This is our old blog. It hasn't been active since 2011. Please see the link above for our current blog or click the logo above to see all of the great data and content on this site.

Blowing 9th-inning leads

Posted by Andy on November 3, 2007

Yesterday, I wrote about the PI's Team Inning Summary function, and here's another thing you can do with it.

For each year, I looked at run scoring for all teams (in total) and extracted the total number of games that a team had a lead going into the 9th inning and then ended up behind at the end of the 9th inning (i.e. losing the game.) Click through for detailed results.

For starters, here is some of the data shown in numbers:

Year    9th inning leads   9th inning losses    %
2007    2210               55                   2.5%
2006    2218               75                   3.4%
2005    2193               59                   2.7%
2004    2188               64                   2.9%
2003    2196               66                   3.0%

That data is pretty self-explanatory, but keep in mind that this is simply 9th-inning leads turning into 9th-inning losses. It does not include, for example, 9th-inning leads the turn into tie extra-inning games that eventually become losses. This data describes only 9th-inning behavior.

So as you can see, in 2007 about 2.5% of all 9th-inning leads turned into 9th-inning losses. (By the way, another 4.1% of those 9th inning leads were blown leading to ties at the end of 9 innings, meaning a total of about 6.6% 9th-inning leads were blown altogether, although not all of those games became losses.)

Well here is all the data since 1957, in graph format:

The red dots are the year-by-year percentage of 9th-inning leads that turned into 9th-inning losses, whereas the black line is the average percentage for the given year plus the 4 previous years. As you can see, the fraction of 9th-inning leads turning to losses has been declining over time, despite some noise up and down from year to year.

Predominantly, this probably has mainly to do with the emergence of the closer and other specialty relievers. These percentages may not seem very significant, but consider this: in 2007, a total of 55 such lead-blows occurred, spread among the 30 teams. The better teams suffered few of these reversals. (Boston, for example, entered the 9th with a lead 91 times this year and kept that 9th-inning lead 90 times.) The worse teams suffered more of these reversals. (Tampa Bay, for example, entered the 8th with a lead 59 times this year and kept that 9th-inning lead only 54 times.) That represents 4 or so games in the standings right there. Not that 4 more games would have put the Devil Rays into the playoffs, but this is one of the many factors that separates the good teams from the bad.

For more on this type of thing, you can read a personal post here.

2 Responses to “Blowing 9th-inning leads”

  1. vonhayes Says:

    What happened in '94? Did the smaller sample size throw everything off?

  2. Andy Says:

    Well, there were many fewer overall games in '94 where a team entered the 9th inning with the lead (1437 as compared to 2045 in '93) and 66 of those leads turned into losses. That yields 4.6%. I don't know why there would have been a higher proportion of blown leads in that year, though. It's worth mentioning that the running average line doesn't simply average the yearly percentages, but rather takes the total number of game situations and calculates an overall percentage. Thus, the 1994 data point has lesser impact than the other surrounding data points.