Editor's note: Tom Tippett is the designer of the Diamond Mind Baseball computer game. More articles by the author can be found at the Diamond Mind web site.
Are you ready for some baseball? We've been ready for weeks, so rather than wait for opening day, we decided to get an early start on the season.
As we've done every spring since 1998, we projected stats and assigned ratings for more than 1500 players, including hundreds of top prospects, and then carried out a series of computer simulations. In a minute, we'll show you the projected final standings for the 2004 season, but before we do, let's take a few moments to talk about how we got here.
Using major-league and minor-league statistics (provided by STATS Inc. and SportsTicker), we projected the 2004 performance of each player by evaluating his statistics over the past three years. Each stat line was adjusted for the level of offense in the league, the home park (for both major-league and minor-league stats), the effect of facing the DH (for minor-league and AL pitchers), and the competitive level (majors, Japan, Triple-A, Double-A).
For each player, these adjusted stat lines were averaged (with more weight on recent seasons, performances at higher levels, and seasons with lots of playing time), adjusted for age, and projected into the league and park where he will compete in the coming year.
We also projected each player's left/right splits and assigned ratings for skills such as baserunning, throwing, defensive range, and bunting.
After the players had been rated, we assigned them to their new rosters and set up a manager profile for each team. These profiles consist of starting rotations, bullpen assignments, projected starting lineups against left- and right-handed pitchers, and depth charts for each position.
With everything in place, we played out the 2004 season using our Diamond Mind Baseball game. This program simulates every pitch, with the game's computer manager making the decisions about starting pitchers and lineups, game tactics and substitutions. And injuries are very much a part of these games, just as in real life.
Because luck can be a significant factor in any single season, we simulated the coming season 100 times and averaged the results.
Interpreting the results
The effect of averaging these results is to shoot down the middle, neither being overly optimistic nor pessimistic about a team's chances.
There's nothing wrong with being optimistic, of course. After all, this could be the year when everyone who struggled last year bounces back, every youngster who played well in a three-week trial blossoms into a solid everyday player, the team stays healthy, the older players have one good year left in them, the breaks go their way in the close games, and so on.
In other words, if your favorite team didn't show as well in these simulations as you would have hoped or expected, don't get too discouraged. It certainly doesn't mean that your team can't be this year's version of the 2003 Marlins or the 2002 Angels. It just means they don't look like the favorites on paper right now.
A lot of unexpected things are going to happen in the next seven months, and we make no claims about being able to predict that future. We don't know whether Ken Griffey will give the Reds 150 games of superstar-level hitting, whose pitching ace will go under the knife and miss the season, whether young prospects like Bobby Crosby will live up to their potential right away, which general managers will win the trading deadline sweepstakes, or who's going to going to come through in the clutch.
But that doesn't mean this project is a waste of time, either. It forces us to take a hard look at all the roster moves that have taken place in the past several months, evaluate injury reports, and assess about how each team is going to approach the coming year. That process leaves us with a much better feel for the keys to success for each team, and it fills our heads with things to watch for as the 2004 season unfolds.
Projected 2004 standings
Here are the projected final standings, based on the 100 seasons we simulated on March 17th.
Keep in mind that some things have changed since March 17th. Trades, roster decisions, and new injury reports have altered the landscape. Nothing major, but if we ran the simulations again today, the standings would look slightly different.
W, L, Pct, GB -- average wins, average losses, winning percentage, games behind leader
RF, RA -- average runs for and against
#DIV, #WC -- number of division titles and wild cards (fractions given for ties)
AL East W L Pct GB RF RA #DIV #WC
Boston 102 60 .630 - 890 680 54.5 42.0
New York 101 61 .623 1 899 703 45.5 46.5
Toronto 79 83 .488 23 819 845 1.0
Baltimore 75 87 .463 27 829 872
Tampa Bay 64 98 .395 38 706 874
AL Central W L Pct GB RF RA #DIV #WC
Minnesota 83 79 .512 - 787 765 42.0
Chicago 82 80 .506 1 755 760 34.0
Cleveland 78 84 .481 5 740 756 15.5
Kansas City 78 84 .481 5 812 828 8.5
Detroit 61 101 .377 22 703 917
AL West W L Pct GB RF RA #DIV #WC
Oakland 92 70 .568 - 791 672 56.3 1.5
Seattle 87 75 .537 5 768 720 25.8 2.5
Anaheim 86 76 .531 6 758 700 16.8 6.5
Texas 72 90 .444 20 789 890 1.0
NL East W L Pct GB RF RA #DIV #WC
Philadelphia 97 65 .599 - 855 706 86.5 3.2
Atlanta 86 76 .531 11 822 769 12.0 26.0
Florida 78 84 .481 19 733 752 1.5
Montreal 77 85 .475 20 784 828 1.0 1.0
New York 73 89 .451 24 742 808 .5
NL Central W L Pct GB RF RA #DIV #WC
St. Louis 92 70 .568 - 834 735 45.5 19.3
Chicago 90 72 .556 2 742 655 25.0 22.8
Houston 90 72 .556 2 768 695 29.5 17.2
Cincinnati 75 87 .463 17 719 778
Pittsburgh 67 95 .414 25 698 821
Milwaukee 67 95 .414 25 701 836
NL West W L Pct GB RF RA #DIV #WC
San Francisco 87 75 .537 - 768 707 50.5 3.5
San Diego 83 79 .512 4 771 738 29.5 4.5
Arizona 78 84 .481 9 761 793 8.0 1.0
Los Angeles 77 85 .475 10 657 712 10.0
Colorado 73 89 .451 14 830 916 2.0
The offseason began with the promise of a tighter AL East race as the also-rans in the division groomed some good young talent and/or made some interesting moves. But that was before Boston and New York loaded up on talent (and payroll) and widened the gap between the haves and have-nots. As a result, these two teams reached the postseason in over 90% of our simulated seasons, seriously reducing the chances for an upstart to snag a wild card and make a Marlin-esque October run.
Even so, because the other two division races are much closer, only three AL teams were completely shut out of the postseason in our 100 simulations. Baltimore and Tampa Bay are in the wrong place at the wrong time, while Detroit is in the right place at the right time but has way too far to go after an atrocious 2003 campaign.
Parity or mediocrity? That argument is bound to arise with four teams having a legitimate shot at the top spot in the AL Central. In our simulations, almost 40% of the races were decided by two games or less, with seven ties for first place. Some of those close races weren't all that inspiring, though, as the title was captured eleven times by a team with a winning percentage of .500 or below. Then again, in a third of our simulations, things fell into place for one of the top four teams -- most often Minnesota or Chicago -- and they went on to win 90-plus games and run away with the division.
I'm sure some of you are surprised to see Oakland comfortably atop the AL West again. Anaheim made more headlines during the winter, but Oakland still has all that pitching and quietly improved the offense enough to keep them ahead of their rivals, though they're far from a sure thing in what should be an interesting three-team race. And, yes, we were as surprised as anyone when Texas took one of the 100 division titles. That was an odd season in which the Rangers reached the upper limit of their potential with 83 wins and everyone else faltered just enough for Texas to steal the top spot by a game.
Because the wild card is most likely to come out of the East or West, there's a lot at stake in the AL East race. Chances are the loser of that race will still qualify for the postseason, but their punishment for coming second is most likely to be a date with the Oakland pitching staff. It's quite possible that Boston and New York will clinch playoff spots relatively early, but the promise of a matchup with the AL Central winner should provide plenty of incentive to keep battling down the stretch.
