you who were dubious of the 650 pa standard calibrated your projections appropriately.

you guys are good.

let's play around with these numbers some more; it's february and there's nothing else to do. if rolen meets the performance projected here, how much of a boost will the cardinal offense receive? let's keep him at 550 plate appearances and give the other 150 or so plate appearances to deivi cruz. ZIPS puts deivi at .270 / .299 / .384; PECOTA's got him at .272 / .306 / .383. roll 150 units of that in with 550 units of rolen's projected line, and you get a cumulative .277 / .359 / 478 -- a 130-point boost in ops over last year's third basemen:

last year's third basemen -- nunez, rolen, mabry, seabol, and luna -- created 74 runs as a group, or 4.6 runs per 27 outs. our roughed-out tandem projection for rolen/cruz yields 105 runs created, or 6.4 runs per 27 outs. so we'd be talking about a 30-run improvement at the position, roughly 3 to 4 wins in the standings -- enough of a boost to offset (or at least minimize) the impact of any decreased production at the corner outfield positions.

may it come to pass.

i'm taking a poll to determine the subject of our next community projection -- see the right-hand sidebar.