Simulating Quarterback Injuries With The NFL Simulator

Simulating Quarterback Injuries With The NFL Simulator

Jack Andrews
August 14, 2021

Carson Wentz broken foot


Simulating quarterback injuries and their effect is a key advantage of the Unabated NFL Simulator. With our ability to factor in QB Injury probability you can not only account for the randomness of injuries, but you can see the impact they would have on a team’s season. You can simulate to be prepared to react to breaking news, or at least know how to simulate when news breaks. 


The Case of Carson Wentz

On July 29th, 2021, Carson Wentz injured his foot during training camp for the Indianapolis Colts. A few days later, the team revealed he had a fractured foot that would require surgery. He was slated to miss 5 to 12 weeks. That’s quite a wide window, but apparently at this point the exact timetable couldn’t be known. 

Given the date of the surgery, he was expected to return somewhere between Week 1 and Week 7. How could we have ballparked the effect the Wentz injury may have had on the Colts’ season? We used the Unabated NFL Simulator to run through some scenarios and formulate some bets. 


Creating Baseline Ratings

We first started with the Colts’ baseline power ratings. The preseason ratings when a healthy Carson Wentz was assumed. As the season progresses, Unabated updates the power ratings when new power ratings are released. 

We track several different publicly available ratings. However, I used the Market-Based Ratings for a couple reasons:

  1. Since they are based on sportsbook look-ahead lines, they will have less variance in the output. Ideally a sportsbook’s futures prices should coincide with the implied ratings that built their look-ahead lines. In 2021, the Unabated Market Based Ratings were from the Superbook.
  2. This will also reveal to us if the market seems to have over or under-reacted to the Wentz injury. Obviously, that’s valuable since I can bet into the market. It’s not so easy to bet into Massey-Peabody or ESPN.

Here were the Market Based Ratings before Carson Wentz went down. You can download these ratings and then cut and paste them into the template and use them via the Upload My Own feature.


Unabated NFL Simulator Upload


Here were the results of that simulation. For brevity, I’m only displaying the AFC South.


AFC South Preseason Ratings


As you can see, the Colts were projecting to 9.5 wins, and a strong 47.5% chance of winning their division. 


Carson Wentz Out for the Season

The first scenario we simulated was Carson Wentz out for the season. That wasn’t the prognosis but it allows us to see how much effect he could have on this team. We have two paths we can take to reach this projection. 

First, we could deduct Carson Wentz’s value from the offensive power rating for the Colts. The NFL league average deduction is -3. That’s the default value we list in your simulator design screen in Basic and Advanced modes. However, Wentz’s backup was Jacob Eason, a 4th round 2020 draft pick who hadn’t yet taken a professional snap. The drop off from Wentz to Eason is bigger than the league average. Sometimes simulating quarterback injuries involves some qualitative adjustment. 


Projecting Quarterback Injuries Using Massey-Peabody Research

How do we know how big a drop off it is? Enter Massey-Peabody Analytics. One of the metrics that they analyze is precisely this. The contribution of the starting QB v. the contribution of the backup QB. You can see their projections by creating a new sim in Unabated mode and viewing their ratings.


Massey Peabody settings for QB injury


The drop off from QB1 to QB2 in 2021 was a rather large -6.29! If we wanted to simulate what it would look like for the Colts to be using their backup QB for the entire season, we could go edit the preseason sim we ran and change the Colts Offensive Rating from +1.56 to -4.73 (1.56 – 6.29). Then change the probability of QB1 Injury to 0 percent to signify that this is the QB (Eason) that they’re going to be stuck with all season. This is the easiest way to project any NFL team that loses their starting quarterback for the rest of the season. 

Here’s how that sim played out:


simulator results no wentz


The Colts dropped nearly 3 wins off their projection and now the Titans become odds-on favorites to win the AFC South. Even the Jaguars had a better chance. Yikes. 