For the second year in a row, 13 of 16 National League teams qualified for postseason play at least once. Except for the possibility that Philadelphia could run away with the East title, much as Atlanta did last summer, we could be in for a real free-for-all.
We'll get into the details in the team comments, but we weren't surprised to see Florida drop back into the middle of the pack in the NL East. They played a little over their heads in 2003 and have since lost some important players. If the Phillies don't self-destruct, the top spot should be theirs for the taking, with Atlanta providing their strongest competition despite their losses to free agency.
We were surprised to see St. Louis on top of the NL Central standings. Everybody's been talking about Chicago's young pitchers and Houston's old pitchers, but the Cardinals still have one of the league's best lineups. It all comes down to whether Tony LaRussa and Dave Duncan can put together a respectable pitching staff around Matt Morris, Woody Williams, and Jason Isringhausen. These results suggest that they might just be able to do that. But their margin for error is very slim, and any of the top three teams could be sitting pretty in six months. Not so for the bottom three teams, all of whom were shut out of the postseason a hundred straight times.
The NL West race could be the most fun of all, with no dominant team and only 14 games separating the top from the bottom when our 100 seasons were averaged. The Padres have a genuine opportunity to go from worst to first as they move into their new stadium. San Francisco still looks like the team to beat, but an average of 87 wins is nothing to get too excited about, especially when so much of their success is tied to one 39-year-old player.
For the second year in a row, the NL wild card race is wide open, with 10 of the league's 16 teams making it to October via that route at least once. More often than not, that playoff spot went to the runner-up in the Central.
Along with my comments about each team, I'll list one or more "difference makers". These are players or other factors that might cause the team to perform better or worse than they did in our simulations. Jermaine Dye is a good example. Because he didn't hit at all in 2003 and was quite ordinary in 2002, we don't project him to be an asset this year. But if he's over his injuries and can return to anything resembling his peak form, that would be a very big plus for the Athletics.
In the following comments, we'll make occasional references to two other articles. Our position rankings article identified the offensive strengths and weaknesses of each team based on the OPS compiled by players at each defensive position during the 2003 season. Our team efficiency article compared the statistical output of each team (offense and pitching/defense) to the runs it was able to generate/prevent and to its win-loss record, pointing out certain teams that did more with less and vice versa.
Boston Red Sox (102-60, division title 55%, wild card 42%)
There are at least a dozen reasons why the Red Sox might not be this good. Pedro Martinez's shoulder might not hold up. Trot Nixon hasn't played this spring because of a slightly herniated disk in his back. Several hitters had career years in 2003 and cannot be expected to do that again. Manny Ramirez and Nomar Garciaparra may be more upset about almost being traded than they're letting on. Byung-Hung Kim's having trouble with his shoulder. The club may not respond well to new manager Terry Francona. They have only one proven lefty (Alan Embree) in a division loaded with good left-handed hitters. And, as Yankee fans would be quick to point out, they are the Red Sox, after all.
That said, there's no denying the amount of talent on this roster. Who wouldn't want a rotation headed up by Pedro and Curt Schilling, who project to finish 1-2 in the league in strikeouts and strikeout-to-walk ratio and are both in the top 10 in just about every projected pitching stat of importance? Other than the lack of a second lefty, the bullpen looks good on paper, with new closer Keith Foulke rounding out a core group of proven veterans.
And the offense, while down from last year, is still projected to be the league's second-best scoring machine behind the Yankees. The Sox scored 961 runs in 2003, and tools like Bill James' Runs Created suggest that they were inefficient, that their statistical output would normally produce over 1000 runs. GM Theo Epstein is on record as saying that he doesn't expect that to happen again, and our simulations support that notion. Boston averaged only (!) 890 runs in our 100 seasons.
But with Schilling and Foulke improving an already-capable pitching staff, and Pokey Reese shoring up the infield defense, the Red Sox look to be among the league leaders in pitching. Based on their raw projections, this staff actually looks better than Oakland's. But that's before we take into account the strength of their schedule, and with 57 games against three of the AL's best offenses (New York, Toronto, and Baltimore), the Sox slipped to second in runs allowed.
The bottom line is that Boston finished our simulations with the second-best offense and the second-best pitching. That's a potent combination that gives them a very good chance to win 100 games and go a Little further in the postseason than they did in 2003.
Difference makers: After we ran the simulations, the Red Sox announced that Trot Nixon will miss the first month of the season ... Byung-Hyun Kim's sore shoulder may turn out to be more serious than it appears ... Embree, if he's injured or off his game and the Sox can't find another lefty setup man ... Nomar Garciaparra, who slumped late in 2003, is off to a slow start this spring, and is battling a case of Achilles tendinitis that now seems like to keep him out of action early in the season.
New York Yankees (101-61, division title 46%, wild card 47%)
Although the Yankees finished slightly behind the Red Sox in our simulations, this might as well be a dead heat. In fact, we've been running simulations periodically throughout the winter, and in early March we saw these two teams finish with identical records -- an average of 100.2 wins each and run margins of +200.6 for Boston and +200.0 for New York -- in one set of 150 seasons.
In any one season, there's likely to be a lot more separation than that, but we could be in for one of the great battles of all time. (Imagine how much more exciting it would be if the loser didn't have the wild card to fall back on. On the other hand, the wild card system gave us last year's ALCS, and there's a lot to be said for that, too.)
The trade that brought Alex Rodriguez for Alfonso Soriano didn't improve the Yankee offense by as much as some might think. Yes, Rodriguez is a better overall hitter than Soriano, but A-Rod has played his entire career in hitters parks, and after you put both on a park-neutral basis, Soriano's not that far behind.
Furthermore, the move creates an offensive black hole at second base. We went with a combination of Enrique Wilson and Miguel Cairo in our simulations. Both are hitting very well this spring, but neither projects to do that when the games count. Having said that, there's still time for New York to fix the 2B problem, and if they do, the Yankees would move ahead of the Red Sox.
Difference makers: Jon Lieber, if he's unable to make the 30 starts he averaged in our seasons ... Jeter, if he recognizes that the team is better with A-Rod playing short and volunteers to change positions ... Hideki Matsui, if 2003 turns out to be an adjustment year and he begins to hit for more of the power he showed in Japan ... Gary Sheffield, if his thumb injury doesn't allow him to play 150 games ... Jose Contreras, who averaged 14 wins in our simulations but has only 9 big-league starts under his belt.
Toronto Blue Jays (79-83, wild card 1%)
Although Toronto projects to be 23 games behind the leaders, I won't be surprised if the gap is smaller when the real 2004 season is played. Even with a ton of talent, it's not easy to win 100 games, and chances are something will go wrong with at least one of the front runners. In fact, both of those camps have seen their share of worrisome injuries already this spring. And Toronto has some interesting young talent in a farm system rated the 8th-best in the game by Baseball America, so the Jays may well be a stronger team by August.
Offense shouldn't be a problem, though it'll be difficult for them to match the 894 runs they scored in 2003. In our simulations, they slipped to 819 runs and fourth in the league in scoring. The reasons? Shannon Stewart was dealt to Minnesota last summer. Carlos Delgado and Vernon Wells had monster seasons in 2003 but project to come back a little toward their career averages in 2004. Reed Johnson and Greg Myers are also projected to fall short of their 2003 numbers.
Pitching is the real question mark here. Even with the reigning Cy Young winner, Roy Halladay, Toronto finished 10th in the AL in runs allowed last season. Many of the names have changed, but the staff remained 10th in our simulations. A lot depends on whether Miguel Batista and Pat Hentgen can build on their success in 2003; both dropped back a bit in our simulations because their disappointing 2002 campaigns hurt their projections. But they need a number five starter and the bullpen looks quite thin.
