Offensive Projections
Name P Age AVG OBP SLG G AB R H 2B 3B HR RBI BB K SB CS OPS+
Alex Rodriguez 3b 34 .281 .382 .526 128 477 86 134 25 1 30 93 69 106 15 3 141
Mark Teixeira# 1b 30 .280 .369 .505 147 574 95 161 38 2 29 112 75 111 1 0 132
Hideki Matsui* dh 36 .277 .360 .464 111 390 54 108 20 1 17 64 49 58 1 1 119
Derek Jeter ss 36 .303 .372 .424 139 571 88 173 29 2 12 60 56 86 19 5 114
Johnny Damon* lf 36 .272 .350 .436 130 507 87 138 29 3 16 61 60 83 18 3 110
Nick Swisher# rf 29 .243 .354 .446 147 511 87 124 30 1 24 81 85 137 1 1 113
Robinson Cano* 2b 27 .296 .334 .474 160 614 82 182 42 5 19 88 31 71 3 4 110
Shelley Duncan rf 30 .252 .328 .460 116 417 59 105 22 1 21 72 45 100 3 1 109
Xavier Nady rf 31 .275 .326 .454 85 313 51 86 18 1 12 61 19 67 1 1 107
Jorge Posada# c 38 .256 .336 .430 77 270 33 69 15 1 10 46 32 68 1 0 104
Eric Hinske* rf 32 .233 .324 .424 105 257 45 60 14 1 11 40 31 68 3 1 99
Jesus Montero c 20 .273 .315 .416 115 454 47 124 24 1 13 61 26 78 0 0 94
Melky Cabrera# cf 25 .266 .324 .393 155 519 63 138 27 3 11 64 43 67 11 3 92
Juan Miranda* 1b 27 .246 .317 .391 120 447 52 110 24 1 13 62 44 111 1 1 89
John Rodriguez* lf 32 .244 .321 .391 84 266 39 65 13 1 8 38 25 70 0 1 90
Brett Gardner* cf 26 .253 .328 .344 119 387 66 98 14 6 3 31 42 75 32 7 81
Cody Ransom 3b 34 .233 .304 .401 106 347 54 81 20 1 12 51 33 90 6 2 87
Jerry Hairston Jr. 3b 34 .252 .305 .378 101 294 53 74 17 1 6 37 21 45 8 3 82
Reegie Corona# 2b 23 .250 .313 .347 136 507 72 127 27 2 6 42 46 89 15 4 78
Chris Stewart c 28 .260 .323 .344 84 262 33 68 14 1 2 24 20 36 1 1 80
Kevin Russo 2b 25 .260 .312 .351 94 365 46 95 17 2 4 31 26 65 10 5 78
Francisco Cervelli c 24 .255 .307 .380 61 184 24 47 9 1 4 22 12 37 0 2 83
Freddy Guzman# cf 29 .251 .303 .330 119 446 75 112 16 5 3 33 33 64 44 11 70
Colin Curtis* lf 25 .244 .297 .347 132 513 61 125 23 3 8 50 36 101 5 3 73
Austin Jackson cf 23 .245 .296 .338 138 551 65 135 25 4 6 56 38 136 16 4 70
Austin Romine c 21 .247 .285 .367 118 458 53 113 24 2 9 52 23 92 5 3 74
P.J. Pilittere c 28 .262 .299 .336 83 301 33 79 14 1 2 31 14 34 0 1 70
Doug Bernier# ss 30 .224 .315 .313 96 281 40 63 12 2 3 30 33 75 2 1 70
Eric Duncan* 3b 25 .233 .280 .353 113 408 44 95 20 1 9 45 26 93 2 2 69
Ramiro Pena# ss 24 .249 .296 .332 93 382 47 95 17 3 3 32 25 78 5 5 69
Brian Peterson c 30 .237 .293 .320 52 169 20 40 8 0 2 20 13 37 0 1 65
Jose Molina c 35 .227 .275 .308 68 185 20 42 9 0 2 17 11 39 1 0 57
Kevin Cash c 32 .208 .277 .309 53 149 14 31 6 0 3 18 14 47 0 0 57
Total .258 .324 .397 0 127 178 329 66 60 33 162 115 251 23 80 2936
Defensive Projections
Name CThr 1b 2b 3b ss lf cf rf
Rodriguez Av
Teixeira# Av
Matsui* Pr
Jeter Av
Damon* Vg Pr
Swisher# Av Av Fr Av
Cano* Fr
Duncan Av Fr Fr
Nady Av Av
Posada# Pr Fr
Hinske* Av Pr Av Av
Montero Pr
Cabrera# Vg Av Vg
Miranda* Av
Rodriguez* Fr Fr
Gardner* Vg Vg
Ransom Av Av Av Fr
Hairston Fr Av Fr Av Av Av
Corona# Vg Av
Stewart Vg
Russo Av Av
Cervelli Av
Guzman# Vg Vg
Curtis* Av Av Av
Jackson Av Av
Romine Av
Piliterre Fr Fr
Bernier# Av Av Av Av Av Av
Duncan* Fr Fr Fr Fr
Pena# Vg Vg Vg
Peterson Fr
Molina Av
Cash Av
* - Bats Left
# - Switch Hitter
ODDIBE (Odds of Important Baseball Events)
Name PO EX VG AV FR PO COMP 1 COMP 2 COMP 3
RodriguezAlex 3B 84% 12% 3% 1% 0% BoyerKen SchmidtMike MoraMelvin
TeixeiraMark 1B 35% 44% 15% 6% 1% MurrayEddie DavisChili OrtizDavid
JeterDerek SS 69% 21% 7% 2% 1% RizzutoPhil AparicioLuis ApplingLuke
MatsuiHideki DH 13% 32% 27% 23% 4% FairlyRon BainesHarold MartinezTino
CanoRobinson 2B 45% 19% 16% 13% 7% OrtaJorge VidroJose WalkerTodd
DamonJohnny LF 20% 29% 21% 18% 12% GoslinGoose SlaughterEnos VeachBobby
SwisherNick RF 14% 26% 24% 23% 14% EvansDwight BurroughsJeff DeerRob
DuncanShelley RF 10% 23% 22% 25% 20% BrunanskyTom BuhnerJay DyeJermaine
NadyXavier RF 5% 22% 25% 28% 20% GonzalezJuan YoungbloodJoel GuillenJose
PosadaJorge C 25% 40% 21% 10% 3% SchangWally PiazzaMike HartnettGabby
