Baseball Articles | Index

Projecting the 1999 Season

By Tom Tippett
March 15, 1999

We've just completed the process of developing projected 1999 statistics and ratings for over 1400 players and using this information to run a series of computer simulations of the upcoming season. In this article, I'll present our projected final standings for the 1999 season and share my comments on the outlook for all thirty teams. Before I do that, however, I think it's worth taking some time to explain our methodology and talk about what you can and cannot expect to learn from an exercise such as this one.

Anyone in the business of forecasting the weather, the economy or anything else will agree that nobody can predict the future with certainty. Baseball is no different. Every day, things happen that affect the future performance of players and teams. Players change their workout programs, develop new pitches, adjust their swings, decide whether to sit out or play through injuries, and so on. Managers, general managers, and owners alter the competitive balance with trades, roster cuts, and salary-driven personnel decisions. Guys get hurt. Some injured players come back early and others suffer relapses that cost them the entire season.

So we can't tell you who will make the playoffs. But we can provide a baseline against which you can evaluate the news of the day. These predictions provide a snapshot, based on everything we knew as of March 12, the day we ran the computer simulations. This snapshot reflects all off-season roster moves, the best information about the role each player will play on his team, and any players with injuries that will prevent them from being ready on opening day.

To give just one example, our simulations tell us that the Braves are heavy favorites to win the NL East, but that the Mets are much improved and have a legitimate shot to win the division. Hours after we ran these simulations, we learned that Braves closer Kerry Ligtenberg would miss anywhere from a month to the season with a torn ligament in his elbow. Knowing that they were expected to win 102 games before this injury makes it easier to figure where they might end up if he does indeed miss the season.

So, even though events will make these predictions increasingly obsolete as time goes on, it's still useful to put a stake in the ground and say, "here's how the baseball world looks to us at this moment." That's why we think it's worth developing these projections even though the future holds some surprises for all of us. And, of course, we don't just do this to learn something, we do it because it's a lot of fun.

Our 1998 Results

For over twenty years, Pete Palmer (co-author of Total Baseball and The Hidden Game of Baseball) has been tracking published predictions of team standings. He was gracious enough to share his 1998 data with us so we could see how our 1998 projections compared to those of the fourteen national publications that printed their predicted final standings last spring.

His scoring method is a simple one -- subtract each team's actual placement from their projected placement, square this difference, and add them up for all the teams. For example, if you predict a team will finish fourth, and they finish second, that's a difference of two places. Square the result, and you get four points. Do this for every team and you get a total score. The lower the score, the more accurate your predictions.

Here are the final standings, using Pete's method, for 1998:

Diamond Mind            44.5

Sports Illustrated      54

Sporting News           54

Baseball Digest         58

Spring Training         58

Baseball Weekly         60

Street & Smith          64

ESPN                    64

Sport                   64

SI Baseball '98         64

Inside Sports           72

Athlon                  72

Baseball Illustrated    74

Ultimate Sports         78

Mazeroski               88

This isn't the only scoring system one could use to rank these projections, of course, so it's quite possible the standings would change if a different method was used. By the way, the Diamond Mind score is not a round number because we projected that Texas and Anaheim would tie for second, and we used 2.5 for the projected place for both teams.

Naturally, we were very pleased when our projections ranked so highly. But it's just one year, and only time will tell whether we simply got lucky. There's only one way to find out, and that's for us to make it an ongoing task to develop our projections in the spring, analyze the results after the season, identify ways to improve the methodology, and try to get better the next year.

The Methodology

Anyone who follows baseball can make predictions like this, and many writers do just that every spring. The best of them have the experience and mental acuity to translate everything they know -- past performance, off-season player movement, current injuries, prospect reports -- into a good forecast. Our approach is to take all of this information and put it into a very systematic process that weighs all of these factors in an objective fashion.

We started by putting together 1999 team rosters that reflect all of the off-season player moves. Each team includes the guys most likely to appear on the opening day rosters and the others most likely to be called up in the course of the season. The rosters include top prospects, even if they have never played in the majors.

Using major-league and minor-league statistics for the past three years, we projected the 1999 performance for over 1400 players. Each stat line was adjusted for the effect of the league (deflating Pacific Coast League stats, for example), 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, AAA, AA). For each player, the adjusted stat lines were then averaged (with more weight placed on recent seasons, performances at higher levels, and seasons with lots of playing time), adjusted for age, and then projected into the league and park where he will compete this year. We also projected each player's left/right splits based on their splits for the past three seasons and assigned ratings for other skills such as baserunning, throwing, defensive range, and bunting. A more description of our approach appears on our web site.

We then put together a manager profile for each team based on the best available information as of March 12th. Our software uses a manager profile to determine the roles each player is to play in a simulated season. This profile consists of the starting rotation, spot starters (to replace injured starting pitchers), bullpen assignments (closer, setup, long relief, mopup), starting lineups versus left- and right-handed pitching, platoons, defensive replacements, and utility roles.

We set up the manager profiles to reduce the projected playing time for any player who is currently injured and expected to miss part of the season. In addition, all players (even Albert Belle, the reigning iron man) can get hurt during the simulated season.

Using the actual 1999 schedule, we played out the season using our Diamond Mind Baseball game. This program simulates every pitch, with a computer manager making all decisions about starting pitchers and lineups, game tactics (e.g. bunt, steal, hit&run, swing away, pickoff throws, pitchouts, baserunner advancements) and substitutions (pinch hitting, pinch running, relief pitching, defensive subs, injury replacements) using the information in the manager profiles.

We ran the season twenty times and averaged the results. Why? Because in any one season, the results can be affected by luck -- which teams suffer more injuries, which teams get the breaks in the close games, and so on. By averaging the results from twenty seasons, the luck evens out for the most part.

Assumptions . . . or Timing Is Everything

I've already mentioned that things are going to happen that make our projections obsolete to some degree, but it bears repeating. If we had run our simulations a month ago, we'd still have Roger Clemens on the Toronto roster, and Andres Galarraga and Moises Alou would be projected to play 150 games each. If we had run them two weeks ago, we wouldn't have known that Delino Deshields and Tony Womack will be out for about six weeks each with broken bones. And if we had run them on March 13th instead of March 12th, we would have known that Kerry Ligtenberg's torn elbow ligament might cost him at least a portion of the 1999 season, and quite possibly all of it.