Difference makers: Lefty starter Ted Lilly, who could be very good but was held back for much of the spring due to a wrist injury ... Wells, who's young enough to keep getting better ... manager Carlos Tosca, if he goes ahead with his stated plan to start the light-hitting Kevin Cash at catcher 4-5 times a week ... OF Alexis Rios, if he builds on his strong performance in the Arizona Fall League.
Baltimore Orioles (75-87, no postseason appearances)
With the Albert Belle and Scott Erickson salaries off the books, the Orioles wasted no time in rebuilding their batting order. The addition of Rafael Palmeiro, Javy Lopez, and Miguel Tejada improve a lineup that finished 10th in scoring last summer but added 86 runs and rose to 3rd in our simulations.
The pitching is a very different story. Baltimore dropped to 11th in runs allowed in our hundred seasons, giving up 52 more runs than their 2003 counterparts. Their only significant addition is Sidney Ponson, and he doesn't really count as an upgrade because he made 21 starts for the O's before he was dealt to the Giants. Gone are Hentgen and Kerry Ligtenberg, two of their better hurlers from a year ago. Their average run margin suggests a 77-win season, but they fell two wins short of that mark, perhaps because Jorge Julio projects to be one of the worst closers in the majors.
Difference makers: Rodrigo Lopez, if he pitches more like he did in 2002 (3.57 ERA) than in 2003 (5.82) ... ditto for Julio ... Kurt Ainsworth or Eric DuBose could emerge as solid starters.
Tampa Bay Devil Rays (64-98, no postseason appearances)
Lou Piniella made some very bold statements on the banquet circuit, but his team could improve and still wind up in the basement in a division where the Blue Jays are further along in their development of young talent and the other three clubs are spending heavily on established stars.
Aubrey Huff, Rocco Baldelli, and Carl Crawford are fun to watch. But the lineup is comprised of youngsters who don't walk or lack power (or both), an older Tino Martinez who's best power days are behind him, and a defensive specialist in Rey Sanchez. Only the Tigers scored fewer runs in our simulations.
The Devil Rays have quite a few young arms, but the youngest of them have yet to prove themselves and the ones in their mid-twenties don't appear to have a lot of upside. This system hasn't produced anyone like Mark Prior, Dontrelle Willis, Brandon Webb, or the young guns in Oakland. Maybe that will change in 2004 or 2005, but few of their prospects have the kind of dominant minor-league numbers that suggest immediate success at the highest level. And it's not an easy place to get started because 76 of their games are against the league's four best offensive teams.
Difference makers: It's hard to single anyone out ... any number of young players could develop more quickly than usual, but it would take a bunch of them to make a real difference in baseball's toughest division.
Minnesota Twins (83-79, division title 42%)
Despite losing half its pitching staff, Minnesota wound up on top of the division in our simulations.
Starters Kenny Rogers, Eric Milton, and Rick Reed are gone, and the heart of the bullpen was ripped out when closer Eddie Guardado signed with the Mariners and setup man LaTroy Hawkins landed with the Cubs. But that doesn't mean they'll be a whole lot worse in 2004.
In their place are Joe Nathan, coming off a fine season in San Francisco, and three younger guys who appear to be ready to help. In two minor-league seasons, Jesse Crain has fanned 143 in 112 innings while posting a 1.69 ERA. He could emerge as the new closer. Grant Balfour posted a 2.41 ERA and an excellent strikeout-to-walk ratio of better than 5:1 in Triple-A last year. And Carlos Silva has had enough success in the minors to suggest that he can replace some of the subpar starts the club received from Rick Reed (5.07 ERA) and Joe Mays (6.30) in 2003.
Not much has changed offensively, and that's not a bad thing. A.J. Pierzynski was dealt to the Giants to make room for catching phenom Joe Mauer, a move that is likely to cost the team some runs. But this is a young lineup that figures to get better. Only two players older than 30 (Denny Hocking and Chris Gomez) were given 100 or more atbats last seasons, and both of them are gone. Justin Morneau (age 22) and Mike Restovich (25) have gained some valuable big-league experience. And leadoff hitter Shannon Stewart is on board for a full season.
Difference makers: Unproven kids like Mauer and Crain, who might struggle as they get their first dose of big-league opposition ... the front office, if they choose to deal one of their many outfielders for a pitcher or a middle infielder who can hit.
Chicago White Sox (82-80, division title 34%)
The White Sox should have won the division last year. They outscored their opponents by 76 runs, the best mark in the division and 33 runs better than the Twins. Their offense produced 270 more total bases and walks (TBW) than were conceded by their pitchers and defense. That was also number one in the division and well ahead of Minnesota, who finished with a TBW differential of +113. And they were sitting in first place with less than three weeks remaining before they dropped five straight head-to-head contests with the Twins.
Even though the biggest offseason change was the loss of staff ace Bartolo Colon, the White Sox could actually be better this year. The projections for Paul Konerko and Joe Crede are down because their off-years in 2003 are weighted heavily in our system. Either or both of them could bounce back in a big way. Chicago's center fielders combined to rank only 10th in the league offensively in 2003, and that could change for the better if they get more production out of Aaron Rowand, top prospect Jeremy Reed, or newly-acquired Timo Perez. Reed struggled this spring and was sent down, but he could be back later in the year.
With Colon out of the picture, the league's 3rd-best pitching staff in 2003 dropped to 6th in runs allowed in our simulations. Esteban Loaiza, Mark Buehrle, and Jon Garland head up the rotation, but it's questionable after that. Scott Schoeneweis isn't likely to be the answer. He averaged a 10-11 record and a 4.58 ERA in our simulations, but that may be optimistic, since he's never had a sub-5.00 ERA as a starter and has been pounded this spring. Dan Wright is in the mix, but his career ERA is 5.52. Also contending for a spot is Jason Grilli, a first-round draft pick by the Giants in 1997 who dropped off the radar screen a couple years ago but had some success in the minors last year.
Difference makers: Our simulations were run with Schoeneweis and Wright in the 4-5 slots, so there's plenty of room for improvement if Wright takes a step up or they can find someone better than these two ... we gave the closer job to Damaso Marte because he projects to be much better than Billy Koch, and the record could suffer a little if Koch gets most of the save opportunities ... who knows how the club will respond to new manager Ozzie Guillen? ... Magglio Ordonez, if the club falls out of contention early and he's traded during the year.
Cleveland Indians (78-84, division title 16%)
We could have just as well put Kansas City ahead of the Indians in our projected standings, as the two teams finished in a virtual dead heat in our simulations. Cleveland averaged 79.3 wins, KC 79.4, and both had an average run margin of -16. It's not clear why Cleveland won the division more often when everything else was the same, but that's the way it turned out, so we're listing the Indians third and KC fourth.
The Tribe's offense was pathetic last year, spared the dubious distinction of being last in the AL only by the even-more-awful Tigers. They should be better in 2004, though not by a whole lot. In our simulations, the Indians added 41 runs and passed the Devil Rays to climb all the way up to 12th in the offensive rankings. The difference? More playing time for Victor Martinez, Milton Bradley, Ben Broussard, and Travis Hafner. But the team lacks power and it's far from clear that they'll get any offense out of the middle infield positions.
As was the case in 2003, the pitching looks respectable. CC Sabathia, Jason Davis, Cliff Lee, and Jeff D'Amico seem set, but it remains to be seen whether either Jason Stanford or Chad Durbin is the answer in the #5 spot. The bullpen finished 4th in the league in relief ERA last year and should be a strength.