HinskeEric RF 4% 9% 16% 29% 43% NunnallyJon LowensteinJohn BrunanskyTom
MonteroJesus C 4% 31% 37% 24% 3% RodriguezIvan CardonaJavier AlomarSandy
CabreraMelky CF 5% 10% 28% 38% 20% AlmadaMel GallagherDave CrispCoco
GardnerBrett CF 3% 13% 34% 37% 13% PrietoChris RobertsDave ButlerBrett
MirandaJuan 1B 0% 1% 4% 27% 68% FalconeDave McAnultyPaul StahoviakScott
RodriguezJohn LF 1% 3% 5% 15% 76% StahlLarry KempSteve HollandsworthTodd
RansomCody 3B 1% 5% 14% 28% 51% BooneRay DurhamRay ClaytonRoyce
HairstonJerry 3B 0% 3% 10% 26% 60% BarnesSkeeter O’RourkeFrank BrookensTom
GuzmanFreddy CF 1% 3% 16% 38% 41% BrownAdrian RedmanTike RobertsDave
CoronaReegie 2B 1% 3% 8% 25% 64% PhillipsTony ThomasDerrel CrespoFelipe
StewartChris C 0% 3% 14% 44% 38% TaylorZack WilsonCraig KluttzClyde
CervelliFrancisco C 0% 0% 4% 29% 67% TorrealbaYorvit DempseyPat BellorinEdwin
RussoKevin 2B 0% 1% 4% 16% 78% BloomquistWillie BarfieldJosh LansingMike
JacksonAustin CF 0% 0% 3% 18% 79% TerreroLuis RepkoJason MoranJavon
RomineAustin C 0% 1% 5% 29% 66% HearnEd CotaHumberto MathisJeff
CurtisColin LF 0% 0% 0% 1% 99% SwannPedro VazRoberto MillerDavid
PilittereP.J. C 0% 0% 3% 24% 73% EncarnacionAngelo TaylorZack PhillipsPaul
BernierDouglas SS 0% 0% 5% 20% 75% SchofieldDick RungePaul KoppeJoe
DuncanEric 3B 0% 0% 0% 2% 97% HansonTravis FrostadEmerson BakerDave
PenaRamiro SS 0% 0% 2% 12% 86% WeissWalt MeloJuan OlmedoRay
PetersonBrian C 0% 0% 1% 6% 92% TillmanBob HundleyRandy MahoneyMike
MolinaJose C 0% 0% 1% 4% 95% DifeliceMike KnorrRandy SantiagoBenito
CashKevin C 0% 0% 1% 4% 95% KnorrRandy ChavezRaul DifeliceMike
Name .300 BA .375 OBP .500 SLG 140 OPS+ 45 2B 10 3B 30 HR 30 SB
RodriguezAlex 23% 61% 66% 47% 0% 0% 49% 0%
TeixeiraMark 22% 41% 48% 27% 17% 0% 41% 0%
JeterDerek 56% 44% 4% 8% 3% 0% 1% 6%
MatsuiHideki 21% 27% 18% 11% 0% 0% 2% 0%
CanoRobinson 49% 8% 29% 13% 40% 1% 11% 0%
DamonJohnny 18% 20% 11% 8% 3% 1% 6% 4%
SwisherNick 2% 19% 14% 9% 4% 0% 18% 0%
DuncanShelley 4% 5% 19% 5% 0% 0% 9% 0%
NadyXavier 20% 5% 14% 2% 0% 0% 0% 0%
PosadaJorge 9% 11% 11% 4% 0% 0% 0% 0%
HinskeEric 2% 5% 6% 2% 0% 0% 0% 0%
MonteroJesus 12% 0% 1% 0% 0% 0% 0% 0%
CabreraMelky 9% 2% 0% 0% 1% 1% 0% 0%
GardnerBrett 4% 3% 0% 0% 0% 10% 0% 70%
MirandaJuan 1% 0% 0% 0% 0% 0% 0% 0%
RodriguezJohn 3% 4% 2% 1% 0% 0% 0% 0%
RansomCody 1% 0% 3% 0% 0% 0% 0% 0%
HairstonJerry 6% 1% 0% 0% 0% 0% 0% 0%
GuzmanFreddy 3% 0% 0% 0% 0% 7% 0% 98%
CoronaReegie 2% 1% 0% 0% 0% 0% 0% 0%
StewartChris 9% 4% 0% 0% 0% 0% 0% 0%
CervelliFrancisco 9% 2% 0% 0% 0% 0% 0% 0%
RussoKevin 8% 1% 0% 0% 0% 0% 0% 0%
JacksonAustin 1% 0% 0% 0% 0% 2% 0% 0%
RomineAustin 1% 0% 0% 0% 0% 0% 0% 0%
CurtisColin 0% 0% 0% 0% 0% 0% 0% 0%
PilittereP.J. 10% 0% 0% 0% 0% 0% 0% 0%
BernierDouglas 0% 2% 0% 0% 0% 0% 0% 0%
DuncanEric 0% 0% 0% 0% 0% 0% 0% 0%
PenaRamiro 2% 0% 0% 0% 0% 0% 0% 0%
PetersonBrian 3% 1% 0% 0% 0% 0% 0% 0%
MolinaJose 2% 0% 0% 0% 0% 0% 0% 0%
CashKevin 1% 1% 0% 0% 0% 0% 0% 0%
Extrapolated Career Statistics
Name BA OBP SLG G AB R H 2B 3B HR RBI BB SO HP SB CS OPS+
RodriguezAlex .294 .384 .550 2960 11273 2184 3311 585 32 745 2207 1489 2426 222 379 88 141
TeixeiraMark .276 .365 .502 2293 8837 1417 2439 565 31 456 1650 1147 1800 134 26 3 126
PosadaJorge .273 .370 .471 1894 6425 955 1752 400 13 283 1142 955 1547 73 19 18 120
MatsuiHideki .282 .361 .469 1412 5099 768 1439 283 17 212 863 623 751 34 17 13 119
JeterDerek .310 .380 .444 2832 11532 1993 3578 578 68 276 1322 1155 1917 183 389 106 117
SwisherNick .238 .348 .438 1841 6374 1052 1514 360 15 295 950 1050 1687 74 17 17 109
CanoRobinson .291 .327 .457 2271 8888 1221 2582 589 45 267 1115 443 1034 63 54 67 106
DamonJohnny .283 .352 .434 2702 10644 1851 3009 573 109 272 1231 1119 1419 51 447 111 104
Pitching Statistics - Starters
Name Age ERA W L G GS INN H ER HR BB K ERA+
C.C. Sabathia* 29 3.48 17 9 34 34 235.1 215 91 23 58 201 130
