Seven Systems Project the NL East

by Matt Swartz on Feb 21, 2009 1:48 AM EST Comment 31 comments

(Update: I've fixed up the Braves roster, updated the Baseball Prospectus projections, and replaced Sean's own Chone projections' standings rather than my approximation of his methods)

Unsurprisingly, winning the World Series did not seem to make the Phillies the NL East favorites. The magic bullpen fix that fat men in jerseys on the subway told Omar Minaya was necessary to fix the Mets’ problems apparently did the trick of fixing the Mets, according to the media and Vegas traders. The Mets currently have a 9% chance of winning the World Series according to Vegas and the Phillies only have a 6% chance of repeating. Last week, Baseball Prospectus released their projections for the Phillies. Their projections are proprietary, but there standings’ projections are not, so I will post those here with a link:

Baseball Prospectus Standings (updated based on recent edit by Clay Davenport):

TEAM	W	L
NYM	90	72
PHI	87	75
ATL	85	77
WAS	75	87
FLA	69	93

I did projections for the Phillies, Mets, and Braves, and while I do not entirely go by a formula, I highly doubt I was systematically biased. I do not think their projections for the Phillies were very far off at all, but I think that Baseball Prospectus’ PECOTA system was way off on many Mets and Braves players. I think that Baseball Prospectus’ system is spitting out insanely high BABIP projections for many Mets hitters—for Beltran and Reyes, for example, they project higher BABIPs than any of their BABIPs in the last few years. Rather than questioning all of their methods here, I will simply post my standings for the projections I did, and some implied standings (based on my roster guesses, which probably does not affect the numbers much if I am wrong for a few bench players) for many other projection systems, all of which put the Phillies in first for 2009. Then I will summarize my projections for the Phillies, Mets, and Braves.

MattS’ projections for the Phillies, Mets, and Braves:

TEAM	W	L
PHI	90	72
NYM	87	75
ATL	80	82

As for the other projection systems…

CHONE projections for the Phillies, Mets, and Braves (I've replaced with Sean's actual projections for these, since he seems to have a systematic way of doing this, and may very well be more appropriate):

TEAM	W	L
PHI	87	75
NYM	86	76
ATL	86	76

ZIPS projections for the Phillies, Mets, and Braves:

TEAM	W	L
PHI	91	71
NYM	88	74
ATL	85	77

Marcel projections for the Phillies, Mets, and Braves:

TEAM	W	L
PHI	90	72
NYM	90	72
ATL	89	73

(This actually had the Phillies at about 90.4 and Mets at 90.3, so it’s still technically true that the rest of these systems all had the Phillies ahead.)

Ron Shandler projections for the Phillies, Mets, and Braves: (these are proprietary but that website is where to get them)

TEAM	W	L
PHI	95	67
NYM	92	70
ATL	84	78

Bill James projections for the Phillies, Mets, and Braves: (these are proprietary but that website is where to get them)

TEAM	W	L
PHI	93	69
NYM	92	70
ATL	91	71

(here, the Braves did finish a few points ahead of the Phillies).

Also, none of these Braves’ projections except maybe for Baseball Prospectus, including Kawakami, their new pitcher who does not have MLB or MiLB stats yet.

All in all, this average to:

TEAM	W	L
PHI	90.4	71.6
NYM	89.3	72.7
ATL	85.7	76.3

Looks like another photo phinish.

I updated my Phillies projections a little bit (actually, just for Rollins, up a little bit, since he was the only one who seemed to significantly look different afterwards) based on a slightly updated version of the model that I did on a blog that I am now also writing for called StatSpeak.net, but here are my Phillies, Mets, and Braves projections that led to the projections I got above. I do think that the Braves are a bit more of a risk than my numbers are saying, only because their outfield is such a gaping hole that I think if they are close near the break, they will have a huge ability to improve by bringing in even an average couple of outfielders.

