Separating Strategy From the Execution on the Field

This is the third of a series of posts on decision-making and football.

A very interesting part of football is the way that the larger, more strategic decisions (which play to run, which players to have on the field at which times, whether to punt or go for it on fourth down, whether to kick the field goal or not) have all been moved off of the field itself. The decisions that have usually been moved off of the field are those that take searching through complex semantic information (the kind that is easily written on paper or stored in a computer).[1]

A coach up in the booth has lots of documents, computers, and data available to him. He can search through those possibilities while the players are executing actions on the field. In a sense, the coach in the booth is the Deliberative system, searching through possibilities, adjusting actions based on models of the world that are flexibly updated from ongoing information. (The cornerback is slower than usual today, our wide receiver seems to have a step on him. Our offensive line is holding the pocket open longer than usual, we’ve got time to execute longer pass plays. or The pocket keeps collapsing, we need faster, short plays to get ourselves moving.)

As noted in the first installment of this series, the Deliberative system works by understanding how the world works, having explicitly identifiable (language-based, declarative, semantic) information that one can use to predict how things will go. From this model of the world, one can work out the consequences of one’s choices. This takes time to do because one has to find that information in one’s memory (or one’s notes), to work out how that information bears on the decision at hand, and to work out how a sequence of plays can interact with each other.

The clearest example of Deliberative decision-making in football is the strategic planning that goes into designing specifc plays and packages for each opponent each week. Over the course of the week before a game, coaches will watch tape, try out possibilities, consider multiple match-ups and possibilities. They will (often literally) run simulations of what a specific package and play would likely be able to do against a specific opponent. Working out the consequences of these many different possibilties takes time, which is why the planning is done earlier, off-line, during the week before the game, when the coaches have the time to play out the possibilities.

By separating these decision processes (Procedural on the field, Deliberative in the booth), football has found the same kind of efficient advantages evolution did in our mammalian decision-making systems.

What about the hurry-up offense?

As we noted, these more deliberative decisions are slower and take time to find. So, sometimes, teams shift into a “hurry-up offense” which reduces that planning time.

The hurry-up offense has advantages and disadvantages. The main advantage, of course, is that it is faster. If a team is low on time (because the clock is ticking down), then there is no time to plan the optimal play. Sometimes it’s better to go with a pretty good play chosen quickly than the perfect play chosen slowly. In the decision-making literature, this is known as the speed-accuracy tradeoff. If it takes time to find the best option, there is a tradeoff between being fast and being right.

The hurry-up offense also prevents defensive substitutions, forcing defenses to play with player configurations that are not optimal for the offensive play, and it can even catch the defense off guard. An off-balance defense will be easier to beat. Of course, the offense also needs to line up quickly. An off-balance offense isn’t going to run the play well. (A key to procedural learning is that it depends on action-sequences. Starting the action-sequence in the wrong place can make it very hard to execute the right actions at the right times.)

The disadvantage of the hurry-up offense is that the play either needs to be decided before the players go on the field or there need to be a limited number of plays to search through. Generally, there are a limited number of plays available in the hurry-up offense. Usually, the sequence of plays is already decided before the game. When the quarterback is actually doing the play calling himself, there are only a few plays he is deciding between. Again, a quarterback deciding plays on the field is not deliberating over choices – he is using his categorize and act expertise to select the next play, which means the quarterback has to have had enough experience on the field with that team to know what generally works in those situations and what doesn’t.

[1] Interestingly, football is not the only sport that has separated these decision processes. Notice that in baseball, the pitcher is concerned with getting the action right (Procedural) but the catcher is the one calling the pitches, working a sequence of pitches based on extensive semantic knowledge of a specific batter’s tendencies (Deliberative).