Developing Young Talent in the NFL: An Application of Game Theory
I recently read an article about whether rookie quarterbacks should be starting in the NFL or whether they should learn the game behind a veteran for some time before starting. Today, the value of the quarterback to an NFL team is rising, as seen by rising contract sizes for franchise quarterbacks and the increasing number of first-round draft picks spent on young quarterbacks (20 QBs were taken in the first round in the 1990s, but in the last 10 years 33 QBs were taken in the first round). With these crucial picks being invested in young quarterbacks, more and more NFL teams are having to make the decision of whether to play their experienced veteran QB or their younger QB who may have higher potential. One example of this is the 49ers decision to sit #3 overall pick Trey Lance behind their veteran quarterback Jimmy Garoppolo. However, the opposite happened when the Jaguars decided to immediately start their #1 overall pick Trevor Lawrence in lieu of the more experienced quarterback Gardner Minshew II. A modern example is the Pittsburgh Steelers, who so far have opted to start Mitchell Trubisky, yet their slow start to the season have led to calls for their rookie first round pick Kenny Pickett to take over as the starter. The pros and cons of whether to play the veteran or the experienced QB can be viewed in a playoff matrix of two teams trying to make this decision.
Short term expected payoff:
| Start veteran QB | Start young QB | |
|---|---|---|
| Start veteran QB | +1/2, +1/2 | +1, +0 |
| Start young QB | +0, +1 | +1/2, +1/2 |
In the short term, starting a veteran QB makes a team much more likely to win the game. For simplicity, let’s say that starting a veteran QB against a young QB leads to a win for the veteran. In games where either a veteran is started against another veteran, or a young QB is started against another young QB, let’s say each team is equally likely to win the game. The above payoff matrix shows this example, and that both teams would choose to play their veteran QB because they will either win the game or have a 50/50 shot at winning the game., but it’s incomplete. In real life, the team’s payoff is not solely dependent on the outcome of the game, but also the development of their young star. Some teams choose to start their young QB, knowing it may result in more losses, but also knowing that the experience of starting more games will prepare their young QB and help them grow more comfortable in the NFL. The importance of winning now versus developing your young star also varies by team. For example, the 49ers have a great team besides their quarterback, so they felt it was worth trying to win now with Jimmy Garoppolo (the veteran) rather than waste the prime years of their talented defense and skill players trying to develop a young QB. However, the Jaguars have a pretty awful team, so they would value starting their young QB more and developing him while they already thought they were going to lose games due to the lack of talent surrounding him. So, to make the matrix more accurate, we need to include the benefit of developing a young QB and the idea that different teams may have different ways of calculating their payoffs.
Longer term expected payoff:
| Start veteran QB | Start young QB | |
| Start veteran QB | 1.25, 3.75 | 2.5, 1.25 |
| Start young QB | 7.5, 7.5 | 8.75, 6.25 |
The payoffs in this matrix require a bit more explanation. First, the team on top is more talented now, which means they place greater emphasis on winning now, represented in the payoff matrix by 75% of their payoff based on winning games now and 25% of their payoff based on developing their youth. Meanwhile, the team on the left is less talented now, which means they place greater emphasis on developing their youth, so their payoff is 75% based on winning games and 25% based on developing their youth. Also, instead of a scale from 0 – 1, 0 meaning a loss and 1 meaning a win, the payoff values are now on a scale from 0 – 10, 0 meaning a loss and 10 meaning a win, while starting the young player increases the payoff by 5. In this case, the Nash equilibrium is for the team on top to play their veteran QB and the team on the left to play their young QB, which is what we would expect. It’d be interesting to see how well this lines up with the thoughts of current-day NFL teams who are waiting to start their young QBs.
Sources:
https://www.drafthistory.com/index.php/positions/qb
https://www.theringer.com/nfl/2021/9/22/22687282/nfl-rookie-quarterback-ready-justin-fields-bears-trey-lance-49ers