Furthermore, there are things that cannot be known until some future date. How healthy will Kerry Wood be this year? Will Anaheim trade a hitter for a starting pitcher before the season? Will Kansas City trade Kevin Appier, and if so, where will he end up? Will Mark Wohlers get his game back this year? If so, will it happen from opening day, or will it take a few weeks? If it takes a few weeks, will the club give up on him before it happens? Which of the young prospects will fulfill their promise right away (as did Garciaparra, Rolen, Grieve, Wood and Helton) and which will need more time to develop (like Karim Garcia and Paul Konerko)?

Because these things are unknowable, we have no choice but to make assumptions about them. And because some of these assumptions can have a significant impact on the results of the simulations, I'll mention the most important ones in the team-by-team comments below. That way, you'll have more information with which to evaluate the simulation results.

The Results

Here are the projected final standings, based on the twenty seasons we simulated:

Legend
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 wildcards (.5 given for ties)

AL East         W   L   Pct   GB    RF    RA   #DIV   #WC

New York      104  58  .642    -   934   694   20.0 

Baltimore      83  79  .512   21   805   823          5.0

Boston         80  82  .494   24   829   848          2.0

Toronto        79  83  .488   25   823   839          2.0

Tampa Bay      75  87  .463   29   731   821          3.0



AL Central      W   L   Pct   GB    RF    RA   #DIV   #WC

Cleveland      94  68  .580    -   907   756   19.0    .5

Kansas City    78  84  .481   16   784   805    1.0   2.0

Detroit        76  86  .469   18   775   820

Chicago        66  96  .407   28   756   904

Minnesota      63  99  .389   31   723   911



AL West         W   L   Pct   GB    RF    RA   #DIV   #WC

Texas          89  73  .549    -   894   807   12.5   2.5

Seattle        86  76  .531    3   870   818    5.0   2.5

Anaheim        81  81  .500    8   828   829    1.5    .5

Oakland        77  85  .475   12   856   873    1.0  



NL East         W   L   Pct   GB    RF    RA   #DIV   #WC

Atlanta       102  60  .630    -   791   587   18.0   2.0

New York       93  69  .574    9   767   660    2.0  12.0

Philadelphia   76  86  .469   26   704   748

Montreal       68  94  .420   34   687   812

Florida        64  98  .395   38   608   766



NL Central      W   L   Pct   GB    RF    RA   #DIV   #WC

Houston        92  70  .568    -   824   732   10.0   1.0

St. Louis      90  72  .556    2   845   753    6.5   1.5

Cincinnati     84  78  .519    8   749   720    3.0   1.5

Chicago        80  82  .494   12   773   790     .5   

Milwaukee      73  89  .451   19   709   800

Pittsburgh     72  90  .444   20   679   761



NL West         W   L   Pct   GB    RF    RA   #DIV   #WC

Los Angeles    92  70  .568    -   780   667   14.5   1.0

Arizona        83  79  .512    9   745   724    4.5   1.0

San Francisco  81  81  .500   11   777   756         

Colorado       80  82  .494   12   888   899    1.0

San Diego      69  93  .426   23   685   799

Post-Season Chances

In the American League, eleven of the fourteen teams qualified for the playoffs in at least one of the simulated seasons, with only Detroit, Chicago and Minnesota getting shut out. I'll talk about this more in the team comments, but I think these results overstate Kansas City's chances at the expense of the Tigers. If KC is to do this well, several young players must come through, and Kevin Appier must be healthy, pitch the way he did before the injury, and stay with the Royals for the entire season. They've said they'll trade Appier for salary reasons, meaning they want to deal him for prospects rather than equivalent major-league talent, and if they do, Detroit should inherit their post-season chances.

The AL wild card race appears to be wide open. We have nine teams projected to win between 75 and 86 games. That means that any one of them could grab the wildcard if things fall their way, and in the course of our twenty simulated seasons, eight of them did qualify for the post-season at least once. And the ninth team, Detroit, has as good a chance as several of the others and would have made it in had we run more seasons.

Over in the National League, there appears to be a little less parity. Only nine of the sixteen teams made the playoffs, mainly because there are two very good teams in the East, meaning that the loser of that race is odds-on favorite for the wildcard. As in the AL, one team with a legitimate shot at a wild card spot didn't happen to win one. It would not be fair to say that San Francisco has no chance because they failed to make the post-season in twenty straight simulations. Clearly, they are a contender for both the division title and the wild card. It just didn't happen for them this time around.

The Luck Factor

I'm sure you noticed that KC actually won the division over Cleveland in one of the twenty seasons, despite an average that was 17 wins behind the Indians. How can this be? Is it really possible for a 17-game underdog to finish first in a 162-game season? Yes, provided all or most of the five types of luck -- the good season factor, the injury rate, offensive efficiency, defensive efficiency, and run distribution -- go in the underdog's favor.

When teams go to spring training, they have certain expectations for what every player will contribute. But while many players do indeed have typical seasons, some are up and some are down. For most teams, the ups and downs cancel out and the team performs about as expected. But in any given season, the team might be lucky enough to have more than its share of good seasons, as the Yankees did last year, and they might win more games than expected. And another team might suffer through an unusual number of bad years. This is what I call the good season factor.

The potential impact of injuries is obvious. Every few years, some team has its starting rotation stay healthy all year, enabling their top five starters to make at least 155 of a possible 162 starts. And we've all seen situations (think Phillies) where the entire starting rotation goes down in the span of a few months. If the underdog has one of those low-injury seasons and the favorite gets hit pretty hard, it can make for a big swing in the standings.

The third factor is offensive efficiency. Over the course of the season, a team's offense will produce a certain number of walks, hits, homeruns, stolen bases, double plays, and so on. These events combine to produce runs, and Bill James quantified the relationship between the underlying events and the number of runs they produce in his Runs Created formula. This formula has proven to be quite accurate over the years, and we can use it to identify teams whose offenses produced more or fewer runs than normal during any given season. Last year, for example, Seattle's offense produced 64 fewer runs than the Runs Created formula suggested they would, and that was one of the reasons for their disappointing season.