PHILLIES:

HITTERS (AVG/OBP/SLG)

Ruiz: .266/.348/.370

Howard: .265/.358/.547

Utley: .291/.375/.505

Feliz: .264/.309/.429

Rollins: .291/.351/.462

Ibanez: .280/.340/.468

Victorino: .281/.339/.434

Werth: .266/.360/.478

Coste: .279/.328/.432

Paulino: .272/.324/.390

Dobbs: .275/.320/.467

Bruntlett: .243/.315/.328

Jenkins: .246/.300/.394

Stairs: .254/.340/.416

PITCHERS (IP/ERA)

Hamels: 206/3.45

Myers: 190.2/4.11

Blanton: 197.1/4.83

Moyer: 200.1/5.08

Happ: 171.1/4.83

Lidge: 67.2/3.19

Madson: 80.1/4.26

Romero: 41.1/3.70

Durbin: 74.2/4.70

Eyre: 58/3.72

Condrey: 70/4.89

Park: 98.2/4.65

METS

HITTERS

Schneider: .258/.332/.338

Delgado: .260/.337/.443

Castillo: .281/.360/.337

Wright: .301/.381/.518

Reyes: .293/.350/.454

Tatis: .275/.350/.460

Beltran: .281/.365/.496

Church: .252/.330/.424

Castro: .243/.305/.421

Murphy: .284/.346/.424

Evans: .258/.309/.393

Reed: .242/.277/.335

Cora: .234/.310/.314

PITCHERS

Santana: 234.2/2.76

Maine: 140.1/4.10

Pelfrey: 178.2/4.48

Perez: 186.1/4.65

Redding: 179/4.83

Rodriguez: 66/2.73

Putz: 55.1/3.58

Sanchez: 58.1/3.70

Feliciano: 56.1/3.36

Stokes: 59/4.27

Muniz: 51.1/4.73

BRAVES

HITTERS

McCann: .300/.356/.509

Kotchman: .291/.338/.435

Johnson: .295/.358/.450

C. Jones: .334/.422/.521

Escobar: .301/.367/.421

Diaz: .280/.315/.399

J. Anderson: .250/.294/.328

Francoeur: .261/.305/.409

Sammons: .232/.335/.411

Prado: .298/.354/.398

Infante: .261/.307/.386

Blanco: .243/.360/.322

B. Jones: .263/.306/.401

PITCHERS

Vazquez: 211/4.27

Lowe: 202.1/3.74

Jurrjens: 191/4.43

Kawakami:156.2/4.65

Glavine: 122.2/4.70

Gonzalez: 44/3.89

Soriano: 48.1/4.10

Acosta: 55/4.25

Boyer: 67/5.10

Logan: 50/4.86

Moylan: 50.2/2.84

Bennett: 81.1/3.98

It seems like I regress a bit more to the mean on my projections than most systems. I think that I am somewhat bearish on Ryan Howard for once, and I am certainly very bearish on Delgado and a little bit on Pelfrey too. I also have virtually no confidence in anyone in the Braves’ outfield or bullpen. Keep in mind that if you predict a team’s abilities perfectly, random chance will send them five wins or more in either direction half of the time. So really the Phillies odds of winning the division are probably less than 50%, but given that the NL East has about a 33% chance of having a wild card, I would approximate our playoff odds at 65% or so, making our odds of repeating approximately 8%. It sounds low until you think about what 8% odds of a dynasty might sound like. If I ask who won the World Series in a random year, most baseball fans might not know or remember the answer, but most fans remember the teams that won back-to-back World Series. I’ll take 8% odds of that.

0 recs | Comment 31 comments |

Story-email Email Printer Print

Comments

Regression??

You have a huge regression on pretty much every pitcher on the Braves roster. Maybe you should read a couple of professional publications that suggest that the Braves will have one of the best pitching staffs in the majors this year. The bullpen is so good that they will end up demoting, trading, or cutting at least 4-5 guys that would be on the roster of probably 80% of all other ML teams. I love that you did the work but the bias against Atlanta is so evident that it is almost laughable. Just my thoughts. Have a great weekend.

by KC Ryan on Feb 21, 2009 10:03 AM EST reply actions

proof

Rather than spout “BIAS!” maybe you can show us some proof. Appealing to the authority of the “professional publications” who didn’t give the Phillies a shot in hell at beating the Dodgers or the Rays in the postseason last year is a pretty weak argument, in my opinion. It’s amusing that you’re raking a guy who is really “showing his work” (to use a middle school math class term) over the coals.

http://www.thegoodphight.com

by WholeCamels on Feb 22, 2009 10:41 AM EST up reply actions

Not biased, but would love to know if I'm wrong

Thanks for the comment. I definitely am not biased against the Braves bullpen at all, but if my method could very well be pessimistic. I don’t know a lot of those guys well and mostly used weighted averages of previous K/BB ratios and GB% to help me determine their performance. Of the two main guys I have followed, I have always been pretty bullish on Soriano and Gonzalez both, and have let them frustrate my fantasy teams for the past couple years.

