Wednesday, January 31, 2007

MLB Preview 2007 – NL Bullpens

Today’s post focuses on the quality of the NL bullpens. I discussed the methodology and inherent difficulty in bullpen rankings in yesterday’s post.

Above Average
New York Mets: 3.79 EqERA
Atlanta Braves: 3.90 EqERA
San Diego Padres: 3.90 EqERA
Houston Astros: 3.93 EqERA

As I discussed in my post on NL Rotations, the weakness of the NL at present appears to be in the pitching department. While the quality of starting pitching in the AL is markedly better than in the NL, the difference in relief pitching is insane. Ten of fourteen AL teams finish with above-average bullpens, while only four of sixteen NL teams grade out likewise. Additionally, the Mets bullpen would not crack the top seven in the AL! The difference in bullpen quality is simply massive. The Padres have had some impressive bullpens in recent years, but a lot of the observed quality is park and defense aided. One also has to wonder if the observed dominance of AL pitching is due to an incorrect application of league strength factors by PECOTA (explanation: PECOTA first projects a raw ERA for pitchers, then creates an EqERA based on the ballpark and league they play in. The translation requires an estimation of the magnitude of these two effects). However, I think the most likely explanation is simply that bullpen quality in the NL is quite poor right now.

Los Angeles Dodgers: 3.97 EqERA
Arizona Diamondbacks: 4.01 EqERA
Milwaukee Brewers: 4.04 EqERA

Only three more teams approach the MLB bullpen average EqERA, including popular sleeper picks Arizona and Milwaukee. Maybe I was overly critical of Milwaukee’s break-out chances in the team fielding article.

Have one good reliever
Chicago Cubs: 4.11 EqERA
Philadelphia Phillies: 4.16 EqERA
Cincinnati Reds: 4.17 EqERA
Pittsburgh Pirates: 4.18 EqERA
Washington Nationals: 4.23 EqERA
Colorado Rockies: 4.24 EqERA

A bunch of teams with one or two good relievers, but not enough depth to be a truly strong bullpen. The Cubs have the most upside with Kerry Wood and Neal Cotts, but PECOTA does not expect big things out of them presently. The Phillies have the strongest starting pitching in the group, so the lack of real depth may hurt them least.

Aspire to one day have one good reliever
St. Louis Cardinals: 4.34 EqERA
San Francisco Giants: 4.48 EqERA
Florida Marlins: 4.54 EqERA

The bottom three here really lack both frontline talent and depth in their bullpens. The Marlins, like the Devil Rays and Royals, are trying to develop starting pitching talent before devoting significant resources to the bullpen, but the Cardinals and Giants are both supposed to be contenders this year. Starting pitcher defections, which will force their young starters into the rotation and out of the bullpen, have hurt the Cardinals. The Giants have simply lacked front-line bullpen talent for the past few years. Peter Gammons recently speculated that the Giants could utilize prospects Brian Wilson or Tim Lincecum in the closers role. PECOTA does not like Wilson, but thinks Lincecum is a once-a-generation pitching prospect. Indeed, put Lincecum in the closer’s role, and the Giants pen improves to a 4.14 EqERA. This year’s Jonathan Papelbon?

Tuesday, January 30, 2007

MLB Preview 2007 – AL Bullpens

Today’s post focuses on the quality of the AL bullpens. Ranking bullpens, especially this early in the offseason, is tricky business for three reasons:

1. Relief pitchers are the most volatile and difficult to project species in the MLB universe.
2. Major league bullpens are where you will most likely find reclamation projects, guys given a second chance, and career minor leaguers. Going into the season, teams may only have four or five of the seven bullpen spots assigned, with the last two comprising a rotating group of whoever looks in spring training.
3. All bullpen roles are not created equal. The top three relievers in the bullpen typically pitch the most important innings, led by the putative closer. Thus, the quality of only three pitchers will dominate the overall bullpen ranking.

Fortunately, the last point also allows us to get a general idea of bullpen quality without the benefit of spring training outcomes, because most teams (I’m looking gloweringly at you Boston) already have an idea who the top three relievers are. Still, all these numerical projections should be taken with a grain of salt. I’ll provide comments when the numerical rank may not be providing a true view of reality.

As with the previous posts on starting pitchers, the following projections use PECOTA’s EqERA, with appropriate weighting applied for expected workload. In addition, a multiplier is applied to each relief pitcher’s EqERA that accounts for the importance of the situation he pitches in.

Top 2
Minnesota Twins: 3.20 EqERA
Texas Rangers: 3.23 EqERA

In the first grouping, we already see why the reliever ranking process is so difficult. Minnesota is a legitimate #1 on this list, with six quality relievers. Joe Nathan was the most dominant closer in the game last year, and PECOTA expects that to continue. The Rangers, on the other hand, are ranked here largely on the strength of the Eric Gagne projection (2.17 EqERA). Take out Gagne, and their bullpen EqERA increases to 3.64. This is not to suggest a flaw in the methodology, but rather to suggest that the performance of an entire bullpen can be transformed by the addition of a single player.

