Stiff Arm Trophy

A plea to sportswriters for statistical accuracy

Posted by Kari Chisholm on December 8, 2011 in News.

When comparing year-to-year, use the percentage of the maximum points available. (In other words, the closeness to a unanimous selection.)

Otherwise, you're not comparing apples-to-apples. In almost every single statistical category, you'll find that OJ Simpson leads. But it's not because he was an overwhelming Heisman winner. He was simply the biggest Heisman winner during one four-year stretch when the Heisman had over 1200 voters. Since 1988, the Heisman has standardized on 870 media voters and every living winner. (And even that total changes every year, of course.)

In 2011, there are 927 voters - 870 media, 56 living winners (excluding Reggie Bush), and 1 fan vote. There are a total of 2781 points available to each Heisman candidate. Simply divide the actual point total by 2781, and you'll have the number to use.

It makes zero sense to compare Heisman winners year-to-year based on raw point totals. Even if Griffin - or any modern-day winner got 100% of first-place votes, he wouldn't get more points than OJ Simpson's 2903. There were 1200 voters that year, 274 more than today. A fair comparison requires percentages. Comparing Heisman winners across years using raw point totals is arithmetic fraud

Here's the right comparisons for your coverage (with links to details):

(Wondering what the heck happened in 1952? That year, the Heisman voter pool was expanded dramatically from 987 to 1222 - and 162 different players across the country earned votes.)