I’m also not really sure how you can suggest I read professional publications, since I quite clearly listed 6 publications which I copied data into spreadsheets from in this very article, but I understand that the Braves’ bullpen were certainly the players I knew the least about. Who particularly surprises you? It seems like I was more or less at the median for the how I projected the Braves’ bullpen compared to the 6 projection systems above.

There definitely is a wide disparity in how different projection systems view the Braves, with Bill James picking them to be perhaps the best in the division and ZIPS not really viewing them as legit contenders. I’d love insight if there’s someone that you think should be higher.

by Matt Swartz on Feb 21, 2009 12:19 PM EST reply actions

Could you please post your “method,” so we can evaluate for ourselves why it might be pessimistic?

If you’re mainly based on groundball rates and K/BB, I’m not sure how you can have such high ERAs for Vazquez (incredible K/BB rates) and Jurrjens (great GB%).

Also, leave Sammons off. Dave Ross will be the Braves’ backup catcher.

by tgthree on Feb 21, 2009 5:12 PM EST up reply actions

I figured that Jurrjens would get about 31 starts. I figured 100 pitches per start, and 3.8 pitchers per batter. That gave me 816 batters faced. Last year, he struck out 17.1% of hitters and unintentionally walked 7.5. I put him down for 17.0 and 7.5% for this year. So that was 61 UBB and 139 K. I approximated pretty similar numbers for SH, SF, SB, CS, GDP, ROE, etc. as last year. I figured his IBB were high at 9 and put him down for 5. For balls in play last year, he had 21.9 LD, 51.5 GB, and 26.5 FB. This year, I figured him for 20/50/30 respectively. I put him down for 11 HR/Flyball leaving him with 20 homeruns, unlike last year’s 11. Then I looked at that 20/50/30 split, and based on the Braves’ defensive abilities on those balls in play last year, removed his 20 homeruns, that gave him .308 BABIP, similar to last year’s .307 BABIP. Last year, Jurrjens surrendered 40 2B and 8 3B on 589 BIP. That seemed high for a groundball pitcher, and I bumped him down to 30 2B and 6 3B. For the outs he got that way, and with all the double plays, caught stealings etc., which were similar to last year, that came out to 191 IP. That left him with a .331 OBP and .410 SLG surrendered. Using the RC formula, that approximated 102 runs surrendered. That sounded like about 94 ER in 191 IP. His FIP in this method is just 4.19 and his QERA with this method is 4.30. The 4.43 ERA is a little higher than that, and I suspect that’s because I think his BABIP will be a bit higher than league average.

I did a very similar method for Vazquez, getting 5.9% UBB and 22.5% K, with 38.5 GB, 41.5 FB, and 20% LD. I figured 11% HR/FB as well and ended up with very good FIP and QERA as well. But I guess my suspicion is that his very high XBH% on balls in play will continue, and I brought him down to 45 2B and 5 3B instead of the 54 & 7 from last year, but I still got him surrendering a .424 SLG and .307 OBP.

by Matt Swartz on Feb 21, 2009 5:39 PM EST up reply actions

I don’t think the 4.19 FIP is that far out of line (CHONE has him at 4.17, though other systems are friendlier about his regression in HR allowed). What I don’t understand is where you get that extra 20 points from 4.19 to 4.43. The Braves’ infield defense looks to be very good, so there’s little reason to imagine his BABIP will skyrocket from where it was last year.

As for Vazquez, Turner Field is far roomier than the Cell, and the outfield defense also figures to be excellent (Matt Diaz and Gregor Blanco are plus in left field, Josh Anderson and Jordan Schafer are plus in center and Jeff Francoeur is plus in right). So now, instead of pitching in front of Carlos Quentin (-4 UZR/150), Nick Swisher (-10 UZR/150), Jermaine Dye (-20 UZR/150), he’s pitching in front of three plus defenders, AND the ballpark is going to hold more of his flyballs. I’d think you ought to knock his XBH numbers down even further than that.

by tgthree on Feb 21, 2009 7:13 PM EST up reply actions

Jurrjens: I gave him the same BABIP as last year. I tend to trust QERA more than FIP and 4.30 and 4.43 are pretty similar. Not enough to make much of an adjustment.