The flip side of offensive efficiency is defensive efficiency, and by that I mean the ability to yield fewer runs than normal given the number of offensive events created by the opposing team. Last year, Houston was a good example of a team helped by this factor, as they allowed about 60 fewer runs than would normally be produced by the hits and walks and homers their pitchers allowed. If this is a sign of a bunch of pitchers who are especially good in the clutch, they might do this again next year. Chances are, however, they were lucky last year and things will revert to normal in 1999.

Finally, there's run distribution. Sometimes the runs you score come just at the right time and you have a great record in close games. Sometimes it seems that when your pitchers are on you don't hit, and when your attack is in full swing your hurlers get hammered, and you end up losing a bunch of games 2-1 and 9-8. This effect can be measured by another Bill James formula, the Pythagorean Projection, that computes the expected winning percentage for a team that scores and allows a certain number of runs. The win totals for most teams are right in line with these projections, but sometimes a full season isn't enough time for these things to even out. In 1997, the Giants were outscored by their opponents but still managed to win 90 games and a division title, while Baltimore scored 32 more runs than they allowed but still wound up four games under .500.

So, yes, it's possible for a weaker team to finish ahead of a stronger team in a 162-game season if most or all of these five factors go their way. There are lots of examples of this from baseball history, but one of the better ones is the 1984 Cubs. In 1983, they were not a good team (71-91, 5th place). In 1984, they won the division despite being outdone by the Phillies in team ERA, batting average, on-base percentage, slugging average, stolen bases and other statistical categories. In 1985, they finished below .500 again. I believe their 1984 campaign was a great example of one of those charmed seasons when a mediocre team rises to the top because a lot of things go right. You can attribute their success to clutch performances, a commitment to team goals, or other noble attributes of that group. But I will remain convinced that, for whatever reason, the most talented team did not win in 1984. And that could happen again this year. And it doesn't surprise me that it did happen in one of our twenty simulated seasons.

Team Comments

AL East

New York Yankees (104-58, 20 division titles)

Even though the Yankees are projected to have the best record in baseball and were the only team to win its division all twenty times, I feel the need to explain why their projection isn't better than it is. After all, they won 114 games last year, brought back all their starters except Tim Raines, and then added Roger Clemens to the league's best rotation. How can they be ten games worse than last year?

Well, age is a factor, as quite a few of their key players are on the wrong side of 30. Last year, they had a very good ratio of career years to bad seasons, and that's not likely to be repeated. And there's a tendency for all teams to move toward the middle of the pack each year, as the weaker clubs try to get better and the stronger ones stand pat. Last August, I wrote about the many similarities between this team and the great 1939 Yankees club. That team stood pat and proceeded to fall 22 games in the standings and finish third.

But let's focus on the positive. In our twenty seasons, Roger Clemens averaged a 20-7 record and a 2.98 ERA, and his best season was 24-6, 2.33. We assumed that Andy Pettitte would be healthy enough to take his regular turn in the rotation, and he averaged 18-7, 3.53. If he's not at full strength, it could hurt the team, though with a pitcher like Mendoza in the wings, it shouldn't hurt much. Overall, this team is projected to lead the league in staff ERA by a good margin.

Like last year, the Yankees will field a well-balanced offense that should once again lead the league in scoring. It's hard to know who'll play LF for them, but we gave most of the playing time to Ricky Ledee, the lefty hitter in a platoon with Chad Curtis. But the Yankees have enough good options at this position that it shouldn't affect the results if they go in a different direction.

Baltimore Orioles (83-79, 5 wildcards)

I'm not convinced that they've improved their team as much as many are saying. Yes, the addition of Albert Belle will help the offense, but they lost Eric Davis's .327 average and 28 homers. Yes, they signed Delino DeShields, but he's not as good as Roberto Alomar. At first base, they gave up Rafael Palmeiro's .296 average and 43 homers and replaced it with Will Clark's .305 average and 23 homers. And Palmeiro is a better fielder. The rotation could be better if Mike Mussina stays healthy all year, Scott Kamieniecki comes back, and Juan Guzman has one of his better seasons, but I think they're really going to miss Armando Benitez in the bullpen. And they're still one of the older teams around.

In our simulations, Belle averaged .296 with 39 homers and 116 RBI, figures that are right in line with his career averages. I've heard quite a few people say that Belle could challenge the homerun record playing in Camden Yard, but that park has only increased homers by 7% for right-handed batters over the past three years, and I think they're making way too much of this.

There is upside potential in Mike Mussina. Because our projections use data from the past three years, his poor 1996 season is still dragging him down a little. He averaged 16-11 with a 3.93 ERA in our simulations, but I think there's a very good chance he will be better than that.

Boston Red Sox (80-82, 2 wildcards)

Last year, we pegged the Red Sox as the favorite to win the AL wildcard and they came through with the second best record in the league. This year, with the loss of Mo Vaughn and the failure to sign any of the marquee free agents, many have written this team off altogether. Don't forget that this club outscored it's opponents by 147 runs last year. Even if the loss of Vaughn costs them 50 runs offensively, they would still have a substantially positive run margin that would put them into the thick of the wildcard race.

So why are they pegged to win only 80 games? After Pedro Martinez, the starting pitchers are old and have a history of injuries. It's very unlikely that Tom Gordon, even if he's 100% healthy and at the top of his game, will be able to go a full season with only one blown save as he did last year. And some of their hitters are coming off above-average seasons that are unlikely to be repeated.

Despite these factors, there is some reason for optimism. A year ago, the Red Sox signed Mark Lemke in a last-minute attempt to find someone to be their everyday second basemen after Jeff Frye went down with a season-ending knee injury. How things change. This year, they have Jose Offerman, Frye, Lou Merloni and Donnie Sadler battling for that spot, and if Frye is able to return to form, they could be in a position to trade a 2B for help in other places. And Ramon Martinez could come back at full strength for the second half of the season.

On the other hand, Bret Saberhagen and Mark Portugal could find that their oft-repaired arms won't hold up, and their rotation could be in big trouble.