Strong to…quite strong
Oakland Athletics: 3.52 EqERA
Anaheim Angels: 3.56 EqERA
New York Yankees: 3.58 EqERA
Toronto Blue Jays: 3.62 EqERA

All of the teams above are characterized by a relief ace (which I’ll define as a top-10 closer) and at least two additional quality relievers. The thing that separates them from the top two is generally weakness in the 4-7 spots of the bullpen.

Baltimore Orioles: 3.77 EqERA
Cleveland Indians: 3.82 EqERA
Detroit Tigers: 3.84 EqERA
Chicago White Sox: 3.92 EqERA
Seattle Mariners: 3.98 EqERA

Continuing with the theme of AL pitching dominance, a full ten AL pens come in below the league average EqERA of 3.96, with the Mariners close at 3.98. The five pens listed here in general have closers of a caliber below that of the previous groups, and thus receive a lower ranking. Cleveland is a tough bullpen to decipher because it is not clear who will be pitching the important innings. PECOTA likes Keith Foulke and Rafael Betancourt a lot, and this projection assumes they will be pitching key innings next year. However, both struggled last year. Additionally, Cleveland has imported “proven closer” Joe Borowski, who projects as quite poor. Plugging him into a critical role drops the Indians out of this grouping entirely.

Late inning losses are our specialty
Boston Red Sox: 4.12 EqERA
Tampa Bay Devil Rays: 4.21 EqERA
Kansas City Royals: 4.31 EqERA

The bottom ranking contains two teams you would expect to find here, and one surprise. Tampa Bay and Kansas City are not surprise cellar dwellers, as it does not make sense for a team to devote resources to the bullpen until the have developed starting pitching assets. The Red Sox, by contrast, have the best projected starting pitching in MLB. However, the bullpen looks like a mess at this late date. They have some average-ish pitchers in Brendan Donnelly, Mike Timlin, and Manny Delcarmen, an unknown in Japanese import Hideki Okijama, a rookie with upside in Craig Hansen, and a converted starter in Joel Pineiro. PECOTA is not high on any of these players. Much like the Mets in the starting pitching department, the Red Sox had a clear need this offseason (bullpen), and found the available options unpalatable. Perhaps history will vindicate them for not overspending on risky free agents like Eric Gagne, but I think the Red Sox will regret having such an obvious hole in the team.

Monday, January 29, 2007

MLB Preview 2007 – NL Rotation Analysis

As with yesterday’s post, I’ll be looking at projected starting rotations, this time of the NL teams. This is the first time I have ever attempted such a venture, so the method may be a bit raw, and in need of refinement. You can see the algorithm used in the previous post.

Without further ado, the results:

Tier One
Arizona Diamondbacks: 4.29 EqERA, 977 IP
St. Louis Cardinals: 4.36 EqERA, 944 IP
San Diego Padres: 4.36 EqERA, 980 IP
Philadelphia Phillies: 4.39 EqERA, 966 IP
Milwaukee Brewers: 4.40 EqERA, 990 IP

It’s difficult to come up with a snappy title for the first group of NL teams, since only the Diamondbacks would crack the top five of the AL list. The NL lacks both top-line talent and depth when compared to their AL counterparts. The Diamondbacks show up on the list thanks to two of the NL’s best starters, Brandon Webb and the returning Randy Johnson (PECOTA sees a big bounce-back for Mullet Man). I am shocked – SHOCKED – to see St. Louis in the second spot on the list. PECOTA loves three of their young starters in Anthony Reyes, Adam Wainwright, and Brad Thompson. Still, the young age of their rotation is reflected in the low projected Innings total. The composition of the St. Louis bullpen will be critical to their success next year. The Padres and Phillies have the deepest rotations in the NL. The Brewers probably have the strongest starting trio in the NL with Ben Sheets, Chris Capuano, and Dave Bush, who can all pitch deep into games. However, PECOTA does not like Jeff Suppan, projecting a nearly 5.00 ERA. Still, having four starters who have the potential to average 6 innings or more per start is incredibly valuable.

Florida Marlins: 4.49 EqERA, 964 IP
Chicago Cubs: 4.51 EqERA, 950 IP
Los Angeles Dodgers: 4.54 EqERA, 948 IP
Atlanta Braves: 4.56 EqERA, 930 IP

The team average EqERA is 4.50, making this quartet decidedly average. However, they achieve average-ness in quite unique ways. My intuition was that the Dodgers would appear far higher on the list, but the reality is that none of their top three (Jason Schmidt, Derek Lowe, Brad Penny) is truly an elite starter, and Chad Billingsley is nowhere near the pitcher he appeared to be last year. Florida surprised the world last year with their pitching staff, which has the advantage of being young, but once again they truly lack a star-level performer on the staff. The Cubs projection includes a projected 4.66 EqERA from Mark Prior in the five spot. Prior seems to me to be an extremely high variance player at this point. He could fail miserably, or he could far exceed that projection. If Prior regains some of his lost ability, the Cubs will definitely contend for strongest rotation in the NL. Prior could also fail to produce, forcing more starts to go to Jason Marquis, Wade Miller, or Sean Marshall, and the Cubs could rank in the bottom half of the league. As Chuck Klosterman would say, it's 50/50.

