Dynasty, in Theory: Imagining Profiles In More Dimensions

Photo: Gregory Fisher, USA Today Sports

Join the Footballguys Daily Update

Start your morning with our roundup of the most important stories in football - with the fantasy insight you need to make league-winning decisions. Delivered straight to your inbox, 100% free.

Subscribe Now

There's a lot of really strong dynasty analysis out there, especially when compared to five or ten years ago. But most of it is so dang practical-- Player X is undervalued, Player Y's workload is troubling, the market at this position is irrational, and take this specific action to win your league. Dynasty, in Theory is meant as a corrective, offering insights and takeaways into the strategic and structural nature of the game that might not lead to an immediate benefit but which should help us become better players over time.

A New Paradigm For Your Dynasty Teams

If you've been reading this column for a while, you know how much I love writing about paradigms, or "frameworks for considering things." Paradigms are the quintessential topic for Dynasty, in Theory because they don't change a thing about our reality; they merely change how we relate to it. You can believe that seasons are only successful if you win a championship, or you can believe that any season is successful as long as you had fun, but either paradigm you choose won't alter the fact that you missed the playoffs.

Last year, a few in football analytics started working on evaluating offensive and defensive identity using something called n-dimensional representative ideaspace. I've found it a useful framework for evaluating... all manner of things, really. So today I wanted to walk through the concept, how it works, and how we can apply it to our dynasty teams (and more).

What is N-Dimensional Representative Ideaspace?

Let's imagine we charted every NFL defense on two factors: what percentage of the time they called a blitz and what percentage of the time they played zone defense. Both of these values exist on a scale, so we could graph them against each other. In the top-right quadrant would be teams that called lots of blitzes and played lots of zone coverage. In the bottom right would be teams that blitzed and played man. The bottom left would be teams that primarily rushed 4 and played man, while the top left would be teams that rushed 4 and played zone.

A crucial point is that neither axis deals with how good a team is at what it is doing; it merely describes how often it does it. You could (and likely would) have good and bad teams in all four quadrants.

This would be an interesting two-dimensional breakdown of defenses, but no defense can be perfectly described with just two dimensions. Let's say we wanted to track a third factor, such as how often they played 4 defensive backs vs. 5 defensive backs. We could imagine making the last graph three-dimensional, with teams that play a lot of nickel defense moving higher up and teams that play less nickel defense moving down.

Let's say we wanted to add another factor, perhaps how often the team called a zone blitz (where a player who typically drops into coverage will instead rush the passer while a player who typically rushes the passer drops back into coverage). And another factor, such as how far the linebackers line up relative to the line of scrimmage. And another factor, and another factor, and another factor.

If we have N different factors that we are considering, the range of possible locations a defense could fall would encompass the n-dimensional representative ideaspace. Defenses with similar tendencies will be located close to each other in space, while defenses with divergent philosophies will be located far apart. And if you squint hard enough, way off in the distance, you might even be able to spot Brian Flores' Minnesota Vikings.

Vikings defensive scheme superlatives:

Most Cover 0 (by far)
Least Cover 1
Most Cover 2 (by far)
2nd most rotations to open
Lowest avg. LB depth
Most avg. players lined up at the line
Highest blitz % (by far)
Most drop 8 coverage (by far)

— Timo Riske (@PFF_Moo) December 22, 2023

When clustering all defenses of the last 5 years (i.e. 160 defenses total) based on schematic indicators, the 2023 Vikings defense is the furthest away from all other 159 defenses.

In other words, the most unique defense of the last 5 years.

— Timo Riske (@PFF_Moo) December 22, 2023

Last year, the Vikings ranked 26th in yards allowed; this year, they rank 9th. Last year, Minnesota gave up 2.9 more points per game than its opponents averaged against the rest of their schedule; this year, they've allowed 2.5 fewer. Minnesota is running the most unique defense in the league, and the results have been dramatic. On the other side of the ball, if you envisioned all NFL offenses in n-dimensional representative ideaspace, the San Francisco 49ers and Miami Dolphins would likely stand apart from the rest of the league.

Is It Good To Be Different?

From the examples above, you'd assume the answer would be "of course". But remember, none of the dimensions here represent "good play" or "bad play"; they describe scheme, not execution.

if you think about what this represents for a second, it gets a lot more complicated. The ideaspace is, more or less, the boundary of all possible schemes a team could run. If teams tend to cluster in a specific region of that space, there's probably a good reason. A section of that space will represent teams running West Coast concepts on offense, and that section will be very heavily populated because the West Coast Offense is one of the best and most effective offenses ever devised.

On the other hand, the vast majority of potential schemes a team could use are objectively moronic. The sector of the ideaspace representing "put five fullbacks on the field and have them all run go routes on every single down" is completely unpopulated... because it's a terrible idea and any team that tried it would get humiliated. Almost all unclaimed territory in n-dimensional representative ideaspace is terrible.