I'm really hoping you guys can help me figure this one out.
As I wrote Monday, before this season only six teams in the last 20 years opened their season with six straight wins. Five of those teams ultimately made the playoffs. Granted, that's a small sample. But that success rate did strike me as moderately significant, at least. And thus relevant in discussions of this season's Royals (now 7-0) and Tigers (6-0).
Meanwhile, the Twins and Marlins both started the season 1-6: the Panic Number, as Jeff Passan dubbed it. According to Passan's research: In the last 31 seasons, 56 teams have started the season 1-6 or worse, and only two of those 56 teams made the playoffs.
So it does seem that the first six or seven games of the season really are predictive, right? I mean, 2 for 56.
Ah, but here comes Joe Sheehan in his newsletter this morning:
Projecting from that 7-0 is where you lose me. First of all, analysis that treats the first seven games of the season as a thing separate from other seven-game stretches isn't grounded in reality.
--snip--
Set the "when" aside, and just consider the fact of a seven-game winning streak. Knowing nothing else, does a seven-game winning streak hold signature significance? Not really. Last year, 11 teams had streaks of at least seven wins; of those, four made the postseason. Just five finished above .500. The Rangers won seven in a row…and went 67-95 overall. The Rays had a nine-game winning streak on their way to 77-85. As many last-place teams (two) had seven-game winning streaks as first-place teams did.
Last year was actually a light year for seven-game streaks. In 2013, 17 teams ripped off at least seven in a row. In 2012, 13 teams did. Roughly speaking, in recent years about half the league has a winning streak of at least seven games, and many of those teams are unimpressive. Two years ago, the Blue Jays had an 11-game winning streak on their way to 74-88, and three other teams won at least seven in a row while failing to reach .500. A seven-game winning streak doesn't tell us much about a team that we didn't know from other information. Even bad teams -- the 2012 Marlins or the 2011 Twins or the 2010 Indians -- can have a great week.
Actually, what I found is that while bad teams can have good weeks, it's pretty uncommon. Anyway, that's a lot of numbers for you. What I'm looking for is some help from someone (or someones) better at maths than I ...
My guess is that Joe's wrong about this, because while a seven-game streak in July might tell us nothing more than a seven-game streak in April would, a) seven-game streaks at any point in the season are somewhat more likely for a good team than a poor one, and b) while one-seven game streak might not mean a great deal, streaking at the start of the season leaves a great deal of room for more streaks.
So who's right here? Mathematically speaking, I mean.