Joba Chamberlain 24 4.21 8 5 38 26 147.1 143 69 17 63 136 107
A.J. Burnett 33 4.46 12 11 31 31 195.2 186 97 25 86 178 102
Andy Pettitte* 38 4.48 11 11 31 31 190.2 200 95 21 67 130 101
Chad Gaudin 27 5.02 10 9 40 24 154.1 167 86 18 69 126 90
Sergio Mitre 29 5.07 6 5 23 19 108.1 129 61 14 31 64 89
Zachary McAllister 22 5.08 8 10 24 24 124.0 139 70 17 45 66 89
Chien-Ming Wang 30 5.13 6 8 19 17 105.1 114 60 10 40 58 88
Ian Kennedy 25 5.22 4 4 17 16 81.0 83 47 9 40 62 87
Ivan Nova 23 5.87 6 11 25 24 130.1 156 85 16 76 64 77
Josh Towers 33 6.11 5 8 25 16 101.2 130 69 19 29 50 74
Kei Igawa* 30 6.39 5 12 25 24 132.1 158 94 30 54 75 71
Pitching Statistics - Relievers
Name Age ERA W L G GS INN H ER HR BB K ERA+
Mariano Rivera 40 3.18 5 2 62 0 62.1 53 22 7 13 64 141
David Robertson 25 3.84 6 3 54 0 70.1 59 30 5 38 81 120
Philip Hughes 24 4.32 7 5 40 12 93.2 93 45 11 33 84 105
Edwar Ramirez 29 4.62 3 3 52 0 64.1 60 33 8 38 68 97
Alfredo Aceves 27 4.70 5 4 62 0 76.2 81 40 10 23 55 95
Brian Bruney 28 4.72 3 2 52 2 47.2 47 25 5 29 40 94
Phil Coke* 27 4.91 4 3 79 0 69.2 72 38 9 29 57 93
Mark Melancon 25 4.96 4 3 47 0 74.1 78 41 9 29 55 92
Damaso Marte* 35 4.97 3 2 53 0 41.2 41 23 5 17 40 92
Jonathan Albaladejo 27 5.06 7 6 74 1 90.2 100 51 13 36 62 89
Humberto Sanchez 27 5.32 2 2 19 3 23.2 25 14 3 13 15 85
Michael Dunn* 25 5.50 4 4 44 0 75.1 76 46 10 57 69 83
Eric Wordekemper 26 5.61 2 3 40 0 51.1 60 32 8 26 28 80
Romulo Sanchez 26 5.69 3 6 39 6 68.0 72 43 10 39 44 80
Kevin Whelan 26 5.91 2 3 39 2 56.1 53 37 7 57 49 77
Zachary Kroenke* 26 6.14 2 5 42 1 58.2 64 40 10 41 37 74
Jose Valdez 27 6.16 3 5 42 2 57.0 62 39 10 40 40 74
Grant Duff 27 7.07 3 9 40 10 85.1 103 67 16 68 44 64
* - Throws Left
ODDIBE (Odds of Important Baseball Events)
Player PO TOP MID BOT COMP 1 COMP 2 COMP 3
RiveraMariano RP 75% 23% 2% JonesDoug HoffmanTrevor AndersenLarry
SabathiaC.C. SP 89% 10% 0% KoufaxSandy GuidryRon CandelariaJohn
RobertsonDavid RP 46% 36% 18% ManteiMatt WilliamsonScott LittellMark
ChamberlainJoba SP 46% 41% 12% MaloneyJim RijoJose PizarroJuan
HughesPhilip RP 24% 46% 30% BantaJack CaudillBill AyalaBobby
BurnettA.J. SP 30% 58% 12% NomoHideo ClemensRoger StottlemyreTodd
PettitteAndy SP 23% 64% 13% RogersKenny KeyJimmy ReussJerry
RamirezEdwar RP 22% 36% 42% NelsonJoe SimpsonAllan DohmannScott
AcevesAlfredo RP 19% 30% 50% FoulkeKeith DeLeonLuis BellHeath
BruneyBrian RP 24% 28% 49% BalfourGrant NelsonJoe LaCorteFrank
CokePhil RP 15% 31% 54% LittlefieldDick AyalaBobby GuardadoEddie
MelanconMark RP 17% 30% 53% AyalaBobby YorkJim BeardDave
MarteDamaso RP 20% 32% 48% PercivalTroy DotelOctavio GordonTom
GaudinChad SP 19% 40% 41% DarwinDanny WeaverJim GreggHal
AlbaladejoJonathan RP 12% 28% 59% BelindaStan LaddPete JohnsonKen
MitreSergio SP 20% 35% 46% PavanoCarl EthertonSeth EatonAdam
McAllisterZachary SP 7% 54% 39% LittellMark BurdickStacey JohnsonJoe
WangChien-Ming SP 9% 47% 44% WojnaEd KisonBruce HamiltonJoey
KennedyIan SP 20% 30% 50% HardenRich D’AcquistoJohn FoppertJesse
SanchezHumberto RP 16% 29% 54% JonesGordon CoombsDanny DeanPaul
DunnMichael RP 10% 21% 69% WilliamsMitch MeyerJack BruneyBrian
SanchezRomulo RP 1% 15% 84% MarquezJeff ValdezCarlos PattersonJeff
WordekemperEric RP 1% 19% 80% CrowDean RoehlScott HinesCarlos
NovaIvan SP 0% 15% 85% WellsJared RitzKevin MoehlerBrian
WhelanKevin RP 1% 12% 87% BowlesBrian VoylesBrad van BurenJermaine
TowersJosh SP 5% 18% 77% SchmidtDave SmithsonMike TollbergBrian
KroenkeZachary RP 0% 7% 93% FeshSean NanceShane DuquetteBryan
ValdezJose RP 0% 6% 94% WilmetPaul MiadichBart VillanoMike
IgawaKei SP 0% 6% 94% OlinSteve YoungCurt LorraineAndrew
DuffGrant RP 0% 0% 100% SextonJeff NelsonJeff HolmanBrad
Player 130 ERA+ 100 ERA+ K/9 >8 BB/9 <2 HR/9 <1
RiveraMariano 68% 96% 85% 62% 62%
SabathiaC.C. 55% 97% 34% 32% 73%
RobertsonDavid 41% 75% 85% 1% 85%
ChamberlainJoba 17% 69% 57% 1% 52%
HughesPhilip 18% 59% 51% 8% 61%
