October 25, 2014

And your 2011 World Series Winner is…

October 17, 2011 by · 4 Comments 

About 30 years ago, Bill James introduced a prediction system that picked the World Series winner with 70% accuracy. He wrote about the system for Inside Sports magazine in 1982, then expounded on it in his 1984 Baseball Abstract. He developed the system in 1972 and it accurately predicted the World Series winner at a 68% clip for 12 years. Then he went back and used it to predict all of the postseason series of the 20th century and it returned a success rate of 73%. So I’ve decided to apply the system to this year’s Fall Classic to see which team is expected to win.

The system hasn’t been updated since 1984 (as far as I know), so the point distribution may not be as precise as it should be. For example, James awards 19 points to the team that threw more shutouts during the regular season because at the time of the original research, teams with more shutouts had won 19 more postseason series than teams with fewer shutouts. There have been almost 40 years of postseason games since he developed the system and I have no doubt the point distribution should be adjusted based on the last four decades of playoff games, but for now we’ll stick with the old system and see how it does. Maybe someday I’ll update it and see how it works moving forward (or maybe James will).

Here’s how the system distributes points:

  • Compare the won/lost records of the two teams involved. Award one point per half-game difference to the team with the better record.
  • Give 3 points to the team that has scored more runs.
  • Give 14 points to the team that has hit fewer doubles.
  • Give 12 points to the team that has hit more triples.
  • Give 10 points to the team that has hit more home runs.
  • Give 8 points to the team with the lower team batting average.
  • Give 8 points to the team that has committed fewer errors.
  • Give 7 points to the team that has turned more double plays.
  • Give 7 points to the team whose pitchers have walked more men.
  • Give 19 points to the team that has thrown more shutouts.
  • Give 15 points to the team whose ERA is further below the league ERA.
  • Give 12 points to the team that has been in postseason play more recently. If both last appeared in postseason play in the same year, award the points to the team that was more successful at the time.

James included a 13th rule that pertained only to league or divisional playoffs and that awarded 12 points based on head-to-head competition during the season. At the time he developed the system there was no inter-league play, so it wouldn’t have applied to the World Series participants since the Fall Classic would have been the first time an American League team would have played a National League team (except for the All-Star game and during Spring Training, of course). FYI, according to the Boston Globe, the only time these teams have met in inter-league play came back in 2004 when the Cardinals took two of three in Texas.

Let’s see how the points are distributed between Texas and St. Louis:

  • Compare the won/lost records of the two teams involved. Award one point per half-game difference to the team with the better record.

Texas had a regular season record of 96-66, while St. Louis posted a mark of 90-72. That’s a 6-game difference, which gives the Rangers 12 points. They’re up 12-0.

  • Give 3 points to the team that has scored more runs.

The Cardinals paced the NL in runs scored with 762. Texas was third in runs scored in the AL, but plated 93 more runs than the Cards. That’s another three points for the Rangers and they now hold a 15-0 lead.

  • Give 14 points to the team that has hit fewer doubles.

This was a close one. The Cardinals smacked 308 doubles to the Rangers’ 310, giving St. Louis the points and cutting Texas’ lead to 15-14.

I should explain why the team with fewer doubles gets the points before I move on. James concluded that teams with more doubles are more aggressive on the base paths and that they exploited weaknesses during the regular season that won’t be present in the World Series. Based on Texas’ stolen base attempts, they do appear to be much more aggressive on the bases than St. Louis, who attempted fewer than 100 steals on the year. At the time James devised this formula, teams with more doubles had lost approximately 60% of the time in the postseason. So, the Red Birds get 14 unanswered points and are within one of the Rangers.

  • Give 12 points to the team that has hit more triples.

The Rangers had 10 more three-baggers than the Cardinals, lacing out 32 triples to St. Louis’ 22. Texas gets the points to move ahead, 27-14.

  • Give 10 points to the team that has hit more home runs.

This one is a blowout in favor of the Rangers, who belted 210 homers to St. Louis’ 162. Suddenly Texas has a sizable lead, 37-14.

Another interesting argument can be made that perhaps ballpark factors need to be considered in future versions of the prediction system. Rangers Ballpark ranked first in HR factor at 1.500, while Busch Stadium ranked 27th with a much lower .774. On the other hand, teams that hit more homers have probably won 10 more postseason series than those who don’t (hence the 10 points) and this is James’ system, so regardless of ballpark factors, Texas gets the points.

  • Give 8 points to the team with the lower team batting average.

Both teams led their leagues in batting. The Rangers hit .283 on the season and the Cards hit .273 to earn the points here.

