 |  |  | ||||||||||
|
|  | | |||||||||
 |  |  | ||||||||||
 | | | ||
| ESPN.com NFL | |  | |
| | | ||
Wednesday, August 16
Blowouts and the value of closers
Special to ESPN.com
Editor's note: The team of writers from the Baseball Prospectus (tm) will be writing twice a week for ESPN.com. You can check out more of their work at their website at baseballprospectus.com.
With blowouts of a different type in the news, it seems to be an opportune time to talk about blowouts in baseball. The offensive explosion of the last eight years, while decried in some quarters, is defended by others as a good thing since it supposedly generates more fan interest, because fans like more scoring and more home runs. But judging by the number of fans leaving in the seventh inning of a 12-1 game I recently attended, they like games to be close as well.
Has the increased offense of recent years resulted in more 12-1 yawners, or have yesterday's 2-1 pitchers' duels become 9-8 hitters' battles? To investigate this question, I looked at all games over a 21-year-span, from 1979 to 1999. The data was made available thanks a combination of sources: Retrosheet, the Baseball Workshop and Total Sports among them.
Taking a cue from the definition of a save, we'll consider any game that ends with a three-run or less margin of victory as "close." We'll double that threshold and call any game won by six or more runs a "blowout." Let's look at the frequency of close games and blowouts over the years:
Year Close Blowout 1979 65.5% 17.0% 1980 65.8% 14.3% 1981 63.6% 14.3% 1982 65.6% 15.8% 1983 65.6% 15.7% 1984 65.0% 14.8% 1985 63.2% 17.3% 1986 64.7% 16.4% 1987 61.8% 18.8% 1988 65.0% 15.8% 1989 63.4% 16.5% 1990 63.0% 16.4% 1991 64.6% 15.7% 1992 66.1% 15.4% 1993 63.3% 16.8% 1994 61.1% 20.0% 1995 60.5% 20.3% 1996 60.8% 20.6% 1997 61.2% 20.0% 1998 62.0% 18.8% 1999 59.3% 22.2% Avg. 63.4% 17.4%The past six seasons, along with the rabbit-ball year of 1987, are the seven years with the highest percentage of blowouts since 1979. Similarly, the number of close games has declined by five to six percentage points. Perhaps this shouldn't be too surprising -- we'd expect the highest scoring seasons to have the most blowouts -- but it confirms that the rising offensive levels are turning 4-2 games into, say, 12-4 blowouts more often than into 9-7 games. The percentage differences aren't small either. The gap between the high and low points in the above chart means that teams in 1999 were playing in 10 to 11 fewer "close" games, and were involved in 12 to 13 more blowouts, than they were in previous decades. Allowing for some overlap between the two, teams are seeing a tenth of the season pushed into decidedly less interesting categories of games. I wonder what that's costing teams in late-inning concession sales. Of course, the final outcome of the game may not reflect the actual level of tension during the game. A rally in the ninth inning could turn a one-run squeaker into a foregone conclusion. So to account for that possibility, let's look at the run differential between the two teams at any point during the last three innings of the game. We'll again use the three-run margin for close games, and a six-run gap as a blowout:
Year Close Late Early Blowout 1979 79.4% 9.2% 1980 80.7% 7.7% 1981 79.7% 7.7% 1982 79.5% 7.2% 1983 80.3% 7.2% 1984 79.6% 8.1% 1985 78.8% 8.4% 1986 79.6% 8.0% 1987 77.1% 9.9% 1988 79.6% 7.9% 1989 78.9% 8.6% 1990 78.9% 7.8% 1991 79.2% 8.3% 1992 80.4% 7.0% 1993 79.5% 7.7% 1994 76.6% 10.3% 1995 76.0% 10.6% 1996 75.9% 10.3% 1997 77.4% 10.0% 1998 79.1% 8.9% 1999 76.2% 10.2% Average 78.7% 8.6%The percentage of games that are close at any point in the last third of the games increases, which makes sense since the previous category is a subset of the games we're looking at now. Still, we again see a decline in games that are still close in the late innings and an increase in early blowouts. We're still looking at an increase of about one game in 14 where the differential between teams late in the game makes the outcome seem like a foregone conclusion. Is baseball seven percent less interesting nowadays (Hmmm ... perhaps that's