Below Average
Cincinnati Reds: 4.62 EqERA, 955 IP
Colorado Rockies: 4.65 EqERA, 949 IP
Pittsburgh Pirates: 4.68 EqERA, 931 IP
San Francisco Giants: 4.72 EqERA, 956 IP
New York Mets: 4.73 EqERA, 904 IP
Houston Astros: 4.74 EqERA, 938 IP

A diverse group of teams here, from perennial rebuilders Colorado and Pittsburgh to perennial playoff contenders New York and Houston. Houston has very little pitching depth after the loss of both Clemens and Pettitte. I think New York is going to truly regret not overpaying for at least one starter this offseason, as the rest of the team has the talent to take them to the World Series. As it stands, I think the rotation is weak enough to open the door for both Atlanta and Philadelphia to win the division. Cincinnati, while still below average, has improved their pitching tremendously in the last year, when they had exactly zero pitchers (starters or relievers) projected to be above average. San Francisco seems to me to be the White Sox of the NL in terms of starters – lots of recognizable names, lots of innings eaters, but no true stars (although Matt Cain could be that guy).

Bottom of the Barrel
Washington Nationals: 5.29 EqERA, 841 IP

Quite simply, the KC Royals of the NL, without the promise of better times ahead. In my NL Lineup preview, Washington came out last after the Pittsburgh addition of Adam LaRoche, and they come in last in pitching by a wide margin here. After John Patterson, the Nationals will run out the motley collection of Sean Hill, Jerome Williams and Mike O’Conner. They definitely have a very good shot at a 100 loss season.

Incorporating Innings Pitched
As a last step, I incorporate an innings pitched credit or debit. Teams with more starter IP get a credit, teams with less a debit. This produces a linearized ranking of NL teams that is intended to represent true rotation value. The ranking is as follows:

1. Arizona Diamondbacks
2. Milwaukee Brewers
3. San Diego Padres
4. Philadelphia Phillies
5. Florida Marlins
6. St. Louis Cardinals
7. Chicago Cubs
8. Los Angeles Dodgers
9. Cincinnati Reds
10. Colorado Rockies
11. San Francisco Giants
12. Atlanta Braves
13. Pittsburgh Pirates
14. Houston Astros
15. New York Mets
16. Washington Nationals

Milwaukee moves up to number 2 due to their high projected innings pitched. Looking at the list above, one can see why Arizona and Milwaukee are popular picks to break out this season.

Sunday, January 28, 2007

MLB Preview 2007 - AL Rotation Analysis

Today’s post is a look at the projected starting rotations of AL teams. This is the first time I have ever attempted such a venture, so the method may be a bit raw, and in need of refinement. The general procedure was as follows:

  • For each team, I identified the projected starting five, and up to two alternates. These alternates were in some cases true swingmen (like Kirk Saarloos on Cincinnati), in other cases rookies expected to make midseason call-ups (think Philip Hughes of the Yankees).
  • With great difficulty, I restrained myself from choosing my preferred rotation, instead relying on my best estimate of the team’s projected rotation. So, my hometown Twins will be running out Ramon Ortiz and Sidney Ponson, instead of Matt Garza and Scott Baker.
  • Pitchers 1-4 are projected to get 32 starts, number five gets 19, six gets 9, and seven gets 7.
  • Pitcher projections are PECOTA’s EqERA, which is ERA adjusted for league, park, and team defense.
  • Pitcher Innings Pitched per start is also projected by PECOTA.
  • Teams are ranked in order of projected ERA. This ranking method does not truly capture rotation value, because it does not account for the number of innings contributed by starting pitchers. All other things equal, more IP contributed by starters is better, because in MLB, the starting pitchers are in general possessed of superior talent to the relievers. For this reason, I have also included a projection of the number of innings contributed by starters (out of 1440 IP).
  • The average rotation EqERA on an MLB scale is 4.50 EqERA.
Without further ado, the results:

Top 2
Boston Red Sox: 4.01 EqERA, 992 IP
Anaheim Angels: 4.06 EqERA, 994 IP

Boston gets top billing on the strength of its top five, with Schilling and Matsuzaka projected amongst the league’s top starters. Wakefield, Papelbon, and Beckett all project as above-average starters as well. However, Boston has almost no depth beyond their top five (Kyle Snyder anyone?), so will definitely deal with downside risk. The quality of the Angels top five is slightly below Boston’s, but they have two reliable backups in Joe Saunders and rookie Nick Adenhart.

Above Average
New York Yankees: 4.22 EqERA, 993 IP
Detroit Tigers: 4.24 EqERA, 968 IP
Oakland Athletics: 4.31 EqERA, 956 IP
Cleveland Indians: 4.39 EqERA, 977 IP
Toronto Blue Jays: 4.40 EqERA, 921 IP
Minnesota Twins: 4.43 EqERA, 944 IP
Tampa Bay Devil Rays: 4.44 EqERA, 897 IP

The starting rotation analysis starts to indicate why it is that the AL totally pwn3d the NL last season – it’s the pitching, stupid! Whereas the NL lineups stacked up with the best of the AL, the pitching differential is huge, with nine of 14 AL teams above average. Oakland and Cleveland appear here due to superior depth rather than star talent, while the Blue Jays are dragged above average due almost entirely to Roy Halladay and A.J. Burnett. Tampa Bay puts up good rate stats, but have the second lowest projected IP total in the league – if those projections are correct, that is going to be a taxing burden on a weak bullpen.

