Cloud Cover for the Rays
Last night the Rays lost the evening portion of a noon-dusk doubleheader. Not only did this put them at a somber 0-4 on the weekend and 0-8 on the week, but the loss actually marked the eleventh in a row. That is the longest such streak in the major leagues this season and comes from arguably one of its better squads.
Now, this team isn’t quite the one from a month ago. Carlos Pena broke a pair of fingers on Labor Day, sidelining him for the remainder of the season. In his place is Chris Richard. The rotation is featuring three pitchers in their first full pro years, including Wade Davis who went from masterful start to atrocious start in the span of a week. The bullpen has imploded on itself and the offense has chosen a poor occasion to mail at-bats in.
Losing 11 in a row is a pretty implausible for any team. Take a 50 win (.309) team and ignore strength of schedule as well as home field advantage. Just assume they have a 69.1% chance at losing each game. Over a course of 11 games that ragtag group will have about a 1.7% chance at losing all 11 in a row. That’s for a 112 loss team. The Rays are considerably better than that. To avoid argument, let’s call them a .500 true talent unit – I don’t necessarily believe that to be the case, but it’s not overly important whether you think they are a 81 or 88 win team here – and rerun the binomial distribution.
The chance of them losing 11 in a row is something like 0.05%. Yes, that’s already in percentage form, not decimal. The accompanying graph looks like this, with the win total on the x-axis and the likelihood of at least that win total occurring on the y-axis:
The Rays had similarly microscopic playoff chances entering this span of games. They pulled a statistical oddity, unfortunately for them, it’s not the opposite of what they needed.
28
So when the Rockies went on that lossless tear at the end of 2007, we all wondered where those wins came from. Now we know: in exchange for their souls, the devil sold them wins borrowed from the 2009 Rays. (You knew they were going to have to pay something for taking his name off the team, right?)
The chances of a true talent .500 team losing all of 11 of a particular 11 game stretch are very low. The chances that a .500 team will lose 11 in a row at some point are much higher.
Exactly. And those chances are even higher when you are playing three of the strongest teams in the league over the 11 games.
Low probability events happen all of the time. For example, last night Reid Brignac got 4 hits in 4 plate appearances. If we assume that he will get hits in 1/4 of all plate appearances, the odds of this happening are 0.39% (ignoring correlation, similar to the odds computed above). But while there is a low likelihood of Brignac getting 4 hits in a given game, there is a good chance that some player will do so on a given night. Similarly, while there is a low chance that the Rays will lose a given stretch of 11 games, there is a high chance that some team will at some point lose 11 in a row. And when that happens, everyone will write “what are the odds…”
Yes. There is a 99.95% chance that a .500 team will NOT lose a particular 11 in a row. However, a 162-game schedule has 152 possible stretches of 11 games where that losing streak could happen. You’d have to dodge not losing 11-straight 152 consecutive times (.9995)^152. You’re chances of doing that are only 92.9%.
In a league of 30 teams, it’s unlikely they all get out unscathed.
Nice inverted binomial distribution CDF you have there.
Well, what were the chances of them getting to the WS last year? Doesn’t this streak/year kinda equal everything out?
Why, yes. Yes, it does.
What b_rider said. And just for fun, that important statistical point is well described in this excellent episode of Radiolab:
http://blogs.wnyc.org/radiolab/2009/06/15/stochasticity/
And the funny part is, had they won 11 in a row (which of course is more likely), they’d be right back in the race.
Based off the assumption that they are a true .500 team both are equally likely.
“They pulled a statistical oddity, unfortunately for them, it’s not the opposite of what they needed.”
I thought it WAS the opposite of what they needed… just saying.
I remember that back in 2004, the Rays pulled off a unique streak… a bad team with a 12-game winning streak. I remember thinking at the time how odd it was for a lousy team to win a dozen games in a row. They ripped off those wins in June… five wins against very good Giants and Padres teams, four against lousy Rockies and Jays squads, and three against the abysmal D-Backs of that season.
I seem to recall that the Rays owned the longest winning streak that season and I would wager it is rare that a losing team — much less a 70-91 team — finishes with that distinction in a MLB season. I would not be surprised if they own the all-time record for best streak for a 70-win team in a full season.
Slightly less fluky, that same (Devil) Rays team also pulled off a 12-game losing streak later in the season. What an odd year for that team.