Getting More out of your Analytics Program

During this offseason, there have been a lot of unprecedented happenings. A lot of

coaches have spent their time improving their scheme, working on recruiting, and learning more about the history of the game. I have tried doing all of that, but I have spent most of my time improving my analytical abilities. I want to use this article to share with you what I think you can implement in order to improve your analytics program. I also want to give a special thanks to Football Coaches Forum, as the last analytics article I wrote is a large part of what helped me in getting my new job as an analyst at the University of Akron.

First, I will dive into how to evaluate analytically, and what analytics can help you with

in football. It is obvious that analytics can help determine play calling and tendencies, given that quality control departments have been doing that for years. Analytics can be applied easily to the big picture. You can use it to determine efficiency, tendencies, and what play calls should be most effective for your team and against your opponent. Another type of analytics which I have been exploring is the analysis of individuals and position groups. There is one clear advantage of this and one clear disadvantage. The advantage is that looking at a player’s alignments and personal tendencies will give away what they are going to do much more often than schematic analysis. If you look at a position group, their tendencies will be mostly based on communication and alignment. You will normally see they have a particular sign they give each other which means they are running a stunt or anything out of the ordinary. The disadvantage is obviously that with a substitution, you have less or no information to apply at that point in the game to that particular position.

These analyses require a great deal of film study. I learned what to look for and how to

spot it last year while working with the IUP Offensive Line. The OL Coach would ask each

individual player in the unit meeting to identify a tendency of a particular player on opponent film. Humans are creatures of habit, and that is why reading player tendencies is very effective. If a defensive end tightens his stance 90% of the time he is long sticking, that is something the offensive linemen need to know and communicate. Another great example is the CAT (Corner) blitz. From all the film I have watched, the defensive back group has always had a tell when CAT was coming. There was often a safety rolling down to replace the corner, but the corner made it even more obvious. As I am sure most offensive coaches know, it is almost impossible for a corner to not tip his blitz by eyes or alignment. Typically, they will move laterally pre-snap knowing they have significant ground to cover, and they will almost always be looking straight through the outside gap and into the backfield. The thing I am starting to do is take these tendencies and analyze them. Therefore, we can provide our players with accurate information about the signs and tells which opposing players show.

The other type of analyses I use are schematic tendencies.These can be applied to

determine your own efficiency, what your play calls should be, the opponent’s play calls, and

how to counter them. The simplest way to exploit an opponent’s play calls is a universal check. When I first started analyzing, that is one of the first things I had to look for. These are so important to an offense because if you are able to see a defense checks to cover 2 every time they are facing 11 personnel 2x2, perhaps you should implement a Texas concept in order to exploit that. Universal checks are rarely this simple at a high level, but it is very important you look for them and determine what you can exploit. Another great advantage of schematic analysis is determining efficiency. This is because it will allow you to determine what the formula is for creating your best plays. If your top scheme is power, personnel is 21, and direction is right, perhaps those should all be used in tandem on must-convert plays. These can be cross referenced with defensive tendencies too. The biggest way I have been implementing analytics, however, is to create formulas to determine what will be done on fourth downs, kickoffs, PAT attempts, and need-to-convert downs. I determine our own expected points and what the average return (on investment) would be against the opponent. I then use the opponent’s statistics to determine where they will normally begin to make certain choices on the field, such as conversion attempts. This type of analysis is based on both film study and statistics, but it is not mutually exclusive from scheme. It is important that analysis work in tandem with scheme to create your best play calling.

The first step in your analysis is to determine what you are looking for. This means your

research must have a purpose in what you want to get out of it. That could be when in the game a team starts to use all four downs, the distance and field position where they will go for it on fourth down, or pretty much anything else you can imagine. You must then determine how you will use your findings to help you in game planning and on gameday.

The next step is gathering information. I think the most important thing in any gathering

of data is to exhaust all relevant factors. What I mean by that is if you are looking at an

opponent’s attempts to convert on fourth down, you must know the down and distance, time in the game, margin, and field position. All of the relevant factors must be included in order to make your study as accurate as possible.

If you are doing a study which requires you to use multiple unalike variables, I highly

recommend creating an index. If you are not comfortable doing this, you may simply sort them separately. However, the great thing about using an index is that it allows you to sort unalike variables together. I know it may seem complicated, but I would recommend doing some reading on statistical analysis if you are not familiar with it. Often, standard deviations can be used to determine how a team compares to other teams in factors like turnovers, yardage, and touchdowns. While these seem impossible to sort together, creating your own index will allow you to do so. It is important also to use established principles. If there is ever an accurate and useful index which has already been proven, perhaps that is the one you should be using. While a lot of analytics is finding your own ways to determine what will work best for your team, many indexes are already out there to help you make that determination.

The final step is applying the information. First, take the information and determine how

you will apply it. This allows you to extrapolate your findings and use them in your game

planning and play calling. Next, if you are comparing yourself to other teams, capitalize on your strengths and close the gaps on your weaknesses. This also includes your play calling. If a play is terrible, maybe it is time to take it out of the game plan. If a play is okay, use your findings to make it better. If a play is great, continue to capitalize by using your findings to make it even better and more efficient. The biggest point about applying the information is understanding how it can be applied on gameday. In my last article I discussed the study I did to determine what field position and down and distance were best for punts and field goals, and which were best for conversion attempts. Since then, I put together a chart where I made it possible to make the determination in-game without having to solve the entire formula. There are only a few seconds which can be used for determining whether to send out your offense or punt team, so it is crucial you apply these data in a useful manner. My chart stated based on the amount of expected points, what decision should be made based on the distance of the line to gain. If you want to simplify it even more, put the actual yard lines of your expected points in order to determine what situations favor going for it. I highlighted all situations on the table in one of three colors. Red was for the

situations which had a negative expected outcome, yellow was for less than one point, and green was for one point or more. Therefore, if the situation we are in is located in the green area of the chart, it favors our expected points by over one point to go for it.

Thank you for your positive responses to my last few articles. I hope to continue writing

these to help improve analytics departments everywhere and help small schools as they begin to implement them.

Joshua DePasquale