BurnettA.J. 6% 59% 57% 0% 38%
PettitteAndy 3% 54% 2% 1% 52%
RamirezEdwar 19% 50% 71% 1% 51%
AcevesAlfredo 17% 43% 17% 27% 51%
BruneyBrian 19% 46% 40% 1% 61%
CokePhil 13% 42% 34% 5% 53%
MelanconMark 15% 40% 21% 8% 47%
MarteDamaso 20% 46% 59% 8% 59%
GaudinChad 4% 35% 32% 0% 56%
AlbaladejoJonathan 10% 36% 13% 9% 41%
MitreSergio 6% 33% 1% 27% 47%
McAllisterZachary 1% 23% 0% 1% 31%
WangChien-Ming 1% 24% 0% 1% 76%
KennedyIan 8% 32% 20% 1% 62%
SanchezHumberto 16% 46% 13% 11% 52%
DunnMichael 9% 28% 53% 0% 44%
SanchezRomulo 1% 10% 0% 0% 28%
WordekemperEric 1% 15% 0% 0% 30%
NovaIvan 0% 3% 0% 0% 46%
WhelanKevin 0% 10% 35% 0% 52%
TowersJosh 1% 11% 1% 24% 22%
KroenkeZachary 0% 5% 0% 0% 18%
ValdezJose 0% 3% 2% 0% 15%
IgawaKei 0% 1% 1% 0% 1%
DuffGrant 0% 0% 0% 0% 7%
Extrapolated Career Statistics
Player W L S ERA G GS IP H ER HR BB SO ERA+
BurnettA.J. 158 142 0 4.22 408 404 2580.0 2411 1209 277 1089 2372 105
PettitteAndy 268 172 0 4.07 573 563 3557.7 3762 1609 319 1146 2592 115
RiveraMariano 82 56 608 2.37 1038 10 1206.0 955 317 72 282 1132 190
SabathiaC.C. 290 181 0 3.80 618 618 4113.3 3985 1735 410 1167 3380 120
All figures in % based on projection playing time
Disclaimer: ZiPS projections are computer-based projections of performance.
Performances have not been allocated to predicted playing time in the majors -
many of the players listed above are unlikely to play in the majors at all in 2009.
ZiPS is projecting equivalent production - a .240 ZiPS projection may end up
being .280 in AAA or .300 in AA, for example. Whether or not a player will play
is one of many non-statistical factors one has to take into account when predicting
the future.
Players are listed with their most recent teams unless Dan has made a mistake.
This is very possible as a lot of minor-league signings are generally unreported in
the offseason.
ZiPS is projecting based on the AL having a 4.46 ERA and the NL having a 4.41 ERA.
Players that are expected to be out due to injury are still projected. More information
is always better than less information and a computer isn’t what should be projecting
the injury status of, for example, a pitcher with Tommy John surgery.
Positional offense is ranked by RC/27 and divided into quintiles based on what the
most frequent starting players at each position did in 2007-2009. Excellent is the top
quintile, Very Good the 2nd quintile and so on.
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
1. . . . . . . Posted: October 04, 2009 at 09:59 PM (#3340141)Doug Jones, Age 40 - 80.1 IP, 2.02 ERA, 231 ERA+.
That line looks pretty much like a standard Rivera season, and there aren't many 40 year old closers to compare against.
What's the avg OPS+ for catchers? Romine projecting to 74 already seems very encouraging.
Can't wait for the Red Sox 2010 DIPS that shows Buchholz with a 1.03 ERA.
What would you expect when 8 of 9 position player starters had a 2009 OPS+ above their career averages?
Can't wait for the Red Sox 2010 DIPS that shows Buchholz with a 1.03 ERA.
An Oriole fan writing a projection system with a pro-Red Sox bias would be a rather odd choice, I imagine.
Dan, I have what I'm sure is going to be shocking news for you. There are a certain number of Yankee fans who have used the last eight years of no title to fuel one of the most intensely wacky persecution complexes in the history of psychiatry.
I think that Jeter projection is pretty good, 114 OPS+ and average defense at SS. That's a very valuable season. Of course I'd love him to hit .330 again but I would take that projection.
I'm surprised about Melancon's awful projection. I know he's struggled in Major League time but his minor league numbers are really good. Are his control issues in the Majors bad enough to make his projection that low or are his minor league numbers not as impressive as I think they are? To some extent my question also applies to Robertson.
