News Reports re Red Sox Victory Fail Daubert Scrutiny
Various news organizations are reporting today that the Boston Red Rox defeated the New York Yankees in the opening game of the American League championship series, by a score of 5-2. These reports are an example of how even respectable sports "journalists" can bamboozle the public by piecing together superficially plausible stories based on purely anecdotal data. Such "news" accounts may hold sentimental appeal, but a rigorous evaluation of the facts, in light of Daubert, demonstrates why these reports should be disregarded.
1. Has the theory in question been tested?
This factor, by itself, would be decisive. In numerous experimental trials, the hypothesis that the Red Sox can win an important post-season game has been repeatedly falsified. (See Buckner 1986.) Although scattered anecdotal reports of Red Sox victories in post-season play can be found in the literature, these are probably attributable to observational error. At best, they represent statistical outliers. For example, working from a sample of World Series games in which the Red Sox have participated in the past fifty years, the probability of a Red Sox victory in any given post-season game of importance would rise no higher than .429. Even that figure must be considered an over-optimistic appraisal, since the sample is intrinsically biased in favor of seasons in which the Red Sox possessed sufficient talent to claw their way to an American League pennant. Over the same period, meanwhile, the Yankees boast a .598 success rate in World Series play. Plainly, any sports writer contending that the Red Sox have won a significant post-season victory should have the burden of proof by a preponderance of the evidence. And the numbers don't lie. Based on any fair and neutral assessment of the historical and statistical data, it must be reckoned "likelier than not" that the Yankees won last night's game -- probably in a ninth inning comeback sparked by a Boston fielding error.
2. Has the theory been subjected to peer-review and publication?
Isolated reports in the popular press scarcely form a reliable basis for so implausible a conclusion as a Game 1 victory by the Red Sox in the Yankees' home park. The journalists claiming a Red Sox win offer no citations supporting their conclusions from respected peer-reviewed journals such as The New England Journal of Medicine, or Lancet.
3. What is the known error rate?
As any longstanding student of this problem is aware, the error rate is high. (See Buckner, op cit.)
4. Is the theory generally accepted?
Based on decades of study and analysis, the Curse-of-the-Bambino Model enjoys almost universal acceptance among experts in the field and has amply demonstrated its strength as a predictive heuristic. Any suggestion that the Red Sox wrested the home field advantage from the Yankees in the series opener, in a game where Pedro Martinez did not throw a single pitch, flies in the face of that settled paradigm.
1. Has the theory in question been tested?
This factor, by itself, would be decisive. In numerous experimental trials, the hypothesis that the Red Sox can win an important post-season game has been repeatedly falsified. (See Buckner 1986.) Although scattered anecdotal reports of Red Sox victories in post-season play can be found in the literature, these are probably attributable to observational error. At best, they represent statistical outliers. For example, working from a sample of World Series games in which the Red Sox have participated in the past fifty years, the probability of a Red Sox victory in any given post-season game of importance would rise no higher than .429. Even that figure must be considered an over-optimistic appraisal, since the sample is intrinsically biased in favor of seasons in which the Red Sox possessed sufficient talent to claw their way to an American League pennant. Over the same period, meanwhile, the Yankees boast a .598 success rate in World Series play. Plainly, any sports writer contending that the Red Sox have won a significant post-season victory should have the burden of proof by a preponderance of the evidence. And the numbers don't lie. Based on any fair and neutral assessment of the historical and statistical data, it must be reckoned "likelier than not" that the Yankees won last night's game -- probably in a ninth inning comeback sparked by a Boston fielding error.
2. Has the theory been subjected to peer-review and publication?
Isolated reports in the popular press scarcely form a reliable basis for so implausible a conclusion as a Game 1 victory by the Red Sox in the Yankees' home park. The journalists claiming a Red Sox win offer no citations supporting their conclusions from respected peer-reviewed journals such as The New England Journal of Medicine, or Lancet.
3. What is the known error rate?
As any longstanding student of this problem is aware, the error rate is high. (See Buckner, op cit.)
4. Is the theory generally accepted?
Based on decades of study and analysis, the Curse-of-the-Bambino Model enjoys almost universal acceptance among experts in the field and has amply demonstrated its strength as a predictive heuristic. Any suggestion that the Red Sox wrested the home field advantage from the Yankees in the series opener, in a game where Pedro Martinez did not throw a single pitch, flies in the face of that settled paradigm.
<< Home