Difference makers: We ran our simulations with Bob Wickman as the closer, but we've since learned that he'll miss at least the first half ... David Riske gets the job now, and even if he succeeds, it could have a ripple effect as others are moved into more demanding roles and the depth of the pen is tested ... we had Jake Westbrook and Jason Bere in the rotation, but Bere is hurting and they're planning to use Westbrook in relief.
Kansas City Royals (78-84, division title 9%)
Seems odd, doesn't it? The Royals won 83 games last year, brought back most of the key contributors from that squad, added slugger Juan Gonzalez and several mid-level free agents (including Scott Sullivan, Benito Santiago, and Matt Stairs), and yet we seem to be saying they'll most likely drop five games in the standings.
Except that we're not saying that at all. We're saying that they are a better team than they were in 2003. It's just that Kansas City wasn't really an 83-win team to begin with. For one thing, they gave up 31 more runs than they scored, and that run margin normally produces 78 wins, not 83. But it gets worse. That 31-run deficit overstates their underlying statistical performance to similar degree.
The 2003 Royals had baseball's most efficient offense -- fourth in the AL in runs scored, but only sixth in on-base percentage and seventh in slugging. The Runs Created formula says they should have scored about 807 runs, but they pushed across 29 more than that. The difference? Clutch hitting. A team batting average of .300 with runners on base, a full 26 points higher than their overall average. Problem is, this type of clutch performance doesn't tend to be repeated.
We can look at this another way, too. KC's hitters produced 2854 total bases and walks (TBW). Six AL teams produced more bases even though the Royals were aided by one of the better hitting parks in the game. The pitchers and defense allowed 3102 TBW, the 3rd-worst total in the league. Yes, the park made them look worse than they really were in this respect, but the park doesn't affect the differential between offense and defense. And their TBW differential was -248, 4th-worst in the league, a little below Cleveland and Texas and a little above Baltimore and Tampa Bay. That's not good company if you have dreams of playing in October a year later.
Well, enough about last year. Kansas City raised their game a little in our simulations. The offense ranked 5th in runs, and while their park is still giving them a boost (even after moving the fences), clutch hitting had nothing to do with it this time. In any one simulated season, they might repeat their 2003 clutch performance, but when you average the results of 100 seasons, those flukes disappear. The better news came on the other side of the ball, where the Royals shaved 39 runs off their real-life total from 2003.
Difference makers: Miguel Asencio was part of our rotation but will miss the year (Tommy John surgery) ... 22-year-old LHP Jimmy Gobble, who got his first taste of the majors with nine starts last year and didn't embarrass himself ... Zack Greinke, one of the game's best pitching prospects, if he's able to contribute later in the year ... Kevin Appier, who seems to be recovering very quickly after torn-tendon surgery and could make many more starts than we gave him in our simulations ... David DeJesus, an on-base machine in the minors, if he can do the same at the highest level.
Detroit Tigers (61-101, no postseason appearances)
Were the Tigers not coming off a dismal 43-win season, their performance in our simulations would have been considered a disaster, ranking among the most disappointing outcomes in the seven years we've been doing this. It's pretty sad when you can improve by 18 games and still be projected to be the worst team in baseball.
If you're interested in the details, Baseball Prospectus 2004 contains an excellent essay on how the club reached this point and why it's likely to be many years before they can contend again.
I suppose I could try to tell this sordid tale in my own way, but I've already done that a few times in the Diamond Mind blog, and I don't see the need to pile on right now. Instead, I'm going to tell you why there just might be reason to hope for a Motor City miracle.
The 2003 Tigers gave up 669 more total bases and walks than they produced, and that's the 4th-worst deficit in the 30 years for which we have all the stats necessary to compute these numbers. (Doubles and triples allowed by pitchers are not part of the official record.) Let's take a look at what happened to some other teams on this dubious list.
The 1996 Tigers, who topped the list with a -727 mark, rebounded in 1997 to win 79 games and nearly break even on TBW. You'd think that a change of this magnitude would require contributions from every quarter, but in this case it was all pitching. A staff ERA of 6.38 was reduced to a respectable 4.56 in just one season thanks to Justin Thompson (15-11, 3.02 ERA), Willie Blair (16-8, 4.17), Brian Moehler (4.67 ERA in 31 starts), Todd Jones (31 saves, 3.09) and middle relievers Doug Brocail and A.J. Sager.
The 1979 Oakland A's won 54 games and posted the second-worst TBW deficit (-683) in this period before rebounding for 83 wins and a +86 TBW differential in 1980. The offense improved by 289 bases and the pitching/defense by 480 thanks to a highly-successful youth movement featuring a quintet of young starting pitchers (Mike Norris, Matt Keough, Rick Langford, Brian Kingman, and Steve McCatty) and one of the best defensive outfields in the history of the game (Dwayne Murphy, Tony Armas, and Rickey Henderson).
Yeah, I know that the turnaround was short-lived for both of those teams, and this Detroit team hasn't made the wholesale changes that make a quick recovery more plausible. And, no, I'm not predicting that it will happen. I'm just saying that a turnaround of this magnitude is not unprecedented.
Difference makers: Ugueth Urbina, who was signed after we ran the simulations, but who needs time to get his arm ready and won't have many leads to protect anyway.
Oakland Athletics (92-70, division title 56%, wild card 2%)
The win total is down a little, but the percentages of division titles and wild cards are almost identical to those we saw in last year's simulations. As you know, the A's repeated as division winners in the real 2003 season, though not without a strong challenge from the Mariners. Yes, they lost Miguel Tejada and Keith Foulke, but another winter of shrewd moves and the ability to develop young pitchers put them on top in our simulations again.
Despite the loss of Tejada, Oakland's offense is likely to outscore last year's group. It might be asking a lot for Bobby Crosby to fill Tejada's shoes in his first full season in the majors, but Crosby's .939 OPS in AAA and strong spring suggest that the dropoff could be smaller than some people think. In the outfield, Bobby Kielty, Mark Kotsay and Jermaine Dye are upgrades over Terrance Long, Chris Singleton, and ... Jermaine Dye, who batted .172 last year and projects to be much better in 2004. Eric Karros adds a useful right-handed bat at first base, a position where the A's were quite weak in 2003. It's true that new catcher Damian Miller can't replace Ramon Hernandez's bat, but the overall impact of these changes is positive.
The pitching staff is right where you'd expect: at the top. The big three (Tim Hudson, Mark Mulder, Barry Zito) are back, though there are questions about all three. Hudson and Zito were among the league leaders in lowest batting average on balls put in play, and results like that are difficult to repeat. There's some concern about Mulder's health and Zito's declining strikeout rate. But there are precious few teams that wouldn't swap their top three for this group. Rich Harden is available from day one this time. Mark Redman steps in for Ted Lilly. And Arthur Rhodes projects to be a more than adequate replacement for Foulke in the closer role.
Difference makers: Mark Ellis, who was healthy when we ran the simulations but will miss at least a couple of months with a shoulder injury ... Crosby, if he needs time to adjust ... Rhodes, if he's unable to translate his success as a setup man into success as a closer ... top pitching prospect Joe Blanton, who could provide a midseason boost like Hudson, Zito and Harden before him ... Billy Beane, if he can make another of his astute trade-deadline deals to fix any problems that arise in the early going ... Zito, who has adjusted his delivery, if the change is for the better.
Seattle Mariners (87-75, division title 26%, wild card 3%)
This is a meaningful drop from their 93-69 record in the 2003 season, so it's worth taking a moment to see if we can pinpoint the reasons. There's no obvious decline in their list of offseason player moves. Sure, they lost Mike Cameron, Arthur Rhodes, and Kaz Sasaki, but newcomers Raul Ibanez, Eddie Guardado, Rich Aurilia, and Scott Spiezio should compensate in a variety of ways.
