September 12, 2005
Playoff Odds Report, 1964
The Phillies' Collapse
The Playoff Odds Report is one of the most popular stat reports we've created here at BP. Now that we're deep enough into the season for the numbers to be truly meaningful, the report gets some of our highest traffic, and the data is quoted in newspapers around the country.
Because of its popularity, one of the most common requests I get is for Playoff Odds Reports for past seasons, particularly ones with interesting pennant races. It's not as easy as it sounds; the report runs through a million trials of the rest of the season, and doing that for, say, the last 30 days of a season takes considerable time. Still, the Cardinals' "virtual clinch" last week caused Rany Jazayerli to ask about the 1964 Phillies, whose ten-game losing streak down the stretch in 1964 cost them the NL pennant, and I admit I was curious as to what the report might say about them.
I pulled out the software I'd used to run the tests of the playoff odds earlier in the year, and adapted it to run day-by-day through September of 1964. It is based on 10,000 trials instead of a million. Here's the last three weeks of the season, with the odds of winning the pennant (there were no divisions in 1964). Each line represents a team's chances of winning the pennant.
Lost in the Phillies 0-10 swoon were some remarkable streaks by other teams. The Cardinals won eight straight games while the Phillies lost; the Reds had a 10-1 streak. The Yankees won 11 straight games between the 16th and 26th to take over the AL, including two doubleheader sweeps of the Indians on September 22 and 23. They needed to be that good, because their rivals were doing their best to keep up: the White Sox won their last nine games, while the Orioles went 7-1 in the last week, which kept the Yankees from clinching until the next-to-last day.
The Phillies' peak odds for winning actually came three days before the infamous losing streak started, as they hit 96.1% on the 17th of September. They lost their next two games, then won to get back to 96.0%. Then they started their losing streak. By the third loss they had ceded their status as most likely team to win a pennant to the AL-leading Yankees. After six games they had yielded the odds lead to the Reds, even though they were still a half-game ahead in the actual standings: the adjusted standings saw the Reds as a .563 team to the Phils' .534. The seventh lost game saw them drop behind the Cardinals in the odds race. By the tenth loss, they had dropped to less than 0.1%. By the final day of the season, the Reds and Cardinals were tied at 92-69 while the Phillies were a game back at 91-70. The Phillies (now rated at just .526 in the adjusted standings) played at Cincinnati (rated .561), while the Cardinals (.552) hosted the lowly Mets (.386). The Phillies' only hope was to for them to win (a .425 chance, using the log5 method and a .020 adjustment to each team for being home or away) while the Cardinals lost (a .319 chance); that would give them a one-third shot at the pennant, and .425 * .319 * .333 = .045, the Phillies' chances on the final day. The Phillies won their game, but so did the Cardinals, and that was that.