We have always heard as players, fans and the audience for college and NFL games the advice or plan that the team with a lead should run out the clock by calling strictly running plays and hopefully get one, two or three first downs. This general theory avoids the risk of a turnover (pick or fumble due to a QB sack), it is claimed, as well as avoiding the risk of throwing an incomplete pass and stopping the clock. A variation of the general theory suggests that the offense protecting the lead in the final minutes of a game should run on first and second down and then throw a pass on third down if it is more than a couple yards away from a first down after the two runs. Much of this general theory and its variation are based upon the desire to force the opposing team to use up their times out to stop the clock from running.
Does the general theory and its variants make sense? Maybe, but it all depends on many factors, which are game and situation dependent. Teams have started to track data, plug in the known variables and rely less on hunches and the general theory alone. With the help of analysis and knowledge gained before and during the game, I suggest that this area of football coaching, play calling and practice can be improved for any football team.
Major variables include: the score, the time left on the clock, the number of times out left, the average gain per rushing attempt on the season and on that day, the average gain per pass attempt on the season and on that day, the risk of throwing a pick on the season and on that day, the risk of a QB sack with fumble on the season and on that day, the risk of an incomplete pass of different lengths on the season and on that day, how tired both team's DLs are that day, the momentum at that time, whether the team is playing at home or on the road, the physical injury or health condition of the RBs and OL at that point in the game, how many first downs are needed to kill the clock, can we really stop them if they get the ball back, and perhaps several other variables.
Then we should differentiate between a two-minute drill, a four-minute drill and a six-minute drill in "kill the clock" offensive practice and have many repetitions just like is normal for the red zone, two-minute offense and the goal line defense. Put on the play calling card about five or six of your best runs and about the same number of short passes, picks, slants, crossing patterns, a delayed release by the TE and/or RB. We might call a pass on first or second down on the thought that what the real goal is to run out the clock and a completed pass for five to ten yards will be just as good at keeping the clock running as a run provided no one goes out of bounds.
The variables need to be updated by an analyst with a computer on the sidelines. The data from that game will be by far the most relevant. For example, if we are only averaging 2.9 yards rushing per attempt and averaging eight or nine yards per pass attempt, then it is probably worth it to throw a pass on one of the three downs unless no first downs are needed. That is an easy case. Three knees in Victory formation. The harder cases are like what we faced at Ohio in Columbus and Iowa this year. You probably need two or three consecutive first downs to kill the clock even with two TEs and a Fullback. It can be done against many teams and the goal is to be good enough at RB and the OL to do it whenever needed. But the chances of success vary greatly by opponent. In hindsight, we were not good enough to do that at Ohio or on the road at night against Iowa this year .
What to do?
Let's look at Plan B and C, which we need in our arsenal despite conventional wisdom. If the defense is stacked to stop the run with eight or nine defenders in the box, it will be very hard to pick up a first down with rushing only. If the completion percentage is about 60% on the game as a whole, it might reasonably about 75% to 80% with only three DBs back and crashing LBs. If the risk of a pick on the season is normally about 2 or 3 % and maybe it is twice that in this particular game, it is much less against a defense rushing seven or eight defenders and throwing shorter passes. The risk of throwing an incomplete pass is modeled as just the inverse of the completion percentage, so in our example, it might be down to 20 to 20%. But the passes have to be three-step quick deliveries to uncovered receivers or a play action pass.
Basically, the general theory is too general and does not always apply. The score, the time left and the opponent might mean on any given Saturday or Sunday, the general theory does not apply. It might be the case that throwing on first or second down is the best strategy while using the pass to move the chains and run out the clock. Maybe play action works well followed by a seam or double seam pattern with only one safety. That is where the data and extra emphasis in practice comes in.
We can still strive each year to develop the ability to do the idealbest answer, which is an unstoppable run game that can generate six to eight minute drives on the ground when needed. But we will not have that ability every year and that is when we need to be able to accomplish the same thing in different ways.
Does the general theory and its variants make sense? Maybe, but it all depends on many factors, which are game and situation dependent. Teams have started to track data, plug in the known variables and rely less on hunches and the general theory alone. With the help of analysis and knowledge gained before and during the game, I suggest that this area of football coaching, play calling and practice can be improved for any football team.
Major variables include: the score, the time left on the clock, the number of times out left, the average gain per rushing attempt on the season and on that day, the average gain per pass attempt on the season and on that day, the risk of throwing a pick on the season and on that day, the risk of a QB sack with fumble on the season and on that day, the risk of an incomplete pass of different lengths on the season and on that day, how tired both team's DLs are that day, the momentum at that time, whether the team is playing at home or on the road, the physical injury or health condition of the RBs and OL at that point in the game, how many first downs are needed to kill the clock, can we really stop them if they get the ball back, and perhaps several other variables.
Then we should differentiate between a two-minute drill, a four-minute drill and a six-minute drill in "kill the clock" offensive practice and have many repetitions just like is normal for the red zone, two-minute offense and the goal line defense. Put on the play calling card about five or six of your best runs and about the same number of short passes, picks, slants, crossing patterns, a delayed release by the TE and/or RB. We might call a pass on first or second down on the thought that what the real goal is to run out the clock and a completed pass for five to ten yards will be just as good at keeping the clock running as a run provided no one goes out of bounds.
The variables need to be updated by an analyst with a computer on the sidelines. The data from that game will be by far the most relevant. For example, if we are only averaging 2.9 yards rushing per attempt and averaging eight or nine yards per pass attempt, then it is probably worth it to throw a pass on one of the three downs unless no first downs are needed. That is an easy case. Three knees in Victory formation. The harder cases are like what we faced at Ohio in Columbus and Iowa this year. You probably need two or three consecutive first downs to kill the clock even with two TEs and a Fullback. It can be done against many teams and the goal is to be good enough at RB and the OL to do it whenever needed. But the chances of success vary greatly by opponent. In hindsight, we were not good enough to do that at Ohio or on the road at night against Iowa this year .
What to do?
Let's look at Plan B and C, which we need in our arsenal despite conventional wisdom. If the defense is stacked to stop the run with eight or nine defenders in the box, it will be very hard to pick up a first down with rushing only. If the completion percentage is about 60% on the game as a whole, it might reasonably about 75% to 80% with only three DBs back and crashing LBs. If the risk of a pick on the season is normally about 2 or 3 % and maybe it is twice that in this particular game, it is much less against a defense rushing seven or eight defenders and throwing shorter passes. The risk of throwing an incomplete pass is modeled as just the inverse of the completion percentage, so in our example, it might be down to 20 to 20%. But the passes have to be three-step quick deliveries to uncovered receivers or a play action pass.
Basically, the general theory is too general and does not always apply. The score, the time left and the opponent might mean on any given Saturday or Sunday, the general theory does not apply. It might be the case that throwing on first or second down is the best strategy while using the pass to move the chains and run out the clock. Maybe play action works well followed by a seam or double seam pattern with only one safety. That is where the data and extra emphasis in practice comes in.
We can still strive each year to develop the ability to do the idealbest answer, which is an unstoppable run game that can generate six to eight minute drives on the ground when needed. But we will not have that ability every year and that is when we need to be able to accomplish the same thing in different ways.
