Ten Things About Momentum in the Postseasonby Dave Studeman
September 29, 2005
In 1969 the Mets routed the Orioles in the World Series, 4-1. The Mets had gone 100-62 that year, for a winning percentage of .617. Pretty darn good, but not as good as the great Orioles' record of 109-53, .673.
You might say, however, that the Mets had momentum in their favor. In the month of September, the Mets' winning percentage was an incredible .767 (ask the Cubs), .137 points higher than the Orioles'. In fact, that September record was the best of any team that has won the World Series in the last 35 years. It would seem that the Mets' momentum meant more in the Series than the Orioles' consistent excellence.
I've been reading a lot lately about Wild Card teams winning the last three World Series. It's a significant trend, to be sure, and disturbing in some ways. Some say that the Wild Card format gives hot teams a chance to sneak into the playoffs based on late-season momentum, which then gives them an advantage against teams that have been consistently excellent throughout the season but not particularly hot at the end. That feels like an unfair advantage, doesn't it?
Exhibit Number One is the Marlins team of 2003, which had a .692 record in September vs. .562 overall and went on to win the Series crown. In fact, I sense a creeping baseball truism sneaking into the e-mails and blogs I read: Momentum can overcome Good in the postseason.
So my question is, has anyone shown this to be true? I personally haven't seen a thorough analysis of momentum in postseason play anywhere. So, with my best sabermetric hat on my little pointy head, I loaded all the postseason results beginning in 1969 into my Excel spreadsheet to see what I could learn about Good and Momentum.
I chose 1969 because that was the first year of the two-division format; the Wild Card format wasn't fully introduced until 1995. By going back to 1969, I've got 35 years, 184 teams and 145 different series (excluding 1994, when baseball canceled the greatest show on earth). I looked at two winning percentages for each team: winning percentage for the year and winning percentage in September only.
I found out a lot more than 10 things. But that's the name of the column, so let's jump in.
Neither a team's overall record nor its record in September matters a whole lot.
In the 35 years from 1969 through 2004, the team with best overall winning percentage won the World Series only eight times. Let me emphasize: the team with the best regular-season record has won the World Series only 23% of the time. The winners include some of the best and best-known teams of our time: the 1970 Orioles, the Big Red Machine in 1975 and 1976, the Mets in 1986 and the 1998 Yankees.
How about the teams with the best September record? The answer is exactly the same: they won eight World Series, too. Same impact. In six of the eight examples, however, the team with the best September record was also the team with the best overall record. So there's a lot of overlap between the two groups.
In fact, in about half of the last 35 years (17, to be exact), the team with the best regular-season record was also the team with the best record in September. Of those 17 teams, six won the World Series. Even teams that were Good and had Momentum won it all only 35% of the time.
On average, teams that qualified for the postseason had a higher winning percentage in September.
Added together, these 184 teams had an overall winning percentage of .589 and a .612 winning percentage in September. This makes sense. These are the teams that qualified for postseason play, and a number of them made it on the wings of their late-season drives.
Here's a breakout of the overall and September records for all postseason teams over four-year periods.
Years Overall Sept. Diff 1969-1972 0.608 0.627 .019 1973-1976 0.590 0.605 .015 1977-1980 0.595 0.625 .030 1981-1984 0.568 0.583 .015 1985-1988 0.592 0.636 .043 1989-1992 0.584 0.607 .023 1993-1996 0.581 0.607 .026 1997-2000 0.587 0.580 -.007 2001-2004 0.597 0.642 .045 Grand Total 0.589 0.612 .023In the last four years, teams entering the postseason have had a significantly better September record than most previous years. This stands in stark contast to 1997-2000, when postseason teams actually had a worse record in September than their overall record, and it may contribute to the perception that momentum matters. We'll look at this more closely below.
Overall, Momentum is less important than being Good.
Now, here is the same table, but for World Series winners only:
Years Overall Sept. Diff 1969-1972 0.621 0.700 .079 1973-1976 0.608 0.579 -.029 1977-1980 0.599 0.683 .084 1981-1984 0.597 0.573 -.024 1985-1988 0.584 0.594 .010 1989-1992 0.588 0.595 .007 1993-1996 0.593 0.622 .029 1997-2000 0.604 0.502 -.103 2001-2004 0.586 0.620 .033 Grand Total 0.598 0.607 .009Compare the two tables. You can see that the World Champs, on average, had a better regular-season percentage (.598 vs. .589 for all postseason teams). Good does matter. But the champs actually had a slightly lower September record (.607 vs. .612) than the average postseason team. The most recent four-year period, with three Wild Card winners, yielded a lower overall record for the winners than the initial entrants (.586 vs. .597), but the September records were lower too (.620 vs. .642).