Projecting Quarterback Injuries Using Market-Based Ratings

Adjusting the offense for the QB change is the first way of modifying the Colts power ratings. The second method for simulating quarterback injuries is to see what the new look-ahead lines say given Wentz’s absence. I went back to the Market Based Ratings as of Aug. 2, before the market knew the timetable for Wentz’s return. If you’re curious yourself, here were the ratings that day:


simulator ratings with wentz out for season


Here’s a link to the full Market Based-Ratings (based on look-ahead lines) from that day.

The market dinged both the Colts offense and defense. Their overall rating of -3.94 was similar to the adjustment we made which took it to -4.45. It seems the market agreed with the deduction Massey-Peabody had graded the Colts at. Here is how the simulator results looked for the AFC South based on the Market-Based Ratings then:


simulator results with wentz out for season


As expected, the market at the time was down on the Colts chances. Good luck capitalizing on that as most books had Colts team futures off the board. Again, all of this is purely just an exercise since Wentz was not out for the season. He was expected to miss between 1 and 6 games. How do we quantify that?


Simulating When Wentz Returns

Creating a power ratings system that basically accounts for no Wentz for the first 1-6 weeks of the season and then a potentially healthy Wentz for the rest of the season can be a difficult task. To do this we could again go to the current look-ahead lines (at the time, on, August 12, 2021) which only factor a Wentz-less Colts team for the first six weeks of the season. In the remaining games, the market assumed a better Colts team.  

Building a power ratings system from these lines, we get the following ratings values:


Simulator design current market based ratings


If you want to use these settings, you can download them here.

As expected, these results increase the Colts projected wins and improve their chances in the division.


Simulator results with Went back Week 6


However, this model still feels very inexact. After all, what if Wentz comes back sooner? What if Wentz suffers a setback? Will he come back at 100 percent? What if Wentz suffers a different injury? There were a lot of unknowns involving the Colts despite everything we modeled in. Simulating quarterback injuries also requires simulating variance.


Simulating Quarterback Injury Variance

Fortunately, another innovative feature of the Unabated NFL Simulator is that we can tweak the variance in the model using our unique Rating Update Function (RUF). Essentially, you are instructing the model how to act as it progresses through a simulated season. Should it learn from the results in simulated games, or should it stick strictly to the script with how it approaches the next simulated game? 

Rufus spells it out in this article. If a team is going to function exactly how you expect, then you take a more conservative RUF. If there are a lot of wildcards at play, such as there were with this Colts team, you increase the RUF towards Aggressive. There are 11 stops on the sticky slider that makes up RUF. My recommendation is don’t go too extreme in either direction. For now, let’s set the slider towards Aggressive and see what happens to our sim:


Simulator design with Colts set to aggressive RUF


This resulted in a sim where the Colts projected season wins remain relatively unchanged, however their chance of winning the division, the conference, or even the Super Bowl showed increases as some tail events benefit from the variance.


Simulator results with Colts set to aggressive RUF


Our simulator, built on look-ahead lines, differed from the market slightly once books started to introduce Colts futures again. The vast majority of season wins lines for the Colts were more bullish than our simulations. Here is a quick snapshot with our simulator-generated odds in blue:


current market odds for colts season wins


The Colts finished 9-8, second in the AFC South, and missed the playoffs.

Future NFL Simulator Ideas

Going forward, the ideal way to handle the uncertainty of a QB who goes down is to wait until there is some certainty involving his participation. Another benefit of the NFL Simulator is that past results are baked in after they occur and the simulator only deals with future results. Simulating quarterback injuries becomes easier when you know how the team has performed with the quarterback. 

After the quarterback is confirmed to be returning in the next start, you can run a new simulation where the remainder of the season can be played out. More information will be known at that time regarding how a team operates without its starter.

Our stated goal for the tools at Unabated is to give you the ability to find answers to the questions you create. Using the simulator as a what-if tool is a powerful way to use it. Take the ideas from this article to approach other situations where you can simulate the effect of removal to find the answers you seek. 

Copyright 2024 © Unabated Sports, Inc. All Rights Reserved.

This site is strictly for educational and informational purposes only and does not involve any real-money betting. If you or someone you know has a gambling problem and wants help, call 1-800-GAMBLER. This service is intended for adults aged 18 and over only.