I like the rotation depth the Yanks are going to have. Gaudin, Mitre, Wang, Kennedy and MCAllister are all reasonable enough 5th starters and some of them have some pretty high upside for 6th-10th starters.
I hope this year is the beginning of a more consistent stretch for Cano and that in 10 years we look back and wonder what the hell happened to him in 08.
Can't wait for the Red Sox 2010 DIPS that shows Buchholz with a 1.03 ERA.
This is annoying because it's exactly what the posters here expect from Yankee fans and it's just going to lead to more anti-Yankee posts and distract from actually talking about the team.
Buhner, though, was a middle of the order threat on a playoff team, and remained such for a couple years. Why does he come up as a comp for a minor league slugger?
Give that man the 10,000 dollars!
tex: 907 + 948
a-rod: 941 - 933
jete: 791 + 871
posada: 836 + 885
cano: 808 + 871
damon: 800 + 854
matsui: 844 + 874
cabrera: 703 + 752
swisher: 835 + 869
absent a-rod, every starting player is 30-80 points better than their projections. is this all park-related?
Well, if that's all it is, I take back every nasty thing I've ever said about the park, including its financing. But though their .858 / .821 home / road OPS+ split shows that their new Stadium didn't hurt, what does a similar .859 / .753 split for the Red Sox say about the way that Boston hitters use Fenway as a crutch?
First, the baseline for Duncan is a little higher - ZiPS/zMLE have Duncan as a 114 OPS+ in a neutral context.
There's also the issues that I calculated exact component park factors (rather than generalized ones) and with a slightly easier league, Buhner's 126 OPS+ over those years in question drops to 120 OPS+ in a neutral context.
Even that 6 is actually one of the bigger differences among the comps - remember, shape matters, too. If I have a 115 OPS+ hulking slugger, I want a group that contains 110 and 120 OPS+ hulking sluggers, not Carl Crawford!
Some more on Duncan's list:
Jason Lane
Kurt Airoso
Bubba Trammell
Richie Zisk
Carmelo Martinez
Aaron Guiel
Jeff Burroughs
Wally Post
Ed Spiezio
Jerry Morales
Dusty Baker
other Brian Giles
Rob Deer
Andy Kosco
Joe Vitiello
Sam Mele
Willie Kirkland
Dick Gernert
Gene Schall
Chuck Workman
Vic Wertz
Greg Vaughn
Johnny Rizzo
Mike Lum
Cecil Fielder
Ron Kittle
Jack Voigt
Ozzie Timmons
Deron Johnson
Ryan Radmanovich
Ruben Sierra
Richard Hidalgo
Jerry Martin
Dan Ford
Dave Kingman
Ernie Young
Dustan Mohr
Steve Balboni
Babe Dahlgren
And on and on.
absent a-rod, every starting player is 30-80 points better than their projections. is this all park-related?
Part park, part league scoring more than I thought, part everybody but ARod playing better than usual. As noted above, considering that 8 of the 9 regulars played above their career averages, I should have hit low.
Assuming Biggio gets in, if he makes it to 3000, Damon will probably be the first 3K player not to make the Hall, with the exception of those not in for other reasons (Rose of course and very likely Palmeiro).
If Damon gets 3000 hits, I think he gets into the hof, he'll be compared to Brock enough and his superior defense (even with the weak arm) will be enough to get him in.
Maybe eventually, but it's a tough crowd now. When Tim Raines gets 1/3 of what he needs for qualification, there ain't going to be many who are going to "feel" like Johnny Damon is a HOFer.
20 years ago, sure. It's pretty scary that it only took some knee problems to keep Buckner from getting 3000 hits. Now that would be comical.
I think he has it harder than Brock did, but he is going to get compared to Brock, no matter whether that is the best comparison, and in almost every aspect in that comparison Damon is going to grade out better.
cairo: 855
chone: 847
hbt: 860
marcel: 856
pecota: 822
zips: 883
So don't pick on Szym too much. If anything, pick on PECOTA.
Will he? It's getting late out here, so I won't take a deep look at the numbers, but Brock retired with the single-season and career records for stolen bases. For "Fame", that counts for a lot.
I'd say Brock is perhaps the one guy who is non-Hall of Meriter, Hall of Famer, with both groups getting it right.
I don't see Damon having much of a chance of making it (absent more seasons like this one), 3,000 hits or not. Of course, I'm perhaps the most disbelieving of automatic markers here.
The Book Blog thread explaining Regression
Important reading.
Please, please, please don't let Michael Kay find out the Scooter is Jeter's #1 comp.
The Robinson Cano comps have me thinking Jorge Cantu must be #4. :-)
2001 OPS+: 109
10/14/01: Breaks leg
2002 OPS+: 110
2003 OPS+: 38
Yes, he was - better than his contemporary Ray Schalk, who is in the HOF on the strength of being one of the "Clean Sox" of 1919 (and Warren Brown's pushing).
-- MWE
Important reading.
From that link:
The first highlighted passage is mgl's. The second highlighted passage is one that needs more elaboration. In this case, I'm specifically wondering how those numbers for Phil Hughes were arrived at, numbers which seem to imply a mean that's a lot worse than what his performance this year (his first full injury free year) might project. What assumptions were you making about injuries and managerial schizophrenia when you project him to a 4.32 ERA, with 12 starts in 40 games?
I'm obviously missing something here, but what is it? How do you take the sum of Phil Hughes's minor league and Major League career and come up with a line like that? Seems to me that it has to be based on an expectation of another major injury somewhere along the road.
Ugh. This again. MGL should never write about statistics. Let's break it down:
1. any "player will always regress" WRONG. The notion of "always" is antithetical to statistics.
2. "on the average" CONTRADICTS ANY MEANING OF THE WORD "ALWAYS"
3. "towards average" WRONG
....... really, that first sentence is just horrible.