Toronto Blue Jays (79-83, 2 wildcards)

David Wells had a terrific season last year. But many forget that his ERAs in 1996-7 were 5.14 and 4.21 and that he was struggling early in 1998 before turning things around. He turns 36 in May and won't be playing in front of the league's best defense this year. Add this up and you get a forecast of a 15-11 record and a 3.75 ERA. Good enough to make him one of the best lefty starters in the league, but not enough to replace what Clemens did the past two years.

The big questions concern their pitching. Will Pat Hentgen return to his Cy Young winning form? Will Chris Carpenter take another step forward this year? Can Kelvim Escobar again be the dominant starter he was down the stretch last season? How bad will the bullpen be with Mike Timlin, Randy Myers, and Paul Spoljaric having been traded in the past 18 months and Paul Quantrill is out with a broken leg for a while?

Offensively, this team was eighth in the league in runs last year, and that was before Mike Stanley and Jose Canseco left town. The additions of Willie Greene and Homer Bush don't appear to be enough to increase their ranking, even if their good young players continue to develop. By the way, we've got Bush pencilled in as the starting 2B, but it could just as well be Grebeck. Either way, the outcome should be about the same.

Tampa Bay Devil Rays (75-87, 3 wildcards)

Thanks to the addition of Jose Canseco, the development of some young players, and a bounce from several veterans who had off-years in 1998, Tampa Bay looks to improve by about 12 games over their record last year. But I think their three wildcard wins overstate their chances, as they have the lowest projected win total of any of the wildcard contenders. Nevertheless, they could sneak in if things go right.

Their starting rotation is solid through the first three positions, with Rolando Arrojo, Wilson Alvarez, and Tony Saunders. And they have a lot of arms in camp that could enable them to patch together the rest of the rotation, much like the Red Sox did last year. The bullpen looks pretty solid. Right now, we're projecting them to have a staff ERA that's in the middle of the pack. But with several teams bunched very closely together, they could easily match their 1998 ranking, when they were fourth in the league.

Offense is their Achilles heel. Even with Canseco, they were second to last in runs scored in our simulations. But it's better than it sounds. In 1998, they were dead last in runs, 94 tallies shy of the next-worst Royals, and 192 short of the league average. This year, they look to be only 91 runs below the league average. This is where most of the extra wins should come from.

AL Central

Cleveland Indians (94-68, 19 division titles, one tie for the wildcard)

The additions of Roberto Alomar and Wilfredo Cordero should reinforce an already potent attack that projects to finish second in the league in scoring. And it seems they have prospects galore to replace anyone who might falter during the season. It's hard to see how this team can fail to be an offensive powerhouse.

Their pitching also appears to be second best in the league. Their rotation has no ace but no weakness, and that's not counting Orel Hershiser, who failed to make our opening-day rotation and is reported to have a deal that allows him to join another club if he's not in Cleveland's plans. The bullpen looks terrific, with the additions of Steve Reed, Ricardo Rincon, and Jerry Spradlin in the past nine months.

There are other teams with similar projected winning percentages, but playing in the weakest division in baseball gives the Indians a much better chance to reach the playoffs than any team other than the Yankees.

Kansas City Royals (78-84, 1 division title, 2 wildcards)

I spent most of the off-season wondering if this might be one of the worst teams of our generation. They weren't very good last year and then spent the winter unloading their best offensive players. But my work on the projections has revealed some positives, and they could be my favorite team to watch in 1999.

Last year, they scored 714 runs, and we're projecting them to add 70 to that total in 1999. How can they add offense when they lost Dean Palmer, Jose Offerman and Hal Morris? It all depends on three top prospects who project to do very well in their rookie seasons. The first is Jeremy Giambi, who is coming off three terrific minor-league seasons, including a .372 average with 20 homers in 325 ABs at AAA Omaha last year. The second is 2B Carlos Febles, who hit .326, got on base at a .441 clip, slugged .530, and stole 51 bases with a good percentage at AA Wichita in 1998. And the third is CF Carlos Beltran, who batted .352 and slugged 14 homers in 182 atbats as Febles' Wichita teammate. Our system deflates these minor-league stats, so we're not projecting them to hit that well in the big leagues, but we are projecting each of them to make a solid offensive contribution this year, enough to replace what they lost and make the team better.

With the return of ace Kevin Appier, the pitching staff can only get better than the one that finished second-last in the AL in ERA last year. And the addition of defensive wizard Rey Sanchez to play SS will help all the pitchers. In fact, I think this could be one of the best defenses in the league.

So, this could be one of the surprise teams for 1999, and for that reason, they're my current favorite team to watch. But I don't really believe they're going to be a near-.500 team when the dust settles. It's entirely possible that some or all of the rookies will struggle for a while or end up back in the minors for more seasoning. If that happens, they might challenge the Twins as the worst offensive team in the league. And the club has openly said it will trade Appier. Rumors are that Jeff King might be available. And if they begin subtracting talent, this could become a 55-60 win team in a hurry.

Then again, they're currently 10-1 in spring training, and while spring training records often mean very little, there have been many teams that foreshadowed a big move in the standings with a strong spring campaign. I'm definitely going to be watching these guys.

Detroit Tigers (76-86)

The Tigers have been very busy since last October, but I think they're still a year or two away. They were 12th in the league in scoring last year and their off-season moves appear to have added only about 50 runs to that total. It's hard to score a lot of runs with three weak-hitting defensive specialists (Brian Hunter, Deivi Cruz, and Brad Ausmus) in the lineup. The good news is that their strong up-the-middle defense should help an unspectacular but fairly deep pitching staff move into the top half in team ERA. In our simulations, Justin Thompson led the staff with an average record of 16-11 and an ERA of 3.35.

Chicago White Sox (66-96)

This is the first of the teams that appears to have no chance. The loss of Albert Belle and Robin Ventura gutted their offense. The loss of Ventura and Mike Cameron significantly reduces the range of a defense that will be even less mobile if Frank Thomas plays 1B again this year. And, to make matters worse, the White Sox consistently led the league in errors by a wide margin in our simulations. The best reason to watch this team is its youth. There are young arms and young hitters everywhere you look, and it's quite possible that some of them will blossom this year. Not enough to make them a contender, but perhaps enough to make them interesting. By the way, we're projecting Frank Thomas to bounce back to hit .305 with 31 homers and 101 RBI.