An explanation is in order as to why the team with the lower AVG gets the edge. “Why do teams with high batting averages do poorly in World Series play?” asked James. “A simple reason: it takes them too many hits to score.” The Rangers aren’t exactly a Punch-and-Judy club—only the Red Sox had a higher slugging percentage and that was by one point—and they can score runs in bunches, but James’ system gives the points to St. Louis. Texas’ lead is now 37-22.

  • Give 8 points to the team that has committed fewer errors.

No explanation needed here. Both teams were well above league average in committing errors, and by “above” I mean far too many. But Texas (114) committed two fewer than St. Louis and gets the points to extend their cushion to 45-22.

  • Give 7 points to the team that has turned more double plays.

Both teams led their leagues in double plays turned, but St. Louis gets the points for turning 167 to Texas’ 162. The Rangers’ lead is down to 45-29.

  • Give 7 points to the team whose pitchers have walked more men.

Both staffs were among the best at avoiding walks, but Rangers hurlers issued 19 more free passes and they get seven more points to push their lead to 52-29.

To be honest, I’m not quite sure why the team with more walks gets the points here and James doesn’t really explain it. He attributes a lot of the walks to the starters at the back end of the rotation, but admits that those hurlers will have little impact on the outcome of the World Series. I suppose that more walks means fewer chances to get a hit that might plate a run or an extra-base hit that might plate multiple runs while putting runners in scoring position.

  • Give 19 points to the team that has thrown more shutouts.

The Rangers dominated this category during the regular season, tossing a major league best 19 shutouts. St. Louis’ pitchers hurled only 9. This is becoming a runaway as Texas increases its lead to 71-29.

  • Give 15 points to the team whose ERA is further below the league ERA.

Texas’ ERA of 3.79 is 0.29 below the AL average of 4.08. St. Louis’ 3.74 mark is only 0.07 points lower than the NL average. The Rangers score another 15 points to make the score 86-29.

  • Give 12 points to the team that has been in postseason play more recently. If both last appeared in postseason play in the same year, award the points to the team that was more successful at the time.

Well this is a landslide victory in the Rangers’ favor. They’re making their second straight trip to the World Series whereas the Cardinals watched the postseason from the comfort of their couches last year. Texas gets another 12 points to make the final tally 98-29 and make them the favorite to win their first World Series title in franchise history.

If the Series follows the point distribution pattern, this is going to be a short Fall Classic.

Category (points) Edge TEX Pts. STL Pts.
Won/Lost Record (1 per 1/2 game) 99-66 90-72 12 0
More Runs (3) 855 762 3 0
Fewer Doubles (14) 310 308 0 14
More Triples (12) 32 22 12 0
More Home Runs (10) 210 162 10 0
Lower Batting Average (8) .283 .273 0 8
Fewer Errors (8) 114 116 8 0
More Double Plays (7) 162 167 0 7
More Walks Issued (7) 461 448 7 0
More Shutouts (19) 19 9 19 0
ERA Lower than LG (15) 0.29 0.07 15 0
More Recent Postseason (12) 2010 2009 12 0
2011 World Series Champion 98 29

Comments

4 Responses to “And your 2011 World Series Winner is…”
  1. Looking pretty good so far!

  2. Ron Ryan says:

    Of course, it didn’t work out. The formula looks impressive, to be sure, but I wonder. If an educated baseball fan, who follows the teams closely and understands the role of luck in the game, were to chart his WS predictions over twenty years, it wouldn’t surprise me if he were right 70% of the time. I’ve been recording my for 40 years, and I’ve been right 75% of the time, including some upsets. That’s without looking at a single stat.

    Among those successful picks were the 1990 Cincinnati Reds. Under the James forumula, that team finished 70 points behind the Oakland As. However, I picked the Reds because I thought their bullpen, with it’s ability to reduce games to five-inning contest, would provide the crucial difference in the Series. That proved to be the case.

    I don’t understand parts of James’s formula, and I wonder if it needs revisiting. Why would the team with more doubles and a lower batting average LOSE points?

  3. Mike Lynch says:

    Ron,

    I’ve crunched the numbers and will post them to the site soon. I think you’ll find them interesting.

  4. Ron:

    “I don’t understand parts of James’s formula, and I wonder if it needs revisiting. Why would the team with more doubles and a lower batting average LOSE points?”

    Teams that hit more doubles tend to hit less home runs, and teams with lower batting averages tend to be teams that draw more walks and hit with more power, so that’s why they end up losing points for those in James’ formula.

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!