4. "the chances of him being worse is <sic> greater than the chances of him being better" CONTRADICTS #1 AND #3 which clearly state he has no chance of being better (unless he was below average)
5. "ALWAYS" -- wrong (antithetical) but hard to know what this is referring to given the preceding series of conflicting statements
6. "as long as we properly define the mean for the player" -- AYE, THERE'S THE RUB
We don't know what Chipper Jones' "true" mean is. But if we did know Chipper's "true" mean, he would be every bit as likely to over-perform as under-perform that mean (give or take a bit of skew in the true distribution -- i.e. it's not necessarily normal but it's probably close enough).
As to the rest ... a player whose true mean is a 110 OPS+ but who has hit for a 100 OPS+ the last 3 years will, OF COURSE!! ALWAYS!! CHANCES ARE!!, regress towards his true 110 OPS+ -- i.e. he will regress away from average. The projection systems will, on average, under-predict such a player because the projection systems won't know he's a true 110 OPS+ player. That's not a big deal but it's a fact.
Regression towards the mean is really the wrong concept because we don't know what the mean is for any individual player. In a sense, we don't even predict a mean for the individual player. A better explanation of how projection systems work would be something like the following (admittedly short on specifics):
From 2007-2009, 8 players with at least 1200 PA had an OPS+ of 140 or better. However, from 1901-2009, only 57 hitters have had a 140 OPS+ for their careers (minimum PA of 4000). Even if you limit it to ages 25-32 (at least 3000 PA), you only get 90 hitters. Therefore we assume/estimate that it would be incredibly unlikely that the unknown distribution of true talent levels would produce 8 true 140 OPS+ hitters at one time. Therefore some members of this group must be over-performing their (unknown) true talent so we estimate that, on average, members of this group will regress towards an OPS+ of 100. That is, as a group they averaged something like a 150 OPS+ but we'll project the group to average about a 135-140 OPS+ in 2010. We really haven't a clue what any individual member of that group will do (even Teixeira has a 1% chance of being "poor" next year) and we know that it is highly likely that, if only by chance, one or more of them will hit better in 2010 than they did from 2007-2009.
Of course the projections adjust for other predictive factors (most importantly age) and they don't use OPS+, but that's the gist of it.
Or if you prefer ... regression towards the mean would be the correct concept if every player in baseball had the same mean -- i.e. came from the same population. We are very highly confident however that Albert Pujols is not in the same population as Aaron Miles. So the task becomes figuring out which population a given player most likely belongs to. The main drivers behind our guess are the player's recent performance and age. Unfortunately, even using 3-4 seasons, these are small samples to detect small differences in performance that are so important in baseball. So we have a great deal of uncertainty around a mean estimated purely by recent performance and age.
The question which follows is -- do we think that performance-age mean estimate is an unbiased predictor of the player's true mean? That's not an easy question to answer but history and earlier statistical analysis do quite clearly show that such a mean is biased. It is biased upwards for those who performed above average (in the overall league sense) and biased downward for those below average.
But let's be clear about what's "really" going on. There are a group of players who come from the "true 120 OPS+" population. Over the last few years, a few of these guys (say 20%) had an OPS+ between 90 and 110 -- they will all be projected to about a 100 OPS+. 30% were between 110 and 120 and they'll be projected to about a 110 OPS+ (everybody loses 5 points towards average). 30% were between 120 and 130 and they'll be projected to about a 120 OPS+. And 20% were between 130 and 150 and they'll be projected to about a 130-135 OPS+. You'll notice the projection system has correctly projected only 30% of that population to their true mean.
In the subsequent year, those "true 120" guys who were projected to a 100 will be considered "true 100" flukes after putting up a 120 season and will get projected to about a 105 OPS+ even though, were we omnisicient, we'd have no reason to expect them not to repeat their performance.
This is one thing about projection systems -- they can't identify "true 120" players who have underperformed the last 3-4 years. They see them as guys who have put up a 100ish OPS+ and will remain at a 100ish OPS+. They handle average or better players who overperformed their true means fine (by regressing them towards 100); they handle average or worse players who underperformed their true means fine (by regressing them towards 100). But above-average players who underperformed their true means and below-average players who overperformed are not being projected correctly because they can't be identified.
That's not a fault of the projection systems necessarily. But the key is that they treat the mean estimate (after regression) as "true" rather than incorporating the substantial amount of uncertainty that there is in that projection. Unfortunately, the uncertainty introduced into a projection just from basic randomness of a season's worth of PAs is already huge (again, see Teixeira ... or Swisher who could basically perform at any level) and adding in the uncertainty of our mean estimate only shows how little we do know about individual players. And I don't know what can be done to try to improve that mean estimate other than possibly incorporating some biometric data (e.g. eye tracking ability), maybe psychometric, or things like pitch f/x (and what comes after).
By the way, how many of you knew that Matt Holliday is one of the 8 players in that 140+ group? The full list, in order: Pujols, AROD, Mauer, Teixeira, Berkman, Holliday, Pena, Gonzalez. I would have guessed at least those last 3 would have been in the 130s.
ZiPS assumes that a player's recent usage reflects his future usage. A human can better guess how Hughes will in fact be used.
As for the 4.32 ERA, that reflects ZiPS not knowing about the rib and muscle issues. I found back when I started doing this that removing seasons in which a pitcher spent at least 30 days on the DL made projections significantly less accurate - the injury performance doesn't go away because players with an injury have an increased risk of being injured. So what amounts to a 6.27 ERA in 2008 is part of his record.
IOW it's an educated guess that derives from a large database of past cases rather than an educated but non-provable belief (based on how he's performed since his recovery) that the injury was of a flukish sort. That's reasonable, and thanks for the explanation, even if I'd still bet on Hughes far outperforming that projection.
One thing I noticed, Teixeira shows as projecting to "AV" D at first, but has a 79% chance of being VG or EX according to ODDIBE . . . that doesn't make sense to me . . .
Tell me ZiPS - are you by chance a ... pleasure model?
You missed Chipper, Hanley, Manny, and perhaps a few others too. But it doesn't change the main point.