introducing too many numbers ... ) The increase in blowouts is not just a matter of entertainment. It can affect in-season decisions as well, particularly in bullpen management. With more and more games out of hand before you can get to your closer, the number of opportunities to use a top reliever in high-leverage situations decreases. This is counterbalanced, to some degree, by what started this discussion -- offensive levels are rising. Teams are scoring more runs, hence effective run prevention is more valuable and harder to come by. Does the value of run prevention in the late innings balance the lack of opportunities for a closer? Are closers worth more or less in a high-offense environment? Let's consider an idealized closer: one who is capable of shutting down the opposing team completely for one inning, and is brought in solely when the team is winning in the ninth inning. In a previous article I wrote for the Baseball Prospectus website, I developed a formula for estimating how often a team of differing offensive strengths would score 0, 1, 2, 3, etc. runs per inning. I've used this formula to figure out how much impact our ideal closer could have in different offensive environments. Assume two teams with equally potent offensive lineups (they average the same number of runs per game), as we move them into different offensive environments. The odds of one team overcoming a one-, two- or three-run deficit in the ninth inning varies as the offensive level ranges between three and six runs per game:
Chance of blowing a:
1-run lead 2-run lead 3-run lead
3.0 R/G 12.2% 4.6% 1.7%
3.5 R/G 13.6% 5.5% 2.2%
4.0 R/G 14.9% 6.4% 2.7%
4.5 R/G 16.2% 7.3% 3.2%
5.0 R/G 17.5% 8.4% 3.9%
5.5 R/G 18.4% 9.0% 4.4%
6.0 R/G 19.4% 9.7% 4.9%
So, the likelihood of a comeback increases quite a bit as offense levels rise. Naturally, this makes each appearance by the closer in a tight game more valuable, since the odds of the opposing team coming back are greater. The heightened likelihood may make the game more interesting, should it be close enough in the late innings.
Measuring subjective perceptions like this is probably too difficult, but we can reasonably hypothesize that it's related to the chance that the outcome of the game will be reversed in the late innings. We could consider the value of a closer as a proxy for how interesting a game is to a typical fan.
Therefore, the impact of a closer increases as the likelihood of a comeback increases (which, as we have seen, happens when offense increases), and decreases with the chance that the game will be close in the late innings (which decreases as scoring increases). These factors work against each other but don't completely cancel out.
Multiplying these two components together yields a factor proportional to the expected value of a closer. Comparing this factor to the overall offensive level, we find a slight overall increase in the value of a closer as scoring increases. In fact, between the lowest-scoring season and the highest-scoring season, the value of a closer has risen roughly 13 percent (assuming they were to be used in the idealized way described above -- we aren't considering the evolution of the closer strategy over time). The "shape" of that contribution is different, however, as it comes into play in fewer games but has more impact on those games.
We have theoretical support, then, for what common sense may have already told us: that thanks to rising offensive levels, fewer games will be in question in the late innings, but those that do will be real nail-biters. And can anyone who's watched the Mariners' bullpen in recent years claim otherwise?
Keith Woolner writes for the Baseball Prospectus, the annual book by the same name, covering over 1500 players with in-depth statistical analysis and hard-hitting commentary. Keith may be reached at kwoolner@baseballprospectus.com.
 |  |  | ||
|
ESPN INSIDER
Copyright 1995-2000 ESPN/Starwave Partners d/b/a ESPN Internet Ventures. All rights reserved. Do not duplicate or redistribute in any form. ESPN.com Privacy Policy. Use of this site signifies your agreement to the Terms of Service. | | | |
 | |  | ||