Monday, 24 June 2013

On Win Streaks

So Annie Edison is a Jays fan, or a bandwagon jumper.


Jeff Sullivan at Fangraphs had a look at 11+ game winning streaks since 1970, and as Jeff is wont to do, he did a great job at writing something that was kind of funny and pretty informative and all that good stuff that makes the things that people read worth reading.

I had a similar idea that I had been sitting on, but then I forgot until I read that, but now that I've read that, I kind of want to expand on it in my own little way.

More or less, what I was going to do was really just look at teams that had done well over the last few years, and find their longest streaks of the year and go from there or something.  I don't think we would have learned a whole lot though-- we all know that streaks make up such a small portion of a season, and even a scattered loss sandwiched in between 10 wins is almost as good as 11 straight.  Winning 11 straight and then playing .500 ball for the rest of the month isn't really all that different than, say, winning 3 and then losing 1 over and over for a month.  The streak itself doesn't matter so much as just losing less in general over a large sample.

Jeff sorta points that out in his 2004 Tampa example there, I guess, but yeah.  The thing I'm really curious about is the mediocre teams from the last few years.  Good teams are going to go on big rushes of wins, and bad teams are less likely to do so.  Jeff's point is very valid, in that we shouldn't ignore the Jays' poor start just because there's a hot-streak going on here, but stashed within that point is that the Jays might very well be an average-or-so team who just happen to have won 11 straight.  I dunno.  I mean, everyone picked them to win the division before the year, which, granted, is super common to say of the team who has the most active offseason (Miami in 2012 anyone?), and they were projected to get 95 wins or whatever.

In 2012, there weren't a whole lot of teams that were mediocre; teams seemed to really only be good or terrible.  Probably had a lot to do with mediocre teams punting the post-deadline games, but who knows, really?

The White Sox won 85 games, including 9 in a row beginning May 23rd (or 13 of 15 starting May 17) and 9 of 12 starting .  They also lost 10 of their last 14 to finish the season and blow any chance of catching the Tigers.

The Phillies were 81-81, which is probably the definition of mediocre.  Perhaps I need to brush up on my English though.  Either way, they were pretty bad in the first half of the year, and traded away a few pieces, only to play well down the stretch and finish up at .500.  Their last 42 games included 26 wins (.619 win%), but they never won or lost more than 6 in a row at any point. The problem was that they had multiple stretches of losing 4 of 5, 5 of 6, or 7 of 10.

Milwaukee finished with 83 wins, with a huge streak coming down the stretch.  The low point of their year was a 1-9 stretch beginning July 21.  On July 27th, they traded their ace, Zack Greinke, to the Angels for prospects, one of whom being starting SS Jean Segura.  They finished the year playing 34-20 (.629), without ever winning more than 6 in a row.  That doesn't make any sense.

Pittsburgh embarrassed themselves over the last month or so of 2012, winning 7 of their last 23.  They had separate losing streaks of 4, 5, and 7 in September alone.  They also won 15 of 21 heading in to the all-star break.  They didn't make any big trades or anything-- Acquiring Wandy Rodriguez at the deadline was the biggest one-- so the fall from grace wasn't really expected or anything, they just started sucking or something.

Arizona is probably the wackiest case in this study.  Just look.  Win 3, lose 5, win 6, lose 6 of 7.  Crazy that they were never any farther than 4 games above or below .500 after May 16th.

So our sample of 5 cherry-picked teams kind of leads us to believe that, yeah, teams can streak a little.  As is often the case when I try to do something like this, I don't have access to a large enough database to make any worthwhile conclusions here, which largely defeats the purpose of doing this at all.  I know this already.  Going through individual seasons on baseball-reference is tedious.

I dunno, that's pretty much it.  I can't really prove it empirically or anything, and maybe I'm biased, but I think Jeff is downplaying the usefulness of an 11-gamer when evaluating teams and that's really just my opinion and not much else.  Nothing against Jeff; he's smarter than me.  He includes a table that illustrates that a significant percentage of teams that have had 11-game win streaks, from a historical standpoint, weren't <.500 teams.  I doubt many started the year as shittily as the Jays did this year, mind you, but he's the one with access to the data.

No comments:

Post a Comment