I like to refer to the "classic" case of regression to the mean, at least the one I recall most in introductions to the subject, characteristics of offspring. Tall fathers tend to have tall sons (compared to average height), and short fathers tend to have short sons. But tall fathers tend to have sons who are shorter than them, and short fathers have sons who are taller than them.
It is other words, the second random variable (son's height, or year X+1 OPS), conditioned on the first (father's height, or composite of OPS up to year X) will tend to regress part way toward the overall mean.
A couple key things: when projecting, you're conditioning on a random variable, not a known constant. "Tend to" doesn't mean "always." Even if we conditioned on true talent, we'd see some hitters not regressing toward that true talent. And, of course, players get better or worse all the time, so true talent is not constant across time.
I think the 79% is the likelihood of his offensive production being among the top 40% of first basemen.
Edit: Only 12 minutes late! Is that like, a 12 pack of cokes?
Sabathia's top comp might be scarier, given the contract length.
This would have to be quite the Madoff-like catastrophe.
(raises hand)
The guy that did surprise me in that group was Carlos Pena.
-- MWE
2007 really pulls that multi year average up.
FWIW BBREf PI says there are 14 playesr with 1200+ PAS, and OPS+ over 140 from 2007-09:
Cnt Player OPS+ PA From To+----+-----------------+----+-----+----+----+
1 Albert Pujols 178 2020 2007 2009
2 Alex Rodriguez 158 1837 2007 2009
3 Chipper Jones 152 1730 2007 2009
4 Prince Fielder 150 2094 2007 2009
5 Mark Teixeira 149 1967 2007 2009
6 Hanley Ramirez 147 2051 2007 2009
7 Manny Ramirez 147 1654 2007 2009
8 Joe Mauer 145 1704 2007 2009
9 Matt Holliday 144 2006 2007 2009
10 Lance Berkman 143 1895 2007 2009
11 Carlos Pena 142 1789 2007 2009
12 Ryan Braun 141 1863 2007 2009
13 Adrian Gonzalez 141 2101 2007 2009
14 Miguel Cabrera 140 2043 2007 2009
Of the 14, Pena is alone in that he has just 1 year over 130, that would seem to indicate that 140 is likely not his true talent level.
Hmmm ... I used PI too ... I must have accidentally set a criterion I didn't mean to set.
Anyway, the explanation in #51 is clear and simple. The key point I want to add regarding projection systems is that they are regression towards the mean conditional on past performance (usually a weighted function of the last 3-4 years) plus age and a few other variables. And that's fine -- I don't mean to question their utility as standard sorts of models. The problem comes in when they are interpreted as regression towards the player-specific mean.
Dan's presentation (and PECOTA's and I suppose most of them) are good in that they are displaying at least some of the uncertainty. What I'm not clear about is how that uncertainty is being estimated (feel free to link me if you want Dan).
He's an old players skills type (low BA, lots of walks, lots of Ks), so maybe ZiPs sees him petering out soon.
Also, this was just another (good) season for him - it's certainly not clearly his best.
That's certainly a possibility and the only reason that I can think of to explain the projection.
Also, this was just another (good) season for him - it's certainly not clearly his best.
His wOBA this year was .375, his next highest was .368 back in 2006. Maybe 'clearly' was too strong, but this is the best offensive season that he's had in his career. His defense was a bit better in 2006, but that shouldn't affect the offensive projections.
not kewl!
I note that his top comp is Dwight Evans, and Evans had his best offensive seasons from 29-32 and was productive for a long time after that. Burroughs, too, had a couple of productive seasons between 29-31, although he fell off much more rapidly than did Evans. It's very possible that Swisher could be very productive over the next year or two and then fall away fairly quickly.
-- MWE
Everyone's projection models have him slowing down, hitting maybe .280/30/90 next year and steadily declining thereafter, giving him a fairly low (say 35%) chance at 3,000 hits or 700 HR.
I'm having a little bit of trouble swallowing the idea that a guy who was arguably one of the most productive hitters ever up through age 33 is suddenly going to basically curl up and die, having statistics AFTER he turns 34 which aren't anywhere near the best in any category.
Seems to me that if a guy has demonstrated durability and productivity for 14 years at a rate which is arguably just as good as anyone who has played the game in a lot of ways, we ought to presume that will continue. I don't see any signs of him breaking down. He seems fully recovered from the hip. He still has good speed. He can still motor decently well on the bases. He can still hit tape measure HRs. He remains a defensive asset, playing a tough position well. He just doesn't in any way, shape or form look like someone who is done for. In fact, it wouldn't surprise me to see him, like a Ruth or Aaron or Bonds, have a string of great years in his mid 30s.
So, I need to ask this question. What is it about the career projection formulas which is driving this? Is it the fact that he had a huge year two years ago followed by a soso year followed by an injury year? Has anyone ever "gamed" the formulas to see if we need to put in an emergency override when the last couple of seasons have some quirks, such as the first in the series being out-of-context GOOD, followed by a last one out-of-context BAD?
It's just not sitting well to think that a guy can average 40 HR and a bunch of RBI and a great OPS for 14 straight seasons, then with no apparent physical issues all of a sudden he's gonna go from being best-in-class to soso at best, winding down in ignominy. If he was getting fat, had apparent substance abuse issues, couldn't run, SOMETHING, maybe I could go for it. But it just doesn't jive with what I see.
The problem is that you're focusing on the triple crown stats. There's nothing at all shabby about Rodriguez' ZiPS projected rate stats. An OPS+ of 141 isn't a projection of massive decline for a guy who went 150 and 147 in his last two seasons. It's all about playing time, and given that he's missed a good number of games the past two years, projecting him to get 700 PA in 2010 wouldn't make a whole lot of sense. If he plays 150 games this year, his counting stat projections for 2011 will reflect that.
Jimmie Foxx was an alcoholic of massive proportions and it finally caught up to him. I think you would also find that was the case in similar early dropoffs, e.g. Bobby Bonds, Frank Robinson, Mantle, some of the others. Ken Griffey's substance abuse problem was more a function of Dunkin Donuts, thus his mid career decline.