Minnesota Twins (63-99)

Our early pick for the worst record in the majors this year. After Brad Radke and Eric Milton, there aren't any other pitchers to get excited about. Their offense projects to be last in the AL, as even their best hitters project to be only a little above the league average at their positions, with Todd Walker as a possible exception. Like the White Sox, this is a fairly young team, and it'll be interesting to see which of their youngsters takes a big step forward this year.

AL West

Texas Rangers (89-73, 12.5 division titles, 2.5 wildcards)

Projected to be third in the league in scoring, led by Juan Gonzalez, who averaged .311, 44 HR, and 134 RBI in our simulations. In 1998, Tom Goodwin increased his on-base percentage to .378, up 64 points from the year before, but our projections assume he'll fall back to .351 this year. If he can build on last year's success, it could add another ten runs to the team projection.

The pitching rotation is led by Rick Helling and Aaron Sele, two guys who were a little better than the league average but who won 20 and 19 games thanks to terrific run support last year. In our simulations, these guys turned back into the .500 pitchers they were prior to 1998. But the pitching staff is fairly deep, and the bullpen is anchored by one of the best closers in the game, John Wetteland, and this should be enough to get them into the playoffs, either as the winner of a tight division race or as the wildcard team.

Seattle Mariners (86-76, 5 division titles, 2.5 wildcards)

Once again, the Mariners should be one of the best offensive teams in the league. In our simulations, Ken Griffey (.287, 54 homers, 130 RBI), Alex Rodriguez (.313, 33 HR, 45 SB, 120 runs), Jay Buhner (36 homers), and Edgar Martinez (.323 with 99 walks, 43 doubles and 26 homers) continued to lead the attack.

As always, the question is whether Seattle has enough pitching and defense, especially without Randy Johnson. Assuming Butch Henry is healthy enough to take a regular turn in the rotation, and assuming Lou Piniella chooses to use him that way, the rotation could be OK. Henry, Jamie Moyer and Jeff Fassero give them three good starters. After that, it's questionable. And it remains to be seen if the rebuilt relief corps can get the job done, though it would be difficult for them to be worse than last year's crew.

My instincts tell me that the projected 86-76 record may be about as good as it can get. We're assuming a full year of normal production for Buhner plus two more great seasons from Griffey and Rodriguez. More importantly, the pitching could easily be worse than we're projecting. For one thing, we've got four lefties in the rotation, and Piniella is talking of moving one of them to the bullpen. Second, we're assuming Henry comes back at full strength, and that's no certainty. Finally, I have doubts about Piniella's ability to get the best out of his pitchers.

Anaheim Angels (81-81, 1.5 division titles, one tie for the wildcard)

This team doesn't appear to be as good as some of the off-season hype would indicate, but they still have time to make themselves better. In 1998, they were 10th in the league in scoring, 6th in pitching, and scored only four more runs than they allowed. With the addition of Mo Vaughn, we're projecting them to move up to 7th in scoring. It's not clear how the addition of Tim Belcher makes their pitching any better, and they appear to be headed for another middle-of-the-pack finish in staff ERA, especially with hard-throwing setup man Mike James on the sidelines.

But they could be better than this. Todd Greene says he's well enough to do some catching this year, and if he can play 80-100 games behind the plate, it will solve some lineup problems. As it stands, they have four outfielders (Darrin Erstad, Jim Edmonds, Tim Salmon, and Garrett Anderson) and a few guys to share the 1B/3B/DH jobs (Mo Vaughn, Troy Glaus, Dave Hollins and Greene). Putting Vaughn in the lineup takes atbats away from some of these others guys. That's why their projected ranking hasn't risen all that much despite what appears to be a plethora of good-to-great hitters. Having Greene catch gives them one less guy sharing the DH spot.

And that's why I think their potential is greater than these simulations suggest. A well-designed trade of surplus hitting for a front-line starting pitcher should allow the team to more efficiently apply the talent they have.

Oakland Athletics (77-85, one division title)

With the addition of Eric Chavez (projected to hit .298 with 28 homers) to an already good group of hitters, Oakland should be able to score with just about anybody in the league, especially when you consider that they play in a pitchers park. On the other hand, despite Kenny Rogers' decision to stick it out for another year, the pitching looks too thin to justify high expectations for the coming year. But the AL West is a tightly bunched group, and the A's could find themselves in the thick of the race if things go their way.

NL East

Atlanta Braves (102-60, 18 division titles, 2 wildcards)

Before Andres Galarraga was diagnosed with cancer and Kerry Ligtenberg came down with a bad elbow, this appeared to be one of the best of a series of terrific Braves teams. Even without Galarraga, our simulations put them into the playoffs every time and saw them compile the best record in the NL. The team is so deep that they are able to move Klesko to 1B and put some combination of Otis Nixon, Gerald Williams, and prospect George Lombard in left. We ran the simulations with Nixon as the everyday LF and leadoff hitter, figuring that (a) his defense (as a former CF) would make up for any offensive deficiencies and (b) he'd make a slightly better leadoff hitter than Walt Weiss.

The addition of Brian Jordan and Nixon to the outfield might give this club the best defensive outfield of any team in the 1990s. Andruw Jones is already the best defensive CF in baseball and Jordan is in the very top tier of RFs. And Nixon, a converted CF, should be more than adequate in LF even as a 40-year-old. That should make life even better for a pitching staff that is already the best in the business.

With the trade of Denny Neagle to Cincinnati, it's important that Kevin Millwood build on the progress he made last year. And it's not clear what they'll do if Bruce Chen isn't ready to be the #5 starter. If these two don't come through, the rotation starts to looks a little thin.

We ran our simulations before learning of Ligtenberg's elbow problems, so we had him listed as the closer and Wohlers as the setup man. Even with a healthy Ligtenberg, the Braves still lost the division race to the Mets in two of the twenty seasons. The loss of Ligtenberg may or may not affect the outlook for the team. If Mark Wohlers, who has made good progress this spring, comes all the way back to his peak level, they can absorb the loss of Ligtenberg. If Wohlers cannot handle the closer role, however, they'll be forced to turn to 24-year-old John Rocker.