Below Average
Chicago White Sox: 4.51 EqERA, 980 IP
Baltimore Orioles: 4.53 EqERA, 927 IP
Texas Rangers: 4.56 EqERA, 898 IP
Seattle Mariners: 4.63 EqERA, 942 IP

The White Sox have fallen a long way from the championship season in 2005, as every one of their remaining starters has regressed. However, the White Sox starting rotation retains an uncanny ability to eat innings. My fearless prediction: a strong start out of the gate (while other teams are increasing their SP workload), followed by another late-season fade. Book it. The Mariners have spent some money this offseason, but they still look like a last place team to me.

Do you even have to ask?
Kansas City Royals: 4.87 EqERA, 839 IP

KC has lost 100 games in three of the last four seasons. That means top draft picks each of those years. They spent money in the free agent pool this offseason. The result: by far the league’s worst starting rotation, to go with the league’s worst everyday lineup. So, they’ve got that going for them.

Incorporating Innings Pitched
As a last step, I incorporate an innings pitched credit or debit. Teams with more starter IP get a credit, teams with less a debit. This produces a linearized ranking of AL teams that is intended to represent true rotation value. The ranking is as follows:

1. Boston Red Sox
2. Anaheim Angels
3. New York Yankees
4. Detroit Tigers
5. Cleveland Indians
6. Oakland Athletics
7. Chicago White Sox
8. Minnesota Twins
9. Toronto Blue Jays
10. Seattle Mariners
11. Baltimore Orioles
12. Tampa Bay Devil Rays
13. Texas Rangers
14. Kansas City Royals

This list is mostly unchanged, with the exception of Tampa Bay and Texas going down a number of spots, and Chicago improving.

Friday, January 26, 2007

MLB Preview 2007 – Team Fielding

In the last post, I looked at team fielding in the AL West by doing a position-by-position breakdown of each team and developing a consensus ranking for each player. I have modified the process slightly for this post. First, while I am still creating consensus defensive rankings, I am using a straight formula now. The formula is as follows:

50% Chone Smith’s Zone Rating
20% Last Year’s Zone Rating
20% PECOTA Defensive Projection
10% Tangotiger’s Fan Report

Each player is given a consensus defensive ranking in terms of runs above or below average for his position. These rankings are then aggregated as in the last post to give a team defensive ranking. Each team defensive ranking is expressed in terms of runs per 150 games

You’re the Best….Around! These teams make the Karate Kid proud
Detroit Tigers: +34
St. Louis Cardinals: +34
Minnesota Twins: +29
Oakland Athletics: +21

These four teams outpace the rest of the league by a large margin (though, given the 9 run difference between the top 3 and Oakland, it’s debatable whether Oakland really deserves to be in the conversation). Continuing my opinion that fielding matters way more than most people think, notice that all four of these teams made the playoffs last year, and that the top two both made it to a little tournament known as the World Series. Getting a top fielding ranking basically requires average or better at every position. The only projected starter on any of these teams with significantly below-average fielding ability is Gary Sheffield.

Arizona Diamondbacks: +13
Pittsburgh Pirates: +11
Chicago White Sox: +9
Washington Nationals: +7
San Francisco Giants: +6
Baltimore Orioles: +6
San Diego Padres: +5

A few comments: I really like what Arizona is putting together. Last year they were about average defensively, but with all the young talent coming up, they should be very good to excellent. Young fielders often have a difficult transition to majors, however, so we’ll see if the D-backs can actually capitalize on their projection. I was somewhat surprised by the high ranking of the Giants (driven largely by Omar Vizquel and Pedro Feliz????) and the low ranking of the White Sox (who are hurt by Konerko and Pierzkkaaannsjha whatever). The Padres have got some good stuff shaping up. They are slightly above average both batting and fielding, and the combo makes them one of the top NL lineups.

Seattle Mariners: +2
Toronto Blue Jays: +2
Philadelphia Phillies: +1
Anaheim Angels: +1
Colorado Rockies: -1
Cincinnati Reds: -2
Houston Astros: -2
Kansas City Royals: -3
New York Mets: -3
Texas Rangers: -5

These teams will be neither helped nor hurt by their defenses. Cincinnati’s ranking assumes that the reported Griffey to right field goes through. If he remains in center, go ahead and put them in the negative category. Mad props to the Royals for improving all the way to average after being one of the worst fielding teams in the league for the last few years.

Softball League
Los Angeles Dodgers: -6
Tampa Bay Devil Rays: -6
Cleveland Indians: -6
New York Yankees: -8
Atlanta Braves: -9
Boston Red Sox: -9
Milwaukee Brewers: -10
Chicago Cubs: -14
Florida Marlins: -19

Although I’ve ranked all these teams in one group, they really go from meh (the Dodgers and Braves), to really atrocious (Marlins and Cubs). The Cubs are a trendy pick to win the NL Central this year after spending $300 million this offseason, but I think their team fielding is going to be a huge Achilles’ heel. The Red Sox are an interesting situation in that they are projected to be well above average in the infield, and absolutely terrible in the outfield. The big question is whether Coco Crisp is an average defender in CF. If he is, the Red Sox will sniff average. Like the Cubs, the Brewers are picking up some buzz as a potential NL Central contender, but based on fielding, I just don’t see it. You’ve got to have a really strong offense to overcome such a bad defense, and the Brewers offense is in no way strong. Thumbs down for a Brewers break-out.