If you look at the guys who had decent mid-30s careers, one big attribute was, they were apparently free of substance abuse issues, didn't have career-changing injuries, and they stayed in shape. Ruth, Aaron, Cobb, Speaker, Mays, Williams, even guys like Brian Downing or Mickey Tettleton or Joe Morgan. Not that all those guys had their best years in the 33-38 area, but they had some good ones. It wasn't just the flat down curve.
I would suspect Arod will probably manage another 2-3 years with 40+ HR and that 50 again is not out of the question, and if his selectivity grows, he could pair that with 110 walks and have some fabulous offensive years with OPS+ in the 160 or more range.
Having said that, even his severe decline, off a career base that was considerably less than Arod's, still led him to a level that would have Arod scooting comfortably into 3,000 hits, and several more HR.
I'm not disputing that the projection systems often are quite correct. I'm just suggesting that in this case it's not smelling right, that we're missing something, and not by a little, but by an egregious amount, and I'm trying to put my finger on it.
Obviously he isn't going to turn into Emilio Bonifacio overnight but most players do not get get better or even hold steady in terms of either performance or playing time as they age. To suppport your hypothesis in #75 I think the onus is on you to identify something about Rodriguez that makes him likely to age better than the average superstar.
Projection systems are about setting baseline expectations, not about best (or worst) case scenarios. When Dan puts these up, I can look at any given player projection and do a little quick arithmetic in my head to come up with what a great year or a disappointing year will look like for that guy. It sounds like if you were doing the projections, I'd be looking at everybody's 90th percentile expectation. Or maybe 90th percentiles for the good to great players and 10th percentiles for the mediocre to marginal ones.
And since you brought up Ruth, he played 154 games at age 33 and never more than 145 in any subsequent season. He almost certainly would have had three or four more 50+ HR seasons if he'd been able to play every day at ages 34-37, but he wasn't able to do that. Probably had something to do with being 34-37 years old.
I think maybe there are a few things going on.
(1) My guess is the aging curve is flattening out. That players are peaking later than they used to and declining less steeply. Basically because workouts are allowing them to remain closer to their physical prime so that the net tradeoff between experience and loss of physical ability is resulting in less net deteroration. (1a) Sure would be interesting to see how the difference between smart and dumb players affects this. For example, Shawon Dunston retained his physical tools for a long time, but since he was dumb as a bag of rocks, he never really grew as a power hitter or a walker so his value declined when it could have stayed relatively constant.
(2) My guess is the aging curve affects superstars differently than regular players. Again for a few reasons. (a) I think there is some baseline level of ability you need to have to stay in the majors. Marginal players exist on the margin for years, then poof one day their ability is gone and they fall off the edge. Superstars can decline for years and still be viable players, so the dropoff is shallower. If you believe the aging curve, then I would suggest that the earlier one gets to the majors and thrives, there ought to be something in the formula to suggest they will have longevity. If someone was doing well at 19,20 chances are they will do OK at 39,40 relatively speaking. (b) Superstars will tend to get more chances - both because they have extended contracts and because someone will always be hoping they might snap back and be willing to take a chance on them. (c) Since their decline is shallower, they have more time to adjust and develop "old player skills" such as power and on base percentage, to compensate for lack of physical attributes. I'm gonna have to do a little bit of work to see if the decline curve works as well with superstars as it does with average players.
(3) I think there should be a "speed component" to the predictor. You can see when players lose their speed, decline follows quickly. You can also see that those who keep their speed tend to do better.
(4) I'm wondering about the weighting of the calculation. In other words, I see how the one calculation we tend to use works, and what it predicts. But step back. With logic, it would seem that I ought to be able to take a greater timeframe and say hey, if this guy hit more HR than anyone else (by a wide margin) up to age 33, why shouldn't that trend continue indefinitely? Why shouldn't he remain 5th or whatever he is (and has been for several years) in number of hits by age right up to age 40?
(5) What if you take apart last season, and say he came back from the hip too fast? If I recall, on Jul 1 he had missed one month, played poorly for two and was hitting about .240 with 10 HR. The second half of the year he batted about .315 with 20 HR.
I just think there is something wrong here. I don't tend to quibble too much with the projection system results. This one just seems all wrong.
Looking at your five points I would make the following comments;
#1, 2 and 3 are items that I assume are taken into account. I'm guessing tha ZIPS, CHONE and the rest don't assume that players today are aging the same as they did when Cobb played and since these are based on comparisons a lot of times Rodriguez' aging pattern is presumably modeled more after Mays, Mantle, Schmidt etc... than it is after Butch Hobson or Ray Oyler. 3 is a bit of an offshoot of 1. If Sean or Dan or someone who does these stumbles on this they can accurately answer some of what I am assuming here.
#4 - Trends don't continue indefinitely. The top HR hitters through the age of 30 include a bunch of guys but Aaron is "just" tenth, Ruth 15th and Bonds 26th. Players who are great in their 20s are not always great in their 30s. Why would Rodriguez finish ahead of those guys, wouldn't you want to model him after the guys most similar to him, the guys that were 2nd, 3rd or 4th?
#5 - I'm not sure why you would do that. Without looking just about every season by every player in baseball history has ups and down. You seem to be crediting Rodriguez for having the hip injury while the system seems to be debiting him for it. I don't think a serious hip injury is something I would use as a positive indicator. Guys who were hurt at the age of 33 are not less likely to get hurt at 34, 35 or 36.
For one thing, there are several "speed proxies" in the projections -- notably stolen bases, and caught stealing. For another, A-Rod has been losing steps every year. When he came to the Yankees, he was still a relatively fleet SS. Now he's much heavier and slower. He stole half as many bases last year as he did in his first season in NY.
But the really important thing is what Jose said in 80 -- players who miss time in their early 30s are significantly more likely to miss time going forward than players who don't.
You must be Registered and Logged In to post comments.
<< Back to main