Playing Up To The Level Of Competition (Fun Stat Nerd Tidbit: Florida State)

UPDATE: Strategies, Tactics And The Florida State Seminoles is now up at the mothership.

Like chess, football usually boils down to strategies and tactics in the end (unless, as is often the case, one team's talent and athleticism is too great for the other, like they're starting with more pieces on the board). Who you decide to place on the field, and where you place them, constitutes strategy. 4-3 defense? 3-4? 3-3-5? Spread offense? Pro-style? Wishbone? Your strategies are your "base, core plays," your approach, your mindset. Tactics, then, are what you actually do once you have established your strategy: Who do you use to blitz and when? When (and how often) do you call your constraint plays to complement your base plays? You put yourself in position to win with strategy; you win with tactics.

----------

In today's upcoming Florida State piece, I say the following:

Against lesser opponents, teams Florida State could physically dominate, the defense came up big. They allowed just 16.9 Adj. PPG versus the seven lowest-ranked teams on the schedule. The seven highest-ranked teams, on the other hand, averaged 26.0. Now, to some degree it is supposed to work that way (Good teams do better than bad teams? Rocket science!), but FSU's good-to-bad range was a bit larger than most teams'. The defense was young enough that they still have plenty of time to season and mature, but with the early schedule they're facing in 2011 (Oklahoma, @Clemson in Games 3-4), the sooner that maturation occurs, the better.

This claim led me to tackling a topic I'd been intending to tackle for a while: covariance.

The simple 'math' definition of covariance is as follows:

Covariance provides a measure of the strength of the correlation between two or more sets of random variates.

We look at correlations a lot with football stats, and for obvious reasons. It allows us to figure out when our eyes are lying to us, it tells us how seriously we should be taking certain factors (returning talent, recruiting, etc.) when making projections, etc. We're going to use covariance to look at which teams tended to play well against good teams (play up to their level of competition) and which ones tended to only play well against bad ones. Were Florida State's defensive tendencies mentioned above rare? Let's see.

Some notes about the table below:

• For the covariance tables below, we're using correlations between a team's single-game Adj. Point total and the F/+ ranking of the given opponent.
• Teams ranked highest on the list below were the ones who played great against bad teams and less great against good teams. Teams ranked lowest tended to play well against good teams and less well against bad teams.
• You'll see that highly-ranked offensive numbers are positive and highly-ranked defensive numbers are negative.
• Keep in mind one thing: small sample size is an issue here. With just 12-14 games, one strong or weak performance can significantly skew your totals.

The verdict? Florida State was certainly on the higher end of the "played better against bad teams" spectrum, but only in the Top 25, not the Top 10. Evidently I should have either used this for Utah's profile a while back or saved it for Georgia's (whose defensive covariance was CRAZY high).

What does this data mean? Honestly, I'm not entirely sure yet. I'll revisit the topic, possibly for Georgia's profile, to see what kinds of ties I discover between covariance and projections/predictions. I'd love to be able to say that if you were strong on one end of the spectrum, you're more likely to improve/regress the next year ... or once I have multiple years of this data, see if certain coaches have certain tendencies in this regard ... but we'll see. For now it's just something interesting; we'll see if there's more to it than that.

Alright, here's the list:

Full list after the jump.