Roughly three of those wins disappeared because the Mariners averaged 768 runs in our simulations, down 27 from their 2003 total. This seems reasonable to me. Last year's offense was unusually efficient. Because the core veterans are getting up in years, they are projected to decline a little. Aurilia doesn't figure to be any better than Carlos Guillen was last year. Ibanez is not an upgrade on Cameron. At third, Spiezio should improve on Jeff Cirillo, but Spiezio's not projected to hit all that well in this park. And their division rivals have upgraded their pitching. None of these factors is a big deal by itself, but they add up to a small drop in scoring.
The pitching staff led the league with only 637 runs allowed in 2003 but yielded an average of 720 in our simulated seasons. To make a case for a repeat of their league-leading status, you'd have to answer 'yes' to several important questions. Can a 41-year-old Jamie Moyer go 21-7, 3.27 again? Can they expect their top five starters to make every single start, as they did last year? Can Shigetoshi Hasegawa and Rafael Soriano post ERAs in the vicinity of 1.50 two years in a row? Even if Guardado can fill Sasaki's shoes, who's going to replace Rhodes? Can Julio Mateo and Ryan Franklin get through another season with in-play batting averages of .234 and .243, respectively? How much will the staff be hurt by the loss of Cameron, the sport's best defensive center fielder?
Don't get me wrong. This is still a good team. They just don't appear to be quite as good as they were last year.
Difference makers: Freddy Garcia, if the correction of his inner ear problems helps him get back to his 2001 level ... any of the older veterans, if they suffer a sudden decline ... top prospects Clint Nageotte and Travis Blackley, who might be ready to pitch in before season's end.
Anaheim Angels (86-76, division title 17%, wild card 7%)
Two years after the Angels won it all, many are predicting a return to the top of their division in 2004. That could happen, but it doesn't appear to be the most likely outcome. Anaheim was an average team last year -- 11th in scoring, 5th in fewest runs allowed, a run margin of -7, and a TBW differential of -73. To catch Oakland, assuming the A's play to their potential, the Angels will need to improve on their 77-85 record in 2003 by 15 games or more. Have they done enough this winter to make that happen?
Adding Vladimir Guerrero is a very good start, no doubt about that. But Anaheim's hitters ranked in the bottom four in the league in OPS at four positions last year -- shortstop, catcher, center field, and designated hitter. David Eckstein and Bengie Molina are still at short and catcher, respectively. Their "solution" to the center field problem is to move Darin Erstad to first base, an offense-oriented position where he could well be the league's worst hitter. The new DH, 35-year-old Tim Salmon, projects to be a league-average hitter at a position where much more is expected.
The pitching staff does appear to be much improved, even with Erstad's stellar CF defense out of the picture. In our simulations, the Angels trailed only Oakland and Boston in fewest runs allowed. Newcomers Bartolo Colon and Kelvim Escobar should be major upgrades over what Kevin Appier and Aaron Sele gave them in 2003. Each of the returning starters (Jarrod Washburn, John Lackey, and Ramon Ortiz) is projected to improve a little on his 2003 performance. And the bullpen continues to be a major asset.
Difference makers: Troy Glaus, if he can rebound from his injury and his second consecutive subpar season ... Scot Shields, who would make a better #5 starter than either Ortiz or Lackey but who appears to be headed for the bullpen again ... management, if they figure out that Erstad belongs in center ... if they don't want to wait for Casey Kotchman to be ready for the 1B job, they could try moving Jose Guillen for a first sacker who can hit.
Texas Rangers (72-90, division title 1%)
These results reflect a 37-run decline on offense and a 79-run improvement in runs allowed compared with the 2003 Rangers, who (with an assist from a home park that favors hitters) were 5th in scoring and dead last in pitching.
When you lose big bats like Alex Rodriguez, Rafael Palmeiro, Juan Gonzalez, and Carl Everett, you don't expect to score as many runs. But some of the new guys can play, too. Alfonso Soriano isn't A-Rod, but he's not too far behind after you adjust for the difference in parks. New DH Brad Fullmer projects to an OPS over .900 in this environment, and that's more than they got from Palmeiro last year. And Mark Teixeira and Hank Blalock haven't gone anywhere. Overall, it's an average lineup that should finish a little above average in runs thanks to the friendly confines of The Ballpark in Arlington.
The ERAs of the 16 men who started at least one game for the 2003 Rangers were 4.85, 5.09, 5.49, 6.10, 6.23, 6.45, 6.85, 7.01, 7.11, 7.16, 7.30, 7.58, 8.35, 8.53, 11.40, and 12.00. John Thomson, the best of this bunch, is now in Atlanta. The guys who were north of 7.00 amassed a total of 61 starts, so this isn't just a handful of cup-of-coffee September starts that make the overall picture look worse than it really was. You could hardly do worse if you dumped them all and started over with replacement-level pitchers.
In a way, our simulations may prove to be on the optimistic side. We chose the best five starters -- newcomer Kenny Rogers, Joaquin Benoit, Ricardo Rodriguez, R.A. Dickey, and Juan Dominguez -- based on our projections. All of them compiled ERAs in the 5's in our simulations, and while that's not saying much, it's a heck of a lot better than last year's crew.
But Buck Showalter doesn't get to play the season 100 times, and he doesn't get to rely on the fact that everyone will perform roughly at their projected level when you average that many seasons. So one big question is whether Showalter -- who gets only one shot at the real 2004 season and cannot know in advance who's going to perform well -- bets on the right guys.
Difference makers: Juan Dominguez was in our rotation but has since been sent down ... Colby Lewis and Glendon Rusch are in the mix for a rotation spot, and they would drag the results down if we reran the simulations with either or both playing a major role.
Philadelphia Phillies (97-65, division title 87%, wild card 3%)
Well, we picked the Phillies to win the division last year, and we all know that didn't happen. The lineup finished 5th in scoring, but that was a bit of a letdown. When Pat Burrell and David Bell never got going, a potentially great offense was reduced to a good one. The pitching was projected to be quite good, and it was. The combination should have been enough for the wild card, as Philly was considerably better than Florida in both run margin and TBW differential.
Not much has changed on the pitching and defense front. Starter Brandon Duckworth is gone, but Eric Milton should fill his shoes quite nicely. Billy Wagner is a major upgrade over Jose Mesa, who posted a 6.52 ERA but somehow managed to save 24 games in 28 chances anyway. The middle relievers should be okay. All in all, this shapes up to be one of the league's better staffs.
The offense features the same cast of characters, too, so any improvement will have to come from within. The most likely candidates for that are Burrell, Bell, and Jim Thome, who was very, very good but has been even better in the past. If these guys do show some improvement, don't be surprised if the Phillies lead the league in scoring.
Difference makers: Manager Larry Bowa, who hasn't yet shown that he can win even when he has the most talent in the division ... Burrell, who could be comeback player of the year if he bounces all the way back to his 2002 level ... Milton, whose underlying numbers have been better than his ERAs for several years, if he thrives in a more pitcher-friendly league and a park that may turn out to be good for pitchers.
Atlanta Braves (86-76, division title 12%, wild card 26%)
Gary Sheffield is gone. Javy Lopez is gone. Greg Maddux is gone. But their run may not be over.
