Sporting Life Arkansas on the "Gamble"

Going for two leading by 13 is a bad decision because the stats say so. But do those stats really provide useful information?

SportingLifeArkansas has a worthy read on Coach Harsin's decision by J. Frank Parnell. I have to say it is refreshing to read a piece questioning a coach's decision that doesn't take a negative "how dumb can you be?" viewpoint that is unfortunately so common.

Parnell doesn't say anything here that many others haven't said (and I've got the text messages and board posts to prove it) about Coach Harsin's decision to go for two after A-State's second TD. I take that back. Distanced from the emotion of the game it is far more reasoned than those texts and posts.

Of course, had the Wolves kicked an extra point in the first quarter, they would have led, 35-34, early in the fourth quarter. It's just one man's opinion, but it's hard to figure why a coach would go for 2 leading 13-0 in the first quarter. The percentages are certainly against it.

This stream-of-consciousness review of Thursday's barnburner of a game isn't intended as a knock against Harsin's decision; I just found it odd. And I'd like to thank him for making me wonder because I dug up some unexpected facts about the 2-point conversion.


When it comes to statistics, I'm not in the camp of Coach Harsin's comment that "stats are for losers," but I certainly believe that too often stats are used improperly if not foolishly.

Stats require context to be understood and to have any useful meaning. Many years ago TCU came in touted as the seventh best rushing defense in the country. Digging deeper that impressive stat became less impressive once I realized their two opponents hadn't run successfully on anyone else. A-State had a big rushing day.

I don't trust some stats because the context skews the numbers.

Take the success rate of on-side kicks. Most on-side kicks are attempted late in a game by a team trailing against a team expecting the attempt with their "hands" team out there.

Since most on-side kicks are expected and the receiving team is fully prepared, stats about the success rate are not very useful in evaluating the decision of a coach to attempt an on-side kick when tied or leading in the second quarter.

Stats on two-point attempts are similarly skewed by the fact that most attempts come from a trailing team under pressure either playing to win on one play or hoping to tie or to reduce the number of times they have to score to be in position to tie or win.

Teams leading by six or more rarely attempt a two point conversion even though a team leading is generally either the better team or the hotter team at that point in the game and presumably more likely to be successful.

Fans and media are going to debate the wisdom of a failed attempt by a team leading by six or more but the stats on success rate probably aren't relevant enough to the situation to carry much weight in the discussion