Final Note:
You may have noticed that the numbers above don’t sum to zero. I think teams in general have a Plan A lineup that includes good defenders. All good plans eventually go awry which means that some of these players will underperform and others will get injured and be replaced by less talented fielders.

Friday, January 19, 2007

MLB Preview – AL West Team Fielding

Continuing with the aggregated projections, today we’ll look at the fielding ability of each team’s starting eight. Player fielding prediction, despite recent advancements, is still more an art than science. For some current attempts at quantifying defense, see:

The Fans Scouting Report

Zone Rating converted to a run value
John Dewan’s Fielding Bible
Baseball Prospectus Fielding Runs Above Average (FRAA)

By looking at combinations of these metrics, some research of my own, and a dollop of scouting opinion/insight, we should be able to discern the level of defensive performance of teams around the league. Team fielding will be measured as a number of runs above or below average, with positive numbers indicative of better team fielding. To give you an intuitive feel for the numbers, an individual who is 5 runs above average per year is good defensively (say 1 standard deviation above average), and 10 runs represents excellence (2 SDs, or about 2.5% of the MLB population). At a team level, 1 SD is more like 12-15 runs, and it is rare to see extremes beyond +/-30 runs unless driven by an extreme catcher rating (see more on catchers below).

As a first pass, PECOTA projects a FRAA value for every player at his primary position. By aggregating the statistics as in the earlier hitting projections (and assigning a -5 run per year rating to any player not playing their primary position), the projected AL defenses are:

PECOTA Projections
Anaheim Angels: 15 RPY
Oakland Athletics: 7 RPY
Texas Rangers: 4 RPY
Seattle Mariners: -3 RPY

The benefit of using the PECOTA projections is that they are just sitting here. The downside is that FRAA is not based on play-by-play data, instead trying to infer defensive ability from putouts, etc. The results are sometimes a little screwy. It also seems to me that PECOTA projections for catchers are especially wacky, since a catcher’s measurable defensive contribution is largely in controlling the running game. If I remove the catcher’s contribution from above, I get:

PECOTA Projections, No Catchers
Anaheim Angels: 6 RPY
Oakland Athletics: 5 RPY
Texas Rangers: -5 RPY
Seattle Mariners: -5 RPY

As you can see, removing the catcher changes the rankings considerably. PECOTA projects Mike Napoli as +8 RPY defensively, which would put him solidly in the upper quartile of catchers. It seems to me that PECOTA is relying too much on translated minor league statistics to make the projection. Based on his good (but limited) MLB fielding performance thus far, I would say he is more like +2 runs (with a big uncertainty).

The next iteration is to incorporate some other information. We’ll use three sources: Chone Smith’s Zone Rating Projections, my own catcher projections based solely on expected CS%, and Tangotiger’s Fan Scouting Report.


PlayerPosZRrunsFan Score
Casey Kotchman1B258
Howie Kendrick2BN/A52 (at 1B)
Orlando CabreraSS067
Chone Figgins3B-256
Juan RiveraLF060
Garrett AndersonLF-339
Gary Matthews Jr.CF-179
Vladimir GuerreroRF-346
Mike NapoliC149

Looking at the rankings above and combining them with the PECOTA projections shows that there is really only disagreement on Howie Kendrick and Gary Matthews Jr. In the case of Matthews, PECOTA and Zone Rating agree that he is a slightly below average defender in CF. Additionally, the Fan Score prior to this year was significantly less positive than in 2006, leading me to believe that the fan score is being artificially raised by the amazing over-the-wall grab made in 2006. Howie Kendrick is difficult to decipher. His FRAA over the past years is excellent, leading PECOTA to project +14 runs per year. However, his Zone Rating at first base last year was quite average, and he was not considered a top defensive prospect in the minors. Combine that with the low fan score this year, and I am inclined to rate Kendrick fairly low, say +2. Incorporating all publicly available info, the consensus would be something like:

Angels Consensus Rankings

Casey Kotchman1B3
Howie Kendrick2B0
Orlando CabreraSS0
Chone Figgins3B-2
Juan RiveraLF0
Gary Matthews Jr.CF-1
Vladimir GuerreroRF-3
Mike NapoliC2


PlayerPosZRrunsFan Score
Nick Swisher1B-256 (as LF)
Mark Ellis2B1174
Bobby CrosbySS969
Eric Chavez3B785
Bobby KieltyLF239
Mark KotsayCF-682
Milton BradleyRF469
Jason KendallC-258

