In the wake of four consecutive close games including; two gut-wrenching losses, one sigh of relief and one steal, I was wondering how the Noles have fared throughout the years in these type games. I can recall with great clarity EVERY close loss to the Canes from the late 80’s through ’09 and many over the last decade to...just about anybody. However, I’ve had a tough time recollecting many that we’ve actually won. So, that inspired me to do a little research to find out.
I took a look at the past 25 years because it coincides neatly with the rise and fall of the Seminole football empire and it’s a convenient time frame for recent historical reference. First I had to define the criteria for "close games". I chose a final score margin of eight points or less and assumed that over that many years the games that weren’t actually that close would counterbalance the ones that were, but had a higher final margin.
From 1985-2009 the Noles played in 82 such games, winning 43 and either losing or tying 39. That’s a winning percentage of 52%. So, the question is…how good/bad/mediocre is that?
To find out I could do one of two things; get the records of every other Div. IA team or take a sample, apply some statistics and make an educated guess. I’m not a statistician, but since I’m using standard analysis methods I’m willingly opening myself up here for crucifixion from the resident stat heavy folk.
My variable in this analysis is the close game winning percentages with the following assumptions and rationale:
1. A normal distribution when considering all 120 teams with an expected mean of 0.50. Since the vast majority of games are played against other IA teams…for every win there is a loss in the population.
2. By analyzing a sufficiently large sample population (at least 30) I could estimate the true mean and variance of the entire population, thereby saving the trouble of evaluating every team.
3. Since I’m only interested in win %, the actual mean should be less than 50% since ties count against both teams involved.
4. With all 120 listed alphabetically and numbered 1-120 I used a random number generator to pick the 30 for analysis…to limit any bias.
Here are the results:
The 0.48 mean is pretty much as expected. However, the negative skew indicates it’s not exactly a standard normal distribution (bell curve). As the sample population shows, Vandy and Duke are the big culprits in the skew (no corresponding teams in the high 60’s% to balance).
Even though my assumptions aren’t entirely valid we can still get a pretty good indication of where the Noles rank using a normal distribution. Therefore, using standard z-table, if you randomly selected any team there is ~ a 73% chance that team would have a winning percentage less than 0.52. In other words FSU has been better than approximately 3/4ths of all division IA teams in "close" games over the past quarter century.
Breakdown of FSU's games decided by 8 points or less
Decide for yourself what to make of this. If you weren't around for the dynasty years I'll just say there were only a handful of teams that could regularly field talent comparable to FSU.....