Baseball Articles | Index
1998 Team Predictions -- Keeping Score
By Tom Tippett
Every spring, numerous baseball magazines come out with predictions for the coming season. Some go so far as to predict that actual win totals for each team; most stop at the final standings. And then they're forgotten. I don't ever remember seeing an article after the season that looks at how good or bad those projections turned out to be.
This is the first year that Diamond Mind has pre-played the season to get an idea of how we thought each team would do. Since we plan to do this every year, its important for us to measure our progress and see where we can do better in the future.
With a huge assist from Pete Palmer (co-author of Total Baseball and The Hidden Game of Baseball), this article compares our projections and those of fourteen national publications to the actual standings for 1998. Pete's been projecting team standing for over 20 years, and he routinely collects the predictions of others and sees how everyone did at the end of the year. He was gracious enough to share his 1998 data with us so we could see how Diamond Mind stacked up this year.
The good news: according to Pete's ranking system, we were closest to the real standings. 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 this 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, so we used 2.5 for the projected place for both teams.
Much more interesting, in my opinion, than the overall scores of these publications, are the specific teams that forecasters were most right or wrong about. For me, anyway, it was a lot of fun to look at the projections side by side and see how each differed from the others. Here are some of the things that I noticed:
AL East. Twelve publications predicted that the Yankees would take the division, and the others all picked Baltimore. Only Spring Training had the Orioles as low as fourth; three others (including Diamond Mind) had them third. Three publications had the Red Sox in second. Everyone had the Devil Rays at the bottom.
AL Central. You won't be surprised to learn that everyone picked Cleveland to win this weak division. As it turned out, Detroit was the key team in the prediction scores. They finished last, but nobody picked them there. We had them fourth, and everyone else put them second or third.
AL West. Everyone picked Seattle to win the division, and everyone was wrong. All but one picked Oakland to finish last, which they did. There was a mix of opinion about whether Texas would finish ahead of Anaheim.
NL East. Another easy one, as everyone correctly picked Atlanta to finish first. As the season unfolded, we became increasingly wrong as the Marlins plummeted to the basement and stayed there. You see, we called them to finish third, figuring they still had more talent left after their fire sale than did the Phillies and Expos. Even before they got rid of Sheffield and Bonilla and Johnson (for Piazza and Zeile, both of whom were subsequently traded away), they were mired in the basement. So we figured this would be our Achilles heel in comparing our projections to others. But it didn't hurt much, after all, because nobody picked the Marlins for last place, six others also picked them for third, and one even had them in second. Another interesting note is that the Phillies were picked last by twelve publications despite their strong second half in 1997.
NL Central. This was a crapshoot. When the season started, there didn't seem to be any dominant teams or any weaklings in this division. Eleven publications had the Cardinals in first, three had Houston, and we were alone in picking the Reds. (The Reds finished fourth, and this was the only team where we missed by as many as three places.) Only one other publication had the Reds finishing as high as fourth, so we were all alone on this prediction. Milwaukee was projected by thirteen publications to better than their fifth place finish. Pittsburgh was also routinely overrated. Nobody had the Cubs placing higher than third.
NL West. Nine publications had the Dodgers in first, four had Colorado, and two correctly identified the Padres as the division champs. Everyone else had the Rockies finishing in the top three; we were alone in placing them fourth. The Giants were a consensus pick for fourth place, with only Diamond Mind and Mazeroski correctly picking them second. Sport magazine had the Giants coming last, and they were the only ones to pick Arizona for something (fourth) other than the basement.
Looking over all the results, it seems that the Diamond Mind score was best because we were much more realistic about the chances of a few teams (Baltimore, Colorado, Detroit, Milwaukee, St. Louis) that were favored by some publications. And we got away with our "mistakes" (Seattle, Florida, Los Angeles) because almost everyone else got them wrong, too.
I'm not sure what to make of all this. It's just one year, and it's entirely possible that we were just lucky. Time will tell whether our approach to projecting seasons is consistently better than average. But it sure is fun to make those predictions and then take a look at them later, so we'll keep doing it.
Copyright © 1998. Diamond Mind, Inc. All rights reserved.