Momentum was King from 1977 through 1980.
If ever Momentum was king, it was from 1977 through 1980 when the September record of World Series champs was .084 greater than their overall record. 1977 was the year the Yankees returned to glory, winning their first World Series since 1962. This was the Billy Martin/Reggie Jackson fight year, the year that made George Steinbrenner famous for buying his teams. Reggie Jackson established himself as "Mr. October" with his three-homer game in the 1977 Series. But it was their 41-12 record after August 10 that won the pennant.
1978 was even more dramatic, one of the most memorable pennant races of all time. The Yankees were seven games back on August 30, but the Red Sox subsequently went 3-14 (including six losses against the Yankees), to put the Yanks up 3.5 games on September 16. The Red Sox turned it around and won their last eight games to force a one-game playoff, which the Yankees won 5-4 on Bucky "F---ing" Dent's three-run homer. The Yankees played .733 ball that September (the third-highest total of any World Series winner) and went on to beat the Royals and Dodgers (same teams they beat in 1977) for the World crown.
1979 was the year everyone talked like the Pirates, as in We Are Family and Willie Stargell. The Pirates were a .500 team in early July, but they went 61-30, .670 over the last three months to win an exciting pennant race against the Expos. They swept the Reds (who were no longer quite so Big and Red) and won a fine 4-3 Series against the Orioles.
And then came 1980, the only year the Philadelphia Phillies have ever won the World Series. Like the Pirates the year before, the Phillies had a tight division race against the Expos, winning it in a final three-game showdown in Montreal. The Phillies weren't a great club; their 91-71 record was the lowest of the four teams that qualified for postseason play. But they played .655 ball in September and won perhaps the closest League Championship Series ever against the Astros, when the last three games all went into extra innings. They then beat the Royals in the postseason to win it all.
Yet the team with the most Momentum of all didn't even win one series.
The Kansas City Royals haven't always been bad. It just seems that way. In 1977, the Royals had their best season ever, going 102-60, .630. These were the Brett/White/Otis Royals, with Whitey Herzog at the helm. As of September 1, however, the Royals were only playing .585 ball, 2 1/2 games ahead of the White Sox.
After September 1, the Royals went 25-5, .833 to run away with the division. This is the best September record of any team to make the postseason over the past 35 years. So what happened in the postseason? The Yankees beat them 3-2, particularly by scoring four runs in the last two innings of the last game to win 5-3. Which perhaps proves that nothing will stop momentum faster than a faulty bullpen.
Turnaround is fair play, however. In 1980, the year of the Phillies' victory, the Yankees actually had the best September record of any postseason team (.750) after winning a very tough race against the Orioles. Yet the Royals swept them in the League Championship Series.
From 1997 through 2000, Momentum was bad.
Each World Champions from 1997 through 2000 had a worse record in September than the season as a whole. In fact, two of those teams had September records below .500. The 1997 Marlins made it to the postseason through the Wild Card slot and became the first Wild Card team to win it all. However, you can't really say they had momentum on their side, as they played only .444 ball in September, the second-lowest September record of any World Series winner in the past 35 years.
The worst September record for a World Series winner was .433; the Yankees in 2000. In fact, the 2000 Yankees lost 15 of their last 18 games but still managed to finish first in the American League East. They then went on to beat Oakland, Seattle and the Mets for the world title.
By the way, the fourth-worst September record of any World Series champion belongs to the 2001 Diamondbacks, who played .476 ball the last month of the season. But the arms of Johnson and Schilling pulled Arizona through each postseason series until they had won it all.
The irony here is that, for all the recent focus on Momentum and the postseason, September records meant almost nothing a few years ago.
Specific matchups follow the same pattern.
Up to now, this analysis has been relatively superficial. Postseason analysis is complicated because teams don't all play each other. If a hot team matches up with a good team and another one matches up with a relatively worse team, a top-level view may not capture this nuance. So I looked at every postseason series from 1995 through 2004, the Wild Card years, and compared the overall and September records of the teams in each series.