New York Mets (93-69, 2 division titles, 12 wildcards)

For the first time in a while, the Braves face a serious challenger within their division. The Mets have a well-balanced attack featuring a bunch of guys who can hit for average, take walks, and toss in enough homers to bring those runners home. (Sounds a little like the 1998 Yankees, don't you think?) Bobby Bonilla is the only serious weakness in a strong defense. And while the pitching staff lacks stars, it's quite deep, and projects to be in the top three in the league. (They were fourth last year.)

There doesn't appear to be a lot of uncertainty with this club. As far as I know, nobody's been hurt this spring. No key player is trying to come back from a serious injury. They're not relying on unproven prospects. If there's a significant risk, it might be age. Rickey Henderson is 40 and his declining batting average (he hasn't reached .250 in three years) might encourage opposing hurlers to groove more pitches and force him to put the ball in play rather than walking him so often. And several of their key pitchers are well into their thirties. But that's about it for risk factors.

Philadelphia Phillies (76-86)

Offensively, they're projected to tread water. They'll have Ron Gant instead of Gregg Jefferies in left, and that's a plus, since Gant matches Jefferies' on-base percentage and offers more power. Rookie Marlon Anderson takes over from Mark Lewis at 2B, and while Anderson has a bright future, he cannot be expected to produce more offense right away. Several players had career years in 1998, and some are likely to pull back a little from those highs. All things considered, the Phillies project to score about the same number of runs (704) as they did last year (713).

The pitching is anybody's guess right now. Curt Schilling should continue to be one of the league's better pitchers. After that, you've got Paul Byrd, who came out of nowhere to make a few good starts last year; Chad Ogea, who has yet to fulfill his promise; Paul Spoljaric, who is trying to make the transition from reliever to starter; Tyler Green, who has had some nasty injury problems and hasn't had a good year yet; and Carlton Loewer, who seems to be highly regarded by management but had a 6.09 ERA as a rookie last year. And the bullpen, which was OK but nothing special last year, has been depleted by the trades of Jerry Spradlin and Mark Leiter.

Despite all of those questions, the Phillies staff looked all right in our simulations, finishing in the middle of the pack in runs allowed. This assumes that Schilling spends the whole year in Philly, Byrd's 1998 performance was not a mirage, Green is healthy, and Spoljaric is a league average starter. That's a lot of ifs, so I won't be at all surprised if this forecast turns out to be optimistic.

Montreal Expos (68-94)

Is it just my imagination, or did the Expos actually make it through a winter without shedding lots of talent? It seems they did, but they're still young enough, especially on the mound, to keep them from being a contender.

Last year, they were last in the NL in scoring with 644 runs, and our simulations suggest that the growth of their young players should enable them to add about 43 runs this year. Unfortunately, that still leaves them in the bottom quarter of the league, though they'd rank a little higher if they weren't playing in a good pitcher's park.

With Dustin Hermanson, Carl Pavano, and Ugueth Urbina, Montreal has the nucleus of what could soon be a very good pitching staff. But it's just not there yet, and we're projecting them to be near the bottom in pitching as well. Last year, they were 10th in staff ERA playing in a park that suppressed scoring by 21%.

Florida Marlins (64-98)

There aren't many teams who can look at a 98-loss projection as a step forward, but this is ten games better than they did last year. The improvement should come on the hill, as their hitting looks like it'll be worse this year than last.

Last year, they managed to produce 667 runs and finish a lofty 13th in the NL in scoring. This year, however, they won't have the benefit of partial seasons from Gary Sheffield, Bobby Bonilla, and Todd Zeile. It's all up to the kids, and it's just not clear where the offense is going to come from. Their top power threats are guys who are projected to hit 20-25 homers (Cliff Floyd, Derrek Lee, and Mike Lowell), and that's not a lot in this era -- Mark McGwire might hit more than all three of these guys combined. To make matters worse, Lowell's not going to be ready for opening day after his recent (and apparently successful) surgery for testicular cancer.

We're assuming that Alex Fernandez will return to lead the starting rotation, and that's a big reason why the pitching staff is projected to rise from last to 11th in pitching this year. That's still not very good, but it's an improvement.

NL Central

Houston Astros (92-70, 10 division titles, 1 wildcard)

The losses of Randy Johnson and Moises Alou are among the big reasons why this club, which won 102 games last year, averaged only 92 victories in our simulations. The other reasons are age (several key offensive players are in their thirties) and doubt about the ability of Jose Lima and Sean Bergman to maintain the level they were at last year. Still, they had the third best offense and an above-average pitching staff in our simulations, and that's enough to make them the favorites to repeat as division champs.

One reason for their success is depth. They have Richard Hidalgo ready to step into Alou's position. It looks as if Chris Holt is healthy again, and he might win a spot from Lima or Bergman or be ready to step in if either man falters. If Scott Elarton makes that transition from reliever to starter without losing much off his 1998 performance (3.32 ERA, 40 hits allowed in 57 innings), the staff will be that much better. And the growth shown by Lima and Bergman could be real. If it is, this club will have the rarest of luxuries, a surplus of quality starting pitchers.

St. Louis Cardinals (90-72, 6.5 division titles, 1.5 wildcards)

A lot of people will follow the Cardinals this year to see what Mark McGwire does for an encore. I'll be watching McGwire, too, but I'm more interested in the development of JD Drew, Eli Marrero, Edgar Renteria, Matt Morris and Fernando Tatis. Those players will have a lot to do with the Cardinals ability to contend this year.

Despite McGwire's heroics and big seasons from Ray Lankford and Brian Jordan, St. Louis finished only 6th in the NL in scoring. In our simulations, they moved up to second, thanks to a full year from Drew (.304, 98 walks, 34 homers), more offense from the shortstop position (.350 OBP for Renteria), a full season from Marrero (.252 with 15 homers), and, of course, another monster campaign from McGwire, who averaged 65 homers and 144 RBI (with highs of 75 and 168).

Even though they should have Matt Morris (2.54 ERA in 17 starts last year) for a full year, we're projecting their pitching staff to be in the middle of the pack, just as they were in 1998. Part of the reason is defense. Since the middle of the 1998 season, they've lost two players who are among the very best at their positions (Royce Clayton and Brian Jordan) and another (Gary Gaetti) who is still well above average despite his age (40). It's also far from clear that anyone is ready to replace the quality innings that Todd Stottlemyre provided before being traded to Texas last year.