2B Marcus Giles and SS Rafael Furcal were first and third in OPS at their respective positions, and they're still there. Chipper and Andruw Jones are still there. J.D. Drew, if he can stay healthy, is a very good right fielder. 1B Adam LaRoche and C Johnny Estrada bring some very good minor-league hitting numbers to the table. Atlanta won't score 900 runs again, but even if they fall all the way back to the 822 they averaged in our simulations, that should be good enough to put them in the top four offensively.
We really like the acquisition of John Thomson, who could very well give the Braves as much as they got from Maddux a year ago. Mike Hampton, Russ Ortiz, and Horacio Ramirez round out a decent front four. The #5 spot is a big question mark, with Paul Byrd working his way back from Tommy John surgery, but the recent acquisition of Juan Cruz could help there. In the bullpen, some of the names have changed, but other than superstar closer John Smoltz, the relief corps is nothing special.
Difference makers: Drew, who has the potential to put up Sheffield-like numbers, though he hasn't done it for a full season yet ... LaRoche, Estrada, and new 3B Mark DeRosa, who must match or exceed their modest projections if the Braves are going to contend ... perhaps Mike Hampton, if he makes it all the way back to his 1998-2000 form now that he's one more year removed from Coors Field ... Cruz and Chris Reitsma were acquired after we ran the simulations, and that narrows the gap between the Phillies and Braves.
Florida Marlins (78-84, wild card 2%)
I guess it's fitting, and maybe even encouraging, that the Marlins reached the postseason in our simulation only via the wild card. After all, they've won two World Series titles but have never finished first in their division. In 2003, Florida made the comeback an art form, climbing out of a deep hole during the regular season, rallying from the brink of elimination against the Giants and the Cubs to reach the Series, then falling behind the Yankees before capping off their third straight series victory with three wins in a row.
There are reasons to be optimistic -- full seasons from Miguel Cabrera, Dontrelle Willis, and Mike Lowell; the potential return of A.J. Burnett from Tommy John surgery; and the hope that Josh Beckett can deliver 200+ innings at the level he demonstrated in the postseason (43 innings, 21 hits, 2.11 ERA).
But can they replace Ivan Rodriguez, Derrek Lee, Mark Redman, and Juan Encarnacion? A full season from Jeff Conine fills the bill for the Encarnacion slot. If Burnett returns early enough and pitches well, the Redman spot is covered. The other two positions are dicier. At first base, Hee Seop Choi has a promising future but doesn't project to match Lee's output right away, and they're losing Lee's defense, too. Behind the plate, Ramon Castro and Mike Redmond will try to replace a Hall of Famer, and that won't be easy.
Their performance in the simulations may turn out to be a little on the pessimistic side. A good number of their key players have had their ups and downs, and because the downs carry some weight in the projections, some of them took a step back from their 2003 levels in our 100 seasons. That could happen, of course, and if it does, you're looking at a pretty average team. If 2003 is a better indicator of their talent, the Marlins could make another serious run.
Difference makers: Juan Pierre, if his injured pinky turns out to be serious, though the latest reports indicate that he'll be ok ... Ramon Castro's legal problems could negatively affect his availability and/or performance ... Burnett, depending on when he returns and how quickly he gets back to his peak level.
Montreal/San Juan Expos (77-85, division title 1%, wild card 1%)
When people talk about the Expos, they usually focus on the losses of Javier Vazquez and Vladimir Guerrero and the seemingly endless search for a permanent home base. (Not that there's anything wrong with Montreal as a baseball town, by the way, if only someone would make a commitment to winning and marketing the team.)
Amid all of that negativity, it's easy to lose sight of the positives. Montreal was in the middle of the pack in staff ERA last year even though they played 3/4 of their home games in a very good hitters park (Olympic Stadium) and the other 1/4 in a bandbox (Hiram Bithorn Stadium) that routinely turned guys like Jeff DaVanon into Babe Ruth. Yes, they did lose Vazquez, but they still have solid starters in Livan Hernandez, Zach Day, Tomo Ohka, and Claudio Vargas. The bullpen was nothing special last year, but it wasn't a liability, though it looks a little worse this year.
The addition of four players should help Montreal score quite a few more runs even without Guerrero. The 2003 Expos ranked near the bottom in OPS at five positions, and they've addressed three of them with the acquisitions of 1B Nick Johnson, 3B Tony Batista, and OF Carl Everett. If prospect Terrmel Sledge emerges as a capable corner outfielder, as he's projected to do, that makes four. And C Brian Schneider could bounce back after a disappointing 2003 season.
Difference makers: We made Luis Ayala the closer, but if manager Frank Robinson stays with Rocky Biddle, it could cost them a few games ... Tony Armas, coming off shoulder surgery, was in our starting rotation for the full season but had a setback and will miss at least a few starts.
New York Mets (73-89, division title 1%)
73 wins is better than the 66 racked up by the Mets last summer, but not good enough when you're in a major market and you're competing for attention with a Goliath on the other side of town.
The Mets will score more runs. Jeromy Burnitz will be missed, but he's the only 2003 contributor who won't be back. They should get more atbats from Mike Piazza and Cliff Floyd. And newcomers Mike Cameron and Kaz Matsui will help, too. Even so, New York's offense was below the league average in scoring during our simulations. Combine that with an aging, below-average pitching staff, and a front office that made only a couple of bold moves during the winter, and it's hard to find the seeds of a miracle.
Difference makers: Piazza, who could reclaim his spot as the sport's best hitting catcher or become a below-average first baseman at the plate and in the field ... Matsui, simply because nobody really knows how well he'll do on this side of the Pacific ... strong defense up the middle could help the pitchers a lot, perhaps enough to overcome some weaknesses at the corner positions.
St. Louis Cardinals (92-70, division title 46%, wild card 19%)
Albert Pujols, Jim Edmonds, Scott Rolen, and Edgar Renteria are back, so you know they'll score a lot of runs as long as they don't fill the rest of the lineup with out-making machines. Could they really undermine the heart of the order that way?
Probably not, especially with a proven hitter like Reggie Sanders in right field. But St. Louis may get very little out of the catcher (Mike Matheny) and second base (Marlon Anderson, Bo Hart) positions. And after they decide whether to play Pujols at first or in left, it's not clear how much they'll get out of the other position. We put Pujols at first and split the LF time between Ray Lankford and Colin Porter, but that's highly speculative at the moment. With that arrangement, this lineup averaged 834 runs per simulated season, second only to Philly's 855.
It almost goes without saying that the real issue is the pitching. That's what derailed a promising 2003 season, and the Cards must cut their runs allowed and hold more late-inning leads. The good news for St. Louis fans is that they did just that in our simulations.
Eleven relievers not named Jason Isringhausen combined to blow 28 saves last season. Overall, with Isringhausen included, the 2003 bullpen coughed up about 12 more leads than the average team. (Remember, they lost the division by only three games.) But that problem was solved last June when Jason Isringhausen returned from injury last June to notch 22 saves in 25 chances. He's healthy now, so closing games shouldn't be an issue in 2004. Overall, with a full year from Isringhausen, the bullpen could be adequate.
What about the starting rotation? Led by Matt Morris and Woody Williams, augmented by journeyman Jeff Suppan, and rounded out by some combination of Danny Haren, Chris Carpenter, Jason Marquis and Jason Simontacchi, the rotation could be good enough to get the job done. We ran the simulations with Haren and Carpenter in the #4 and #5 spots, but you could substitute the other guys without affecting the team's performance very much.
Difference makers: All teams are vulnerable to injuries, but with so much value tied up in the top four hitters, the top two starters, and the closer, the loss of a couple of key guys could be devastating ... Haren, if the 23-year-old is given a chance and emerges as a solid starter.
