A while back I posted the leaders in SFR (which includes only infielders) for 2007. Well, that was several revisions of the system ago and so here is the new ranking based on refinements that have been made.
I've also included the middle and corner infield numbers from Baseball Info Solutions Plus/Minus system as published in The Hardball Times Annual (which I'm having a wonderful time reading over the break I might add). As a result I've excluded pitchers and catchers in the numbers below.
I also wanted to point out that Phil Birnbaum has an excellent post on his blog that shows the similarities between SFR and Bill James' Win Shares method and how using a more granular data set helps us be a little more discerning. And although I've mentioned it a few times, Sean Smith developed his TotalZone metric back in April of 2007 and so I was basically thinking the same thoughts well after Sean had developed his system. I simply didn't find his article when I sat down to construct the system late one Friday night. As is pointed out in the comments to that post systems like these could be used as is when retrosheet data contains both the fielder who fielded the ball and the hit type. Where the hit type is missing a simpler system would have to be used which would therefore make it less accurate. Where it could be applied immediately, however, is to minor league data, a task on my radar for after the break.
Ex
Team Balls Runners Runners SFR Plus/Minus
TOR 2370 563 498 47.6 81
COL 2414 572 512 45.7 30
SDN 2265 535 480 40.8 -6
SFN 2280 526 474 35.0 47
OAK 2349 550 504 34.7 22
BOS 2179 491 451 28.0 17
KCA 2185 505 492 22.9 49
SLN 2413 547 524 17.5 41
BAL 2270 544 524 16.5 -2
ATL 2234 528 508 13.9 -1
CHN 2126 473 454 13.4 16
NYN 2172 488 471 13.0 26
PHI 2297 533 518 8.7 23
ARI 2173 512 503 6.7 1
MIN 2291 526 516 5.1 30
ANA 2166 504 503 1.7 35
TEX 2326 561 557 0.4 -23
NYA 2237 515 520 -7.0 -3
LAN 2255 532 546 -8.8 -28
WAS 2271 510 519 -9.3 -41
DET 2322 544 555 -10.3 29
HOU 2315 536 553 -11.1 -23
CLE 2424 572 589 -11.9 -4
SEA 2358 541 563 -14.4 -25
PIT 2372 551 583 -19.8 -22
CIN 2216 503 534 -24.5 -22
CHA 2336 536 567 -27.6 -53
TBA 2169 499 555 -36.4 -68
MIL 2177 500 561 -46.4 -47
FLO 2230 514 598 -61.6 -98
The following plot shows the relationship between SFR and Plus/Minus. The correlation coefficient is 0.83 and so the systems agree substantially with the big differences being the Padres, Orioles, Angels, and Twins.
125x125_10off+copy.jpg)
2 comments:
>"Where the hit type is missing a simpler system would have to be used ..."
Wouldn't the simpler system just be the same system, but assuming a league-average ratio of liners/grounders/fly balls?
That's what I was thinking or using a ground ball/fly ball ratio estimated from the ground balls and fly balls that we know were hit to the fielder's area of responsibility (with some possible adjustments for the fact that we don't have trajectories for hits).
Post a Comment