Vazquez: True, the OF defense will be better, but the roomier field will lead to more 2B and 3B, at least according to park factors that I’ve looked at. Vazquez always seem to under-perform his peripherals and has done so for long enough that I’m comfortable projecting him to do so again. Anyway, even if I bumped down his doubles 10, he still would only 3 or 4 runs different. Not really enough to make a huge to difference.

by Matt Swartz on Feb 21, 2009 9:11 PM EST up reply actions

no bias?

I think you have 3 Braves pitchers with an ERA under 4. The Braves starters are supposed to be the best in the division this year, and most critics are just as high on the bullpen. Question, how do you have Jurrjens with such a terrible ERA after it was 3.68 in his rookie season when he was basically our number 1, once Hudson was hurt? Please at least spell his name right.
Good work on the compilation though, thanks.
One more thing, Glavine will be our 5th starter instead of Campillo.

by acie4mvp on Feb 21, 2009 2:02 PM EST reply actions

Jurrjens

Spelled it right, ok? Yeah, I never got the spelling gene, just the math one. Sorry :)

I rarely project anyone under 4 ERA unless I’m convinced. If you look at the other teams, I’m pretty heavily regressive to the mean on pitchers.

I think that Jurjjens got a bit lucky last year in a couple of ways. His HR/FB was only 7. Either he’s got a fancy Mariano Rivera cutter, he was playing in Petco all year, or he was lucky. I put him down for the exact same BB and K% as last year, pretty rare for me to expect a strong repeat performance from a rookie actually, and left his GB% at 50. I just put his HR/FB at 11 (league average) at looked at his expected OBP and SLG and used that for his expected runs surrendered. Also, Jurrjens allowed 87 runs last year but his OBP and SLG suggested he should have allowed 7 runs more. He seemed to get out of a lot of jams, I guess.

I’ll put in Glavine soon, but it won’t change the results much.

by Matt Swartz on Feb 21, 2009 2:42 PM EST up reply actions

well...

I’ll admit I’m biased against the Braves; I hate them the way most here hate the Mets.

But I don’t think it’s their bullpen that will leave them in 3rd or 4th; it’s the lineup. McCann is a stud, and Larry Jones Jr. will be great for the 400 or so plate appearances he’s likely to be around for. Johnson and Escobar are above-average bats for their positions. The other half of the lineup flat-out stinks. How Atlanta didn’t get in on Adam Dunn or Bobby Abreu is absolutely unfathomable to me; either of those guys might have lifted them into rough parity with the Phillies and Mets. As it is, though, they look like an 80 to 85-win team to me.

by dajafi on Feb 21, 2009 2:06 PM EST reply actions

Thanks for the breakdown

Glavine will, unfortunately, be the #5. Hoping Hanson comes up by the ASB. I doubt Carlye, Campillo, AND Bennet will be on the roster, they’re basically the same pitcher.

The OF is still terrible, I hope there’s a trade in the works.

by TradeAndruw on Feb 21, 2009 2:26 PM EST reply actions

True, the Braves could significantly improve with a new OF.

Who do you think will be the bullpen?

by Matt Swartz on Feb 21, 2009 2:43 PM EST up reply actions

Moylan is the big one you’re missing, I’d like to see Stockman too.

RF is going to be awful. CF is going to stink unless Schafer or Anferson take a huge step forward. LF is probably going to be a platoon with Diaz and B. Jones/Edmonds/Conrad. Not pretty. Then again, Nady/Swisher aren’t worth the price.

by TradeAndruw on Feb 21, 2009 3:07 PM EST up reply actions

Why are the Chone standings you list here different than the ones listed on the website of the guy who does those projections. For the record, he has the NL East as:
Phi-87
NYM-86
Atl-86

by quake902 on Feb 21, 2009 4:45 PM EST reply actions

I’m guessing I plugged in different players into the roster or different playing times. Maybe he changed his projections recently, too, but I’m guessing it’s a playing time issue. I’m kind of surprised he got the Phillies at 87 wins though. I’ll have a look.

by Matt Swartz on Feb 21, 2009 4:56 PM EST up reply actions

I looked at it a little more. It seems like he is valuing things a little bit differently than me. While his projections have changed a little bit and playing time might be a bit different in my numbers than his, I think he values performances differently than me. For the ERAs he lists for the Phillies’ pitchers, I think those are worth more runs over replacement than he does, and for the Mets’ pitchers, I think they are worth about the same as the ones he listed, maybe not quite as much. I guess I’m figuring on CBP being a much stronger hitter’s park. I’m not sure how he does his hitter run values but even those he seems to give the Phillies a bit less credit than I do for those numbers themselves. I try to approximate Baseball Prospectus’ VORP numbers for each team and sum them up. I’m not sure how he does his run values. That should be the difference.