Cincinnati Reds (84-78, 3 division titles, 1.5 wildcards)

If the Reds can sort out their crowded outfield situation, they have a chance to be in the thick of things this year. The addition of Denny Neagle gives the Reds a core group of three starting pitchers (Pete Harnisch and Brett Tomko are the others) that is the envy of all but a few clubs, and there are signs that Jason Bere is getting his game back together. Add a deep bullpen (though one that's a little light on lefties) and you have a staff that is projected to be in the top third of NL teams.

But are they going to score enough runs? That's the big question, and it appears their offense will struggle to match the league average in scoring. The heart of the offense -- Sean Casey, Barry Larkin, Greg Vaughn, Dmitri Young -- is quite respectable, but our lineup also included three players -- 2B Pokey Reese, 3B Aaron Boone, and CF Mike Cameron -- who haven't yet proven their ability to hit in the big leagues. And with Reese (who averaged a .291 on-base percentage in our simulations) being tried out as the leadoff hitter, it's far from clear who's going to kick-start this offense. Former 3B Willie Greene would look pretty good in this lineup right about now.

The OF/1B situation seems to be the key. In our simulations, we weren't able to find much playing time for some players (Michael Tucker, Hal Morris, Chris Stynes, Jeffrey Hammonds) who might have enough trade value to attract a leadoff man or another hitter who can extend the attack a little further down the batting order.

Chicago Cubs (80-82, one tie for the division title)

The Cubs offense is projected to be a little weaker this year than last, in part because Sammy Sosa hasn't yet proven that he can hit a ton of homers every year. Before matching McGwire swing for swing in 1998, Sosa had never gone over 40. And because our projections take the last three years into account, he's projected for 'only' a .275 average with 50 homers and 123 RBI. The other reason is age -- every starter other than Jose Hernandez is in his thirties and therefore in the phase of his career where performance tends to decline.

The Cubs pitchers were 11th in the NL in team ERA last year and it looks as if they're destined for a similar ranking in 1999. Over the winter, they lost Mark Clark and added Jon Lieber, but didn't make any moves that would fundamentally increase their mound talent.

The biggest variable is Kerry Wood. We assumed he'd be able to take his regular turn this year, but that's no sure thing after he missed last September with elbow problems, spent a few weeks battling a viral infection this winter, and is now experiencing pain in his elbow again. In addition, after walking 7-8 hitters per game in the minors, he suddenly found the strike zone and cut his walk rate to 4.6 per nine innings. Because our projections take those minor-league seasons into account, he walked 5.9 per game in our simulations, and because he was pitching behind in the count more often, he also gave up a few more homers. But if Wood is 100% healthy and matches what he did last year, you can add a few games to the Cubs projected win total.

Milwaukee Brewers (73-89)

The Brewers don't appear to have a legitimate shot to contend, and with the possible exception of Ron Belliard, they don't have any top prospects bursting on the scene. Most of the questions are about whether certain guys (Jeff D'Amico, Cal Eldred, Bill Pulsipher, Dave Nilsson) can come back from injury or stay healthy for a full season. This is definitely not a bad team, and there are quite a few players that I like (Jeff Cirillo, Fernando Vina, and Jeromy Burnitz, among others), but it doesn't look like it will be an eventful season for the Brewers.

By the way, our simulations assumed that Eldred and Pulsipher would be healthy for the full season, D'Amico would not play a role, and Nilsson would be able to handle the full-time catching job. I'll leave it to Milwaukee fans to judge whether these assumptions are, overall, optimistic or pessimistic.

Pittsburgh Pirates (72-90)

Fascinating. Here is an organization that appears to be blessed with some very good prospects -- Aramis Ramirez, Chad Hermansen, Warren Morris and Abraham Nunez -- and has chosen to delay their development by bringing in some veteran free agents. The stated reason is that the club's 5-22 September shows that it was too young and immature to maintain a professional approach to the game. Let's assume that's true. Let's say they have a genuine need to add veteran leadership. Why bring in a bunch of guys -- Ed Sprague, Mike Benjamin, Pat Meares -- who have established that they are below-average players at their respective positions? Wouldn't it make more sense to marshall all of the resources spent on these players and sign one really good player, someone with great work habits, someone who can teach the kids how to approach the game without taking playing time away from all of them?

In our simulations, we assumed that Morris would take enough playing time away from Benjamin to get about 200 atbats, and that Ramirez, Hermansen, and Nunez would spend the year in the minors. If one or more of these kids gets a chance and make the most of it, the team might be better than their 72-win forecast.

NL West

Los Angeles Dodgers (92-70, 14.5 division titles, one wildcard)

This has become a tough division, and the Dodgers will have to play up to their potential to win it, but the addition of Kevin Brown makes them the favorite. Brown averaged a 17-9 record in our simulations and could easily win twenty if he gets decent run support. The pitching staff is very, very deep, and should finish in the top two or three in team ERA.

There are three main questions on offense. First, will Todd Hundley be healthy enough to be the everyday catcher and to hit like he did before his injury? Second, will Adrian Beltre play well enough to be the everyday 3B for the entire season? And, finally, will it be Mark Grudzielanek or Jose Vizcaino at shortstop? Our assumptions were yes, yes, and Grudz.

Arizona DiamondBacks (83-79, 4.5 division titles, one wildcard)

Arizona was busy this winter, adding pitchers Randy Johnson, Todd Stottlemyre, Armando Reynoso, and Greg Swindell, signing Steve Finley, and trading for Tony Womack and Luis Gonzalez. The talent they've added to their pitching staff should move them from 13th in the league in team ERA to about 5th. Their offense is only slightly improved and should remain in the bottom half of the league.

One big question is the role Tony Womack will play. In Pittsburgh, he was a subpar defensive 2B with two skills -- producing a batting average in the .280 range and stealing lots of bases with an excellent success rate. Arizona wants to move him to the outfield, but Womack wants to stay at 2B. We pencilled him in as the leadoff hitter and everyday LF, minus some playing time to account for the broken bone that will keep him out for the next few weeks, because that's what the club says they're going to do with him. But his career on-base percentage (.325) is nowhere near good enough for a leadoff hitter, and I think the team would be better off if they gave the LF job to Dave Dellucci. Dellucci offers a better on-base percentage, more power, and better defense.