Chicago Cubs (90-72, division title 25%, wild card 23%)
According to our simulations, this is the best pitching staff in the National League. Among the hurlers who won't be back, only relievers Mark Guthrie and (to a lesser extent) Dave Veres will be missed. Newcomers Greg Maddux and LaTroy Hawkins will more than make up for those losses, especially when you consider that the Maddux starts replace those of Shawn Estes and his 5.73 ERA. With these changes, a staff that finished 3rd in the ERA rankings last year should shave another 30 runs off their total.
Ah, but you have to swing the bats, too, and that's why the Cubs aren't prohibitive favorites to win this division. Derrek Lee is the only significant offseason addition to a lineup that was outscored by eight NL teams last year. A full season from Aramis Ramirez at third base will help. But it's hard to see where any other gains will be made, and in our 100 seasons, Chicago added only 18 runs to their 2003 total.
Difference makers: If Mark Prior's Achilles problem causes him to miss a few starts, that could be enough to drop Chicago behind Houston ... Maddux, if his pitching expertise rubs off on his new teammates and they soar to even greater heights as a staff ... Corey Patterson, if he continues to develop as quickly as he did last year, or even if he merely consolidates his game at the new level he established.
Houston Astros (90-72, division title 30%, wild card 17%)
The Astros and the Cubs, who finished in a virtual dead heat for second place in our simulations, are similar in other ways, too:
both have very good pitching staffs and mediocre lineups, though Houston's hitter-friendly park makes them look a little less similar than they really are
each team hired a future Hall of Famer pitcher this winter -- Roger Clemens in Houston, Greg Maddux in Chicago
both feature a trio of young starting pitchers -- Houston's are Roy Oswalt, Wade Miller, and Tim Redding, while the Cubs have Mark Prior, Kerry Wood, and Carlos Zambrano
a trio of aging hitters is key for both lineups -- Jeff Bagwell, Jeff Kent, and Craig Biggio for the Astros; Sammy Sosa, Moises Alou, and Mark Grudzielanek for Chicago
Difference makers: The new closer, Octavio Dotel, who is projected to thrive in that role on the basis of three terrific seasons as a setup man but is unproven in that role.
Cincinnati Reds (75-87, no postseason appearances)
The 2003 season is remembered for a surprisingly weak offense, a very poor pitching staff, and turmoil surrounding the midseason dismantling of the front office and the team. A potentially great outfield was reduced to mediocrity when Adam Dunn couldn't hit for average and Ken Griffey and Austin Kearns lost big chunks of their seasons to injury. Apart from 3B Aaron Boone, the catchers and infielders ranged from below-average to terrible.
Fueled mainly by a presumption of better health in the outfield, the Reds scored 25 more runs in our simulations than they did in the real 2003 season. There was plenty of additional room for improvement, but the front office did nothing to fill the other holes in the lineup.
The good news is that the pitching staff was mostly blown up and reassembled. Cincinnati used 17 different starters last year, and Paul Wilson is the only one who appears likely to get at least 10 starts in both 2003 and 2004. Aaron Harang, Juan Acevedo, Corey Lidle, and Brandon Claussen filled out the rotation for our simulations. Danny Graves has been moved back to the closer role, and that along with a full season from Ryan Wagner should give them a respectable pen. All of these changes added up to a savings of over 100 runs.
In case you're wondering why the Reds improved by more than 125 runs in our simulations but gained only six games in the standings, the 2003 Reds were the NL's worst team, statistically speaking. They allowed 517 more total bases and walks than they produced, and only a spate of clutch hitting allowed them to win 69 games.
Difference makers: Griffey, if he's traded or if he stays and returns to the top echelon of the sport ... any infielder who can provide some offense and who is given the opportunity to do so ... Jimmy Haynes beat out Claussen for a rotation spot but is projected for a higher ERA than Claussen.
Pittsburgh Pirates (67-95, no postseason appearances)
After trading away many of their better players during the 2003 season, the Pirates made no significant effort to reload over the winter. Their farm system is almost ready to infuse the big club with some good young arms (such as John VanBenschoten, Sean Burnett, and Bryan Bullington), but that's not enough to make them a threat in 2003. The front office seems quite willing to lose while they wait for the kids to develop.
Difference makers: Kris Benson, who needs to show that he can come all the way back from his 2001 elbow surgery, assuming he's not traded in the meantime ... Ryan Vogelsong, a once-promising but injury-riddled pitching prospect who is having a very good spring.
Milwaukee Brewers (67-95, no postseason appearances)
The Pirates and Brewers were extremely close in our simulations, but we listed Pittsburgh fifth on the basis of a very small edge in run margin. But the real season could easily go the other way. Baseball America says the Brewers have the #1 farm system in the game, and if the benefits of that system begin to accrue this summer, the Brewers could escape the basement.
Difference makers: 2B prospect Ricky Weeks, who brings some highly-touted hitting credentials but is questionable in the field ... we ran our simulations with top prospect J.J. Hardy getting most of the playing time at SS, but Bill Hall (who's having a great spring) and Craig Counsell appear to have the inside track ... Scott Podsednik, whose strong 2003 season was much better than anyone projected based on his minor-league numbers, if he turns out to be a one-year wonder.
San Francisco Giants (87-75, division title 51%, wild card 4%)
Suppose, just for the sake of argument, age finally catches up with Barry Bonds and he loses 250 points off his OPS. Amazingly, that would still leave him at the level of Todd Helton, Albert Pujols and Manny Ramirez. Not bad company for a 39-year-old, is it? But even though he'd still be one of the best hitters in the game, it would be enough to cost the Giants six wins unless somebody else steps up to share the load.
Offensively, the 2004 Giants are very similar to their real-life 2003 counterparts. In our simulations, Barry averaged .339 with 53 homers, 134 RBI, 130 runs, and 162 walks, 40 of them intentional. Ray Durham and A.J. Pierzynski are assets, but other than that, it's a lineup full of guys who are league-average hitters or worse.
The pitching staff is good but not great. In raw numbers, the Giants may challenge for the league lead in ERA again this season, but that's with the help of one of the best pitcher's parks in the league. Take the park into consideration, and none of the starters after Jason Schmidt are particularly impressive, though 22-year-old Jerome Williams has some upside. The bullpen, a strength in recent years, should be improved by the return of closer Robb Nen but must find a way to replace Joe Nathan and Tim Worrell.
Thanks mainly to Bonds, Schmidt, and Nen, the Giants are once again favored to win the West, but their margin for error continues to shrink even with the division in a bit of a down cycle.
Difference makers: Bonds, if he stops being superhuman (and, yes, I do write this every year) ... Schmidt, who had shoulder surgery after the 2003 season and likely won't be ready to start the year ... Nen, who won't be ready for opening day, if he's not back at full strength soon.
San Diego Padres (83-79, division title 30%, wild card 5%)
Of all the 2003 cellar dwellers, the Padres have the best chance to go from worst to first, mainly because they made themselves a lot better and partly because none of their rivals improved in the past six months.
Two of their better hitters from 2003, Rondell White and Mark Kotsay, won't be back. But that's not a huge loss, as neither reached the .800 mark in OPS. In their place are Brian Giles, who was acquired in August, and Jay Payton, a free agent signing. Ryan Klesko and Phil Nevin could play a lot more in 2004 after missing time with injuries last season, though Nevin is already banged up. Ramon Hernandez, acquired in the Kotsay deal, is a badly-needed upgrade at catcher. All of a sudden, you've got the makings of a pretty darn good lineup, one that was slightly above average in scoring in our simulations despite playing in a new stadium (Petco Park) that we believe will favor pitchers.