by Matt Swartz on Feb 21, 2009 5:17 PM EST up reply actions

Makes sense. TY for the explanation.

by quake902 on Feb 22, 2009 2:32 AM EST reply actions

i dont see Atlanta winning that many games even with Derek Lowe

To me the Eagles are like that girlfriend that pisses u off, then u hate her, but the next morning you wake up and remember that thing u love so much about her

by XxBleedGreen5xX on Feb 23, 2009 8:44 AM EST reply actions

Projecting the NL East

I don’t play fantasy baseball and I have no idea what it is you’re talking about when you discuss the projections from the seven “experts” about who’ll win the NL East in 2009. What I do know is that the game is played on the field. The Phillies in 2008, and all other World Series winners in the past, did not use a calulator to win when their games. All of this mumbo jumbo is for the birds! All of the “experts” projections will be decided when the season ends. Some teams will win and some will lose, and it won’t be because BP or Bill James or anyone else “projected” they’d win. Those of you who are so enamoured with all this projection business should take your calculators and help Mr. Obama project our way out the economic morass in which we find ourselves. That seems to me to be a much better way for all of you who play fantasy baseball and project baseball outcomes to spend your time.

by philliemike on Feb 23, 2009 2:25 PM EST reply actions

CLICHE NONSENSE

A) If you don’t want to read it, don’t read it. It certainly seems like a lot of people care.
B) If you think the Phillies don’t use a calculator somewhere in their front offense to win games, you must be living under a rock. And a very large rock at that.
C) Sharpening skill at evaluating baseball is an extremely valuable way of learning about economics. There are fewer industries with such clear and defined data on productivity and salaries, and there is more than one economist who work on this kind of research because it improves our ability to understand other economic phenomena.

No one thinks that that the Phillies win because I say they will. If that were true, I would certainly write here more and be more optimistic in the process.

by Matt Swartz on Feb 23, 2009 2:37 PM EST up reply actions

No, MattS

The Phillies won last year because of HEART and HUSTLE and GRIT and UNICORNS and FAIRY DUST. It had nothing to do with quantifiable performance on the field. Stop being a poindexter and put your calculator away.

http://www.thegoodphight.com

by WholeCamels on Feb 23, 2009 2:59 PM EST up reply actions

side note

I love how the “fantasy baseball” crowd and the “statistical analysis” types have been conflated by the knuckledraggers because they both use scary, confusing numbers and are, therefore, nerds. Any stathead worth his salt loathes the traditional roto stats.

http://www.thegoodphight.com

by WholeCamels on Feb 23, 2009 2:57 PM EST up reply actions

Great work Matt

When the projections come out this close for the Mets and Phils, it’s pretty much a crapshoot, isn’t it? I know that’s not technical statistical lingo, but that’s the reality at this point?

by David S. Cohen on Feb 23, 2009 3:49 PM EST reply actions

I don’t think any objective person could say the NL East has been anything but a crapshoot for a few years now.

http://www.thegoodphight.com

by WholeCamels on Feb 23, 2009 3:53 PM EST up reply actions

Homer Alert

THe Homer Alarm is going off.

by Metzfan22 on Feb 23, 2009 3:49 PM EST reply actions

greetings

What’s up, man?

http://www.thegoodphight.com

by WholeCamels on Feb 23, 2009 3:58 PM EST up reply actions

Definitely

MattS certainly used Philly-friendly numerals and regression techniques. I think they were developed in the Ancient Greeks.

by David S. Cohen on Feb 23, 2009 4:55 PM EST up reply actions

The very name

“Philadelphia” itself is Greek. Coincidence? I think not!

by phatj on Feb 23, 2009 5:56 PM EST up reply actions

check out my post history and see if i have much of a history of being a homer. no? then i guess it’s just math. i just plugged in numbers. who do you think i’m wrong about and what evidence do you see? my phillies win projection is lower than the average of the other systems.

by Matt Swartz on Feb 23, 2009 8:42 PM EST up reply actions

Mets and Phillies

All the projections are very close. When I look at the teams they seem very even. I would put much more money on this going down to the last week of the season than I would on either team.

by supermets on Feb 23, 2009 6:50 PM EST reply actions