San Francisco Giants (81-81)

The Giants are shaping up to be a team that's in the middle of the pack offensively and defensively, one that will be in the races for the NL West and the wildcard but isn't a favorite in either case.

Because they were second in the league in scoring last year, I was expecting them to rank higher offensively. But Barry Bonds and Ellis Burks are in the phase of their careers where skills tend to erode slowly, Marvin Benard and Jeff Kent are unlikely to match the career years they had in 1998, and they've lost some offense from the catcher position with the departure of Brian Johnson.

One possible upside surprise is JT Snow, who has given up switch hitting. He batted only .164 against lefties last year, and we've projected him to do only a little better than that in 1999. But he can hardly do any worse swinging only from his better side, and it's quite possible he'll add to his output with this move. Besides, if doesn't work, he can always go back to switch-hitting. Seems like this move has no downside and some potential for success, and that can only help the team.

On the hill, our projections assume that Shawn Estes will bounce back. He was terrific (19-5, 3.18) in 1997 but slumped badly (7-12, 5.06) last year. We're also assuming that Osvaldo Fernandez won't play a meaningful role. Fernandez is trying to come back from an elbow injury that has kept him out for a year and a half, so his return is speculative.

Colorado Rockies (80-82, one division title)

Last year, we projected the Rockies to win 77 games. They won 77 games. And a well-respected manager, Don Baylor, was fired for failing to live up to expectations. In my opinion, those expectations were unrealistically high, perhaps because the massive impact of Coors Field (which has increased scoring by 55% over the past three seasons) makes it hard to evaluate talent on both sides of the ball. This year, another well-respected manager, Jim Leyland, takes the helm of a team that appears to us to be mediocre but which probably appears to Colorado management and its fans to be a serious contender.

As usual, the Rockies are projected to lead the league in scoring with a big helping hand from their home park. In 1998, the Rockies hitters were awesome at home -- a .325 average and a .519 slugging percentage -- and below the league average on the road -- .257 and .401.

As usual, the Rockies pitchers are projected to finish last in team ERA, but once again, the home park plays a major role. Last year, this staff had an ERA of 5.70 at home but matched the league average of 4.23 on the road.

A year ago, their big free-agent signing was a pitcher (Darryl Kile) with one good season in the previous three who did his work in a pitcher's park (Astrodome). We figured the three-year averages and adjusted for the difference in ballparks and came up with a projected ERA of 5.14. His actual 1998 ERA was 5.20. And people were disappointed.

I fear we're destined to repeat history. This year, the Rockies' big free-agent signing is Brian Bohanon, a pitcher with one good season in the last three who is moving from a pair of pitcher's parks (Shea Stadium and Dodger Stadium). Sound familiar? I hope, for Brian's sake, that Colorado management isn't expecting way too much.

San Diego Padres (69-93)

The defending NL champs were not as good as their record suggested last year. Their hitters combined for a .253 batting average, .330 on-base percentage, and .409 slugging average, all of which were below the league norm. They outscored their opponents by 749-635, a margin that normally produces 94 wins, four less than the 98 they finished with last year.

So they could ill afford to lose talent over the winter. But lose talent they did. Ace pitcher Kevin Brown went to the Dodgers as a free agent. Ken Caminiti and Steve Finley signed elsewhere. They traded Joey Hamilton to Toronto for Woody Williams, and that seems like a net loss in pitching talent. They moved Greg Vaughn and his 50 homers to Cincinnati in a multi-player deal, getting Reggie Sanders and others in return.

But I don't see this as a repeat of the year-ago dismantling of the Florida Marlins. There are only a few teams that could afford Kevin Brown, and San Diego isn't one of them, so they had no chance. Florida, on the other hand, traded Brown while he was still under contract. Caminiti is an expensive, older, and oft-injured player who was holding back 3B prospect George Arias. Steve Finley just turned 34 and is coming off a bad season. The Vaughn deal may turn out to be a long-term win. The only move I question is the trade of Joey Hamilton.

Even though the players who are left project to be in the bottom quarter of the league both offensively and defensively, this should still be an interesting team to watch. To what extent can Arias replace Caminiti at third? Is Ruben Rivera ready for a breakout season as the new CF? Will pitching phenom Matt Clement be this year's Kerry Wood? Can they find a taker for Randy Myers' salary and improve the team through a trade? Can Reggie Sanders stay healthy and bounce back to his peak level? It wouldn't take many yes answers to move this team into the fringes of the wildcard race this year.

What-if?

What if Moises Alou were healthy? What if Roger Clemens had stayed in Toronto? What if Todd Hundley isn't healthy or cannot hit for power like he once did? What if Todd Greene's shoulder allows him to catch 100 games for the Angels?

To find out, we ran a few more season simulations with these changes. In six tries with Clemens, Toronto won two wildcards, while the Yankees still won the division with ease every time. With Moises Alou in the lineup, the Astros won the NL Central four times in six tries. Hundley's reduced contribution didn't affect the Dodgers much, as they still had enough to win five division titles. Greene had the most impact, helping Anaheim to two division wins and a tie for the wildcard in four seasons.

What if . . . well, I'm sure you can come up with lots of questions of your own. The point is that we (and our customers) can update rosters, adjust the projected stats for individual players, and change the manager profiles in order to answer other questions. And this is a big part of the fun of using these projections. We can't tell you how the season will really come out, but we can provide a laboratory for playing out a variety of scenarios to see what is most likely to happen given different assumptions about performance levels, injuries, and managerial decisions.

Parting Thoughts

We had a blast doing this. These projections are a lot of work, but the moment we kicked off the first simulation of the 1999 season, it was worth it. It's a real kick to watch the season play out in 20 minutes, look at the stats, see why some teams fared better or worse than expected, tune the starting lineups or bullpen assignments, then run the season again.

And now that I've spent the time putting together these rosters and looking over these results, I'll go into the season with dozens of things to look for. I've always enjoyed poring over the boxscores every morning, but it will be even more fun now. I can't wait for opening day.

Copyright © 1999. Diamond Mind, Inc. All rights reserved.