There are big discrepancies with Mark Ellis, Bobby Crosby and Mark Kotsay. Ellis seems to be an error with PECOTA, as stats, scouts, fans, and even FRAA prior to this year agree that Ellis is quite good defensively. PECOTA may be seeing a decline in performance at age 30. Splitting the difference between PECOTA (+3) and Zone Rating seems reasonable. Bobby Crosby’s PECOTA (-3) seems to be out of line with his previous performance and perception as a plus defender, but his weighted ZR is giving too much credit for 2004 and 2005 when he was legitimately excellent. Once again, a midrange guess seems reasonable. The numbers all agree on Kotsay, who may be getting by on reputation (or perhaps there is a trend and CFs all get rated highly in the Fan Report). The consensus:

A's Consensus Rankings

Nick Swisher1B0
Mark Ellis2B6
Bobby CrosbySS4
Eric Chavez3B7
Bobby KieltyLF0
Mark KotsayCF-4
Milton BradleyRF3
Jason KendallC0

PlayerPosZRrunsFan Score
Mark Teixeira1B375
Ian Kinsler2B-360
Michael YoungSS-366
Hank Blalock3B-461
Brad WilkersonLF236
Kenny LoftonCF-330
Frank CatalanottoRF-642
Gerald LairdC668

Reasonable agreement across the board. The final configuration of the Rangers outfield is still undecided. Catalanotto may be used largely at DH, leaving the RF duties to Nelson Cruz (good defensively), or Sammy Sosa (not so much). The consensus:

Ranger's Consensus Rankings

Mark Teixeira1B5
Ian Kinsler2B-3
Michael YoungSS-1
Hank Blalock3B-1
Brad WilkersonLF0
Kenny LoftonCF-3
Frank CatalanottoRF-4
Gerald LairdC8

PlayerPosZRrunsFan Score
Richie Sexson1B-637
Jose Lopez2B256
Yuniesky BetancourtSS-782
Adrian Beltre3B1078
Raul IbanezLF236
Ichiro SuzukiCF-891 (as RF)
Jose GuillenRF3N/A
Kenji JohjimaC249

Lots of areas of discrepancy here. PECOTA grades Sexson as positive defensively, but this doesn’t match with fan perception or the PBP data. The consensus is Sexson = BAD. Jose Lopez, by contrast, grades out badly by PECOTA. Last year’s Zone Rating and Fan Report say average-ish. PECOTA strikes out again. Betancourt is somewhat of an enigma: the fans love his defense, the PBP metrics hate it, and PECOTA says average. Mike Emeigh at Baseball Think Factory has suggested that young defenders often have a difficult first year when transitioning to the majors, but subsequently recover. Betancourt was very well regarded in the minors, and the numbers support that. I would hedge the bets and say Betancourt will be average next year, but his upside is extremely good. Finally we consider Ichiro! The numbers say he has been lackluster in RF the past few years, and thus predict a very bad performance when moving to the more demanding role of CF. However, the scouting reports on Ichiro are universally positive, and I have seen comments saying that his positioning in RF is unique, leading to strange numbers. This is one instance in which I would ignore the numbers. Ichiro will be an average CF, and may end up above average. The consensus:

Mariner's Consensus Rankings

Richie Sexson1B-5
Jose Lopez2B0
Yuniesky BetancourtSS0
Adrian Beltre3B8
Raul IbanezLF-1
Ichiro SuzukiCF0
Jose GuillenRF1
Kenji JohjimaC0
TOTAL+3(with TUP)

The final scorecard, then is one good fielding team (Oakland), 2 average teams (Seattle and Anaheim), and one bad fielding team with a good catcher (Texas).

Thursday, January 18, 2007

MLB 2007 Preview - NL Lineups

Same rules as yesterday, this time for the NL...

  • The Run Per Game projections given below are only for the starting 9. I do not consider the effect of bench (this will be considered in later posts)
  • Ignoring platoon situations. If there is an expected platoon situation (e.g. Cleveland LF will likely be split between Jason Michaels and David Dellucci), I used the player likely to contribute the most at-bats
  • Projected player BA/OBP/SLG are projected from the PECOTA projection system.
  • We use the PECOTA projections for EqBA/EqOBP/EqSLG to remove the effect of park when comparing teams
  • The NL pitcher is projected to hit 0.200/0.200/0.200 (BA/OBP/SLG), which is around the average for the position over the past few years
  • Team BA/OBP/SLG are estimated as a weighted average of the BA/OBP/SLG of each lineup slot, weighted by the percentage of plate appearances that lineup slot typically gets.
  • Team RPG are estimated from a linear regression formula RPG = -6.09+1.36*BA+18.89*OBP+9.75*SLG+5.96E-05*SBA where SBA is stolen base attempts. So, there's no fancy Monte Carlo engine here, just a plain vanilla estimate of the strength of each team's current offensive performance.
  • Projected Lineups are estimated as of 1/16/07.
Onto the results...