Here's an example: the 2003 Marlins (.562 overall, .692 in September) first played the Giants in the postseason, who had a better overall record (.621) and the same September record (.692). They then played the Cubs, who actually had a worse overall record (.543) and a better September record (.704). They beat them anyway. In the World Series, they beat the Yankees, who had a better overall record (.623) and a slightly lower September record (.667). So, compared to their postseason competition, the Marlins were 102 points worse in the regular season, but 14 points better in September.
Now check out this table of the results of every series from 1995-2004 (counted twice, for each team's perspective):
Win% Diff Series Win% <-0.120 5 .200 -0.120--0.04 25 .480 -0.040-0.040 80 .500 0.040-0.120 25 .520 >0.120 5 .800 Grand Total 140 .500The left column is the difference between two teams' regular-season winning percentage. A difference of .120, for instance, might be the difference between a .640 team and a .520 team. Reading the bottom line of the table, teams with a winning percentage at least .120 better than the opposition has won four of five series.
By the way, the data is perfectly symmetrical because it includes the viewpoint of both teams. For every five teams with a record over 120 points higher, there were five teams with a record 120 points lower.
The key point here is that the results are not only symmetrical, they're logical. As the difference between a team's regular-season record and its opponent's moves from negative to positive, so does the team's chance of winning. And most series (130 of 140) are close enough that the outcome is relatively random; if .500 is completely random, these results range from .480 to .520.
Even on a matchup level, Momentum doesn't help.
I apologize, because the next table is kind of complicated (and this article is already long enough!). But the above table got me wondering: could Momentum be the difference in winning and losing when opponents are relatively well-matched?
To find out, I took the results from above and added a wrinkle: each team's relative advantage in their September record. Following is the table of how well each team did (expressed as the series winning percentage) when evenly matched but with differences in September play:
Difference in September Records Win% Diff Worst Bad Better Best -0.12--0.04 .500 .444 .625 .250 -0.04-0.04 .800 .536 .469 .200 0.04-0.12 .750 .333 .545 .500 Grand Total .722 .500 .500 .278To help interpet the table, look at the Grand Total line. What it's saying is that when teams had less Momentum (the "worst" difference in September records) they actually won 72% (.722) of the time. On the other end, teams that had the biggest advantage in September records only won 28% (.278) of the time.
That's right. There has actually been a negative correlation between Momentum and winning in the postseason the last 10 years. I can't explain it; it's probably just random variance. But momentum has meant less than nothing in the postseason over the past 10 years.
Did something happen in 2002?
To be fair, the story has been somewhat different the last three years. In 2002, 2003 and 2004 Momentum has actually been positively correlated with postseason success. One good example would be the 2003 Cubs (.543 winning percentage, .703 in September) beating the Braves (.623, .538) before going on to lose to the Marlins. But this is a recent phenomenon, against a background of 32 previous years in which momentum didn't really matter. Have things changed?
My theory is that we're confusing cause and effect here. Wild Card teams have won the past three World Series; during the same time, Wild Card teams have had great Septembers, with an average September winning percentage of .681. This is why it suddenly looks like Momentum matters. It just so happens that the high-momentum teams won.
Counterpoint: in 2001, the two Wild Card teams (Oakland and St. Louis) had an average September record of .791!! Yet neither one even made it to the World Series.
What this means for 2005.
What does all this mean for this year? It means that some people will pick the hottest team (say, the Indians or Yankees) to win it all, because they have momentum on their side. Based on this study, I'd say that the postseason is still a relatively random event. You're better off using the guidelines developed over a year ago by our own Vinay Kumar.
For instance, look at what the Baseball Savant has done, analyzing the hottest starting pitchers entering the postseason. This sort of analysis is more likely to yield something useful.
Let the real games begin.
References and Resources
The winning percentages and postseason records were calculated with data from the Retrosheet Game Logs, the Lahman database, and Baseball Reference. For my historical research, I leaned heavily on my Sporting News Baseball Guides, Baseball's Pennant Races: A Graphic View by John Warner Davenport and Leonard Koppett's Concise History of Major League Baseball.
Let me also add the following caveats to this analysis:
- Teams don't play the same schedule, so their won/loss records aren't directly comparable. I used winning percentages here as a proxy, but it can certainly be improved upon.
- September records may not be the best definition of "momentum." But I can't think of a better one.
- It may be that "momentum", when combined with other factors, does have an impact on postseason play. That would be another study for another day.
Dave was called a "national treasure" by Rob Neyer. Seriously. Comments about this article can be sent to him through the miracle of e-mail.
<< Return to Article