I was paging through the 2011 NC State football media guide and found this, and in the course of trying to figure out just what the hell that's all about, I ended up falling down a rabbit hole. (The mysterious media guide stat is apparently the sum of the number of turnovers forced and the number of throwaways by opponents to avoid a sack, if you were wondering. So "forces" is not a typo; it's "turnovers [and] forces.")
I eventually wound up in the NCAA record books, where there's a section on trends through the years, including an interesting bit on how often teams have elected to go for two since that rule was installed in 1958. It's July, so what the hell. Here's a breakdown:
| Year/Decade | Kick% | 2PA% | Kick Success Rate | 2PA Success Rate | Avg Kick Pt Value | Avg 2PA Pt Value |
| 1958 | 48.6 | 51.4 | 68.6 | 44.7 | 0.686 | 0.894 |
| 1959 | 59.8 | 40.2 | 75.4 | 40.3 | 0.754 | 0.806 |
| '60s | 81.6 | 18.4 | 83.4 | 42.2 | 0.834 | 0.843 |
| '70s | 88.4 | 11.6 | 90.3 | 42.5 | 0.903 | 0.849 |
| '80s | 89.6 | 10.4 | 94.3 | 41.4 | 0.943 | 0.827 |
| '90s | 91.1 | 8.9 | 93.8 | 42.3 | 0.938 | 0.846 |
| '00s | 94.5 | 5.5 | 95.2 | 41.1 | 0.952 | 0.822 |
So in 1958, the first year teams could try for two, they did so more than half the time. Two-point tries were successful 44.7% of the time, while the average team made XP attempts 68.6% of the time. What's interesting to me is just how often coaches elected to try for the two-point conversion in those first two years with the new rule. Perhaps it was the "ooh, shiny" effect, or an example of the days when men were men and damned the torpedoes. Kicking was far less of a sure thing then, which no doubt was a contributing factor. An average team actually could expect to get more points out of a 2PA than it could out of an extra point attempt.
By the end of the 1960s, the ratios were much closer to those we see today. In 1959 the NCAA expanded the width of the goal posts from 18 feet, 6 inches to 23 feet, 4 inches. That made kicking the ball a considerably more reliable prospect, and accuracy climbed steadily through the '70s and '80s.
In 1963, XP accuracy crossed the 80% threshold and would never dip lower than that again. In 1974, accuracy hit 90% and kept climbing. Kickers were getting better and as success rates climbed year after year, the two-point try continued its slow disappearance. Improvements in offensive effectiveness and efficiency, which created a higher scoring environment, certainly had an impact as well.
The "modern era" of kicking didn't begin until 1993. Kicking from a tee became illegal in 1989, the width of the goal posts was narrowed back to 18-6 in 1991, and finally, the hashmarks were narrowed to their current width in 1993. (Though the latter has no relevance to XPs.) Even after the goal posts were narrowed, XPs remained a gimme proposition, made well over 90% of the time.
Nowadays extra points are attempted after 95% of all touchdowns and converted 95% of the time. It's a bit surprising to me that two-point conversion success rates have been so consistent through the years. I would've guessed that as offenses improved, and as two-point attempts became rarer, that figure would climb some. The low 2PA% in the '00s suggests that the surprise two-point try has essentially disappeared, and that teams now only go for two when absolutely necessary. So in terms of expectations and advance notice, things maybe haven't changed all that much for today's defenses. Not to mention that an opponent can always call timeout to ready its defense for a two-point try it did not see coming, though that's not always a sure thing, is it, Paul Pasqualoni?