Cream of the Crop

New York Mets: 5.23 RPG (848 Runs)

The Mets are the toast of the NL offensively, and they can certainly be considered on par with the Big 3 of the AL. Replacing the pitcher in our projection (0.200/0.200/0.200 BA/OBP/SLG) line with a replacement level DH (0.250/0.300/0.400), we increase the RPG estimate by 0.42 runs, bringing the Mets to 5.67 RPG. The Mets are strong up the middle, with superstar level production from Carlos Beltran and Jose Reyes. The thing that catapults the Mets to the elite, however, is a third superstar-level producer at 3B in David Wright. In fact, Wright is projected to have the 3rd highest VORP in baseball next year (behind Albert Pujols and Miguel Cabrera). New York also has top prospect Lastings Milledge available, and he just happens to play the same position (RF) as the offense’s only weak link: Shawn Green. Replacing Green with Milledge increases New York’s scoring estimate to 5.30 RPG. Milledge will also be available to DH in interleague play, likely giving the Mets the strongest DH of any of the NL teams.

Second Tier
Atlanta Braves: 5.10 RPG (826 Runs)
San Diego Padres: 5.04 RPG (817 Runs)
Philadelphia Phillies: 5.03 RPG (815 Runs)
St. Louis Cardinals: 5.01 RPG (812 Runs)
Florida Marlins: 5.00 RPG (811 Runs)

The NL’s second grouping clearly separates themselves from the Mets above, and the mediocre below. The Braves come out on top of the second tier thanks to strong expected production up the middle (Brian McCann is second only to Joe Mauer for hitting catchers, Edgar Renteria, and Andruw Jones), complemented with good to excellent production from Adam LaRoche, Jeff Francouer, and Chipper Jones. Like the Mets, the Braves also have some upside potential with strong hitting Kelly Johnson perhaps taking over at 2B, and top catching prospect Jarrod Saltalamacchia forcing his way into the lineup. By contrast, the remaining teams seem to me to all have tremendous downside potential (or TDP in the phraseology of Bill Simmons). Little in the way of depth will hurt all of these teams if time is lost due to injury. San Diego and its balanced attack are probably best suited to deal with the loss of a starting player. The Cardinals continue to live and die on the production of Pujols, Rolen, and Edmonds, while the Marlins are relying heavily on Miguel Cabrera to carry the load.

San Francisco Giants: 4.89 RPG (793 Runs)
Chicago Cubs: 4.88 RPG (790 Runs)
Arizona Diamondbacks: 4.82 RPG (780 Runs)
Milwaukee Brewers: 4.77 RPG (772 Runs)
Colorado Rockies: 4.72 RPG (765 Runs)
Cincinnati Reds: 4.69 RPG (761 Runs)
Houston Astros: 4.68 RPG (758 Runs)

What to say about these teams? Give them an extra 0.5 runs to account for the DH difference, and most of the teams slot in with the rest of the mediocre AL offenses. A couple of notes: San Fran’s projection is based on the return and continued performance of Barry Bonds. Replace him with Todd Linden (his current back-up), and that projection drops to 4.62 RPG. From the rest of this group, I think the strength of D. Lee, Aramis Ramirez, and Soriano may provide an uptick in performance to the rest of the Chicago Cubs, and I think the Diamondbacks are poised to be a sleeper. The D-Backs have a lot of young talent coming up from the minors (and unlike Tampa Bay, a longer track-record of success), but currently lack an established superstar. The Brewers and Reds both have depth around the diamond, although both teams have had injury problems with starters in the past. The Astros could substantially improve their numbers by dumping Biggio or Ausmus, moving Chris Burke back to his natural position at 2B, and promoting Hunter Pence. In essence, come the All-Star break, I may end up wishing I had switched the Second Tier and Meh lists.

Los Angeles Dodgers: 4.59 RPG (743 Runs)
Washington Nationals: 4.55 RPG (738 Runs)
Pittsburgh Pirates: 4.51 RPG (731 Runs)

Calling these teams bad is a little unfair, as after adjusting for the DH effect they are projected to score as frequently as many of the AL offenses. The Dodgers have the best chance of beating their projection, as they have three prospects that PECOTA pegs as superior to the current starters – Matt Kemp in the OF, Andy LaRoche at 3B, and James Loney at 1B. I am somewhat surprised by the Nationals low ranking, given the abilities of Ryan Zimmerman, Nick Johnson, and Austin Kearns. However, the “strong-up-the-middle” meme continues to be quite useful. The grouping of Brian Schneider, Felipe Lopez, Christian Guzman, and Nook Logan contains maybe one league average player. Maybe. The Pirates up the middle: Ronny Paulino (not bad actually!), Jose Castillo (second worst everyday player in the MLB), Jack Wilson (like scrappy David Eckstein without the talent), and Chris Duffy (meh).

It appears at first glance that the overall quality of offensive talent may be superior in the NL this year; however, given the well-documented superiority of the AL, this quality differential may be an illusion. I’ll study this in a future post.

NOTE: Late word that the Braves have traded Adam LaRoche to the Pirates for Mike Gonzalez and minor-league SS Brent Lillibridge. Adam LaRoche’s replacement at this point is likely to be Scott Thorman, a sizable downgrade. The Braves projection with Thorman at first is 4.99 RPG, still in the Second Tier, but now on the low side. The Braves may go with Kelly Johnson at 1B and Lillibridge himself at 2B, and PECOTA is totally crushing on both those dudes. Throw that configuration in and you get a 5.14 RPG estimate. I’m skeptical. For the Pirates, the addition of LaRoche raises their projection to 4.59 RPG…or only 2nd worst in the NL.