A similar rebound is needed from a pitching staff that was very bad in 2003. Their old playground, Qualcomm Stadium, was the league's best pitcher's park for the past two years, decreasing runs by 18% in that span, and somehow the Padres managed to finish 13th in team ERA. Sixty starts were made by hurlers who finished the year with an ERA over 5.00. With the arrival of David Wells, Sterling Hitchcock, and Ismael Valdez, none of those 5.00-plus guys played a part in our simulations. Jake Peavy, Adam Eaton, and Brian Lawrence head up the rotation.
Two relievers, one well known and the other a newcomer, improve a bullpen that struggled right along with the rotation last year. Ace closer Trevor Hoffman, who tossed only 9 innings in 2003, is back in his accustomed role. That pushes Rod Beck, a very pleasant surprise last season, into a setup role. The newcomer is Akinori Otsuka, whose Japanese stats were so incredible that he projects to be a real asset for the Padres even with a healthy discount for the higher level of competition.
Difference makers: Lawrence is having a terrible spring, allowing 48 baserunners in 19 innings amid reports of a serious drop in velocity, and that spells trouble if it carries over into the season ... Beck left camp for personal reasons and nobody seems to know when, or even whether, he'll return ... Khalil Greene, if his strong spring is a sign that he's about to top our projection for him by a big margin ... Nevin, if his ability to play defense is limited by attempts to guard against another shoulder injury.
Arizona Diamondbacks (78-84, division title 8%, wild card 1%)
This represents a six-game slide from their 2003 finish, and that's mainly because they gave up a lot more runs in the simulations. The departure of Curt Schilling and (to a lesser extent) Miguel Batista are the most obvious reasons for the change, but there's more to the story. According to our projections, several of their younger pitchers won't be able to match the 2-something ERAs they put up in 2003. I'm referring to starter Brandon Webb and relievers Jose Valverde, Oscar Villareal, and Mike Koplove.
After Randy Johnson and Webb, the rotation is a huge question mark. Elmer Dessens posted a 5.07 ERA last year and is projected for about the same in 2004. Steve Sparks and Shane Reynolds are in line for the remaining two spots, but both are projected for ERAs closer to six than five.
Offensively, the Diamondbacks were very disappointing in 2003, finishing 10th in scoring even with a big assist from a hitter-friendly home park. This year's lineup has two very good hitters in Richie Sexson and Luis Gonzalez, an above-average hitter in Steve Finley, a bunch of guys who are average at best, and a fairly thin bench. The result was a run total that was right in the middle of the pack in our simulations.
Difference makers: Randy Johnson, if he returns to the form that won him those Cy Youngs ... we ran the simulations with John Patterson in the rotation because he projected to be better than Sparks and Reynolds, but he's been traded to Montreal.
Los Angeles Dodgers (77-85, division title 10%)
Last March, I used this space to praise the Dodger pitching and argue that the bottom half of the order would use up so many outs that they just wouldn't score enough runs to contend. In a way, I was wrong on both counts. The pitching wasn't just good, it was awesome. And the offense was worse than I ever imagined.
LA's pitchers held opponents to 70% of the league average rate in scoring, a rate surpassed by only four teams since 1901 (the Cubs of 1905/1906/1909 and the 1939 Yankees). LA's hitters scored runs at 73% of the league average rate, a record of futility exceeded by only five teams since 1901.
An average offense with that pitching staff would have done a lot of damage, so let's see if the Dodgers have made enough moves to become average. Brian Jordan led the club in on-base percentage, and he's gone. Fred McGriff and Jeromy Burnitz combined for 26 homers in 149 games, and both of them are gone. They signed Juan Encarnacion, and he's got a little power, but his career on-base percentage is only .313. The good news is that Shawn Green should be better because he's healthier. This is still a very bad offense, folks, and they finished last in the NL in scoring in our simulations.
That's bad news for a team that won't be able to challenge all-time pitching records this time around. Kevin Brown packed his 32 starts and 2.39 ERA when he climbed on a plane for the trip to New York. Paul Quantrill's 77 innings and 1.75 ERA went with him. And how many of the men who remain in Dodger blue are going to be able to match the career-best numbers they put up last year? Not enough, in our view, and while pitching and defense will continue to be the backbone of this team, it won't be good enough to overcome an anemic offense.
Difference makers: Jeff Weaver, if his troubles were left behind in New York, but who's been hammered this spring ... Nomo, who is struggling this spring after having shoulder surgery ... new GM Paul DePodesta, who can be expected to make some moves to address their deficiencies.
Colorado Rockies (73-89, division title 2%)
I give up. I can't figure out what they're trying to do in Colorado. When GM Dan O'Dowd took over, he talked about how they had this whole playing-at-altitude thing figured out. That their computer models would allow them to project future performance and thereby guide their player moves as they build a perennial winner. Then they spent a ton of money and didn't win. So they turned over half the roster and still didn't win. And all of a sudden O'Dowd is making speeches about how altitude changes everything and makes it exceedingly difficult to succeed.
Well, the Broncos won two Super Bowls not long ago. The Avalanche have won the Stanley Cup. And the Nuggets ... well, they're a bad example, though they're showing some life this year. Why are the Rockies the only Denver team that cries about the devastating effects of playing at altitude? Is this argument just a smoke screen, or is there something unique about baseball that makes the altitude effects so much more powerful?
Two of O'Dowd's arguments are, in fact, unique to baseball. One is that muscles need more than 24 hours to recover in Denver, and that matters more in a sport like baseball that plays almost every day. The other is that the thin air changes the trajectory of breaking pitches, which forces both pitchers and hitters to retrain their brains as they go from home to road and back again. There may be something to those claims. I don't know.
I am pretty sure that this Colorado team isn't very good. But if you're going into the season with 73-win talent, you might as well do it in a division where 85 wins might win the darn thing. If things go your way, overachieving by 12-15 wins is not out of the question.
Difference makers: Larry Walker, whose weight loss and renewed commitment to fitness might make him 2004's version of Javy Lopez, though he'll start the season on the DL with a groin injury ... we ran the simulations with Chin-hui Tsao in the rotation, but Shawn Estes is having a very good spring and has won the job.
Our 2004 simulation results suggest that at least four divisions could produce very close races. Boston and New York are neck and neck, both Central divisions appear to be very tight, and the NL West has no dominant team. Philadelphia and Oakland appear to have the most breathing room, but even they can't take anything for granted. Oakland has two good teams to worry about, and Philly looked just as good in our simulations last year and but couldn't match those results in the real 2003 season.
That's not good news for those of us who publish preseason projections and then take the time to assess their accuracy after the season is over. It wouldn't take a whole lot -- one or two career years, a key injury here or there, a couple of midseason trades -- to make the October 4th standings look very different from those at the top of this article.
But it's great news for baseball fans. If our simulations are any indication, sixteen teams have at least a 10% shot at the postseason. Twenty-four clubs went to the postseason at least once in 100 tries. Half the teams that are projected to finish last are close enough to the leaders to pull off an upset.
In other words, a lot of the day-to-day stuff is going to matter more than usual. When a star goes down for six weeks or three months with an injury, it's likely to have playoff implications. This year's version of Dontrelle Willis or Brandon Webb could make all the difference in a division or wild card race. When the front offices are wheeling and dealing at the trading deadline, a sizable majority could be trying to get better. That's a lot more fun than watching eighteen teams dump salary while a handful of elite clubs try to position themselves for their inevitable October appearances.
Close races are great for the game. Here's hoping we get a bunch of them in 2004.
This article also appears on the Diamond Mind web site.