Wednesday, January 17, 2007

MLB 2007 Preview - AL Lineups

First post here at "Solvent is so Yesterday", it is not about implicit solvent in any way. Instead, I'll be starting my 2007 Major League Baseball Preview with a look at the AL Starting Lineups as presently constructed. The rankings provided below are subject to the following conditions:

  • The Run Per Game projections given below are only for the starting 9. I do not consider the effect of bench (this will be considered in later posts)
  • Ignoring platoon situations. If there is an expected platoon situation (e.g. Cleveland LF will likely be split between Jason Michaels and David Dellucci), I used the player likely to contribute the most at-bats
  • Projected player BA/OBP/SLG are projected from the PECOTA projection system.
  • We use the PECOTA projections for EqBA/EqOBP/EqSLG to remove the effect of park when comparing teams
  • The NL pitcher is projected to hit 0.200/0.200/0.200 (BA/OBP/SLG), which is around the average for the position over the past few years
  • Team BA/OBP/SLG are estimated as a weighted average of the BA/OBP/SLG of each lineup slot, weighted by the percentage of plate appearances that lineup slot typically gets.
  • Team RPG are estimated from a linear regression formula RPG = -6.09+1.36*BA+18.89*OBP+9.75*SLG+5.96E-05*SBA where SBA is stolen base attempts. So, there's no fancy Monte Carlo engine here, just a plain vanilla estimate of the strength of each team's current offensive performance.
  • Projected Lineups are estimated as of 1/16/07.
Onto the results...

The Cream of the Crop - aka "The Big Three"

New York Yankees: 5.98 RPG (969 Runs)
Boston Red Sox: 5.74 RPG (930 Runs)
Cleveland Indians: 5.69 RPG (922 Runs)

The starting lineups for these three teams project as far superior to the rest of the AL. Not coincidentally, all three teams have elected to use the DH position as an offense first position, rather than as a rotating rest day for the position players (the trade-off being that the defenses of the Big 3 are projected to be quite poor). The Big 3 are very unlikely to achieve their lofty projections, as injury is inevitable, and the injury replacements are unlikely to provide a reasonable facsimile of the original production.

By the way, the natural question is: "How does one become a cream of the crop team?". You've got to get superstar-level offensive performance from one player (Giambi, Manny, Ortiz, and Hafner), and combine that with solid production throughout the rest of the lineup. One interesting point, however, is that while the Yankees and Indians have followed the traditional route to offensive dominance of strength up the middle (i.e. strong offensive production from C,2B,SS, and CF), the Red Sox strength comes from LF and DH. Although Varitek, Pedroia, Lugo and Crisp are projected to provide adequate position-adjusted production, they do not drive the Red Sox offense.

I would speculate that the Red Sox are in a better position (offensively) to weather the inevitable injuries of the upcoming year because they are not dependent on their up-the-middle players to drive the offense. Alex Cora is available as a decent replacement for either of the middle infielders, and Wily Mo Pena can get by in center.

Middle of the Pack
Minnesota Twins: 5.25 RPG
Detroit Tigers: 5.22 RPG
Tampa Bay Devil Rays: 5.15 RPG
Texas Rangers: 5.12 RPG
Anaheim Angels: 5.12 RPG
Chicago White Sox: 5.12 RPG
Toronto Blue Jays: 5.10 RPG
Baltimore Orioles: 5.07 RPG
Seattle Mariners: 5.03 RPG
Oakland Athletics: 5.01 RPG

A bit of a misnomer, since all the remaining AL teams save one fall into this group. It appears that, at least amongst the starting position players, there is a fairly level playing field in the AL this year. I'm surprised by the strong showing of the Twins, but I guess I shouldn't be, given their strong up-the-middle personnel (Mauer, Bartlett, Castillo, and Hunter all provide strong production for their position, with Mauer at a superstar level). Amongst the teams listed, I would knock down Tampa Bay (depending on strong performances from a number of rookies) and the Blue Jays (absolutely no depth to speak of). Expect overperformance from the White Sox (because they can so easily upgrade by replacing Podsednik in left) and Anaheim (Howie Kendrick and Casey Kotchman are regarded much more highly than their numerical projections, plus this team always manages to squeeze out a few more runs that predicted).

The Pole Smokers
Kansas City Royals: 4.74 RPG (768 Runs)

Following up on the question I made regarding the Big 3: how can one team project to be so bad? Well, combine
awful production from the corners (Emil Brown? Reggie Sanders? Ryan Shealy? Seriously?) with Angel Berroa (the worst everyday player in baseball), and you have the worst offense in the AL (and likely the worst offense in all of MLB). However, there is room for hope in Alex Gordon (baseball's #1 prospect) and Billy Butler (a future DH to be sure, but for the purposes of offense only, he carries a big stick). If the Royals follow the expected plan of calling up Gordon and moving Teahen to LF, they improve the team projection all the way to 4.87 RPG. Additionally, if the Royals are willing to bite the bullet and replace Berroa with someone - maybe Esteban German can fake it at SS? - and that takes you all the way up to 5.04 RPG, or no longer the worst team in the AL.