Last week, I published my first set of 2024 projections on the site. In previous years, I have shared a note like the one below when introducing them for the first time each season.

Creating accurate fantasy football projections is a time-consuming process, but it’s one that I’ve grown to appreciate. It’s an important part of my annual offseason preparation because it helps me understand a player’s potential outcomes, the key factors that will shape his season, and the portion of his team’s offense he might realistically control. Assigning market shares across an offense can be eye-opening and shows just how important, yet elusive, opportunity is. In many ways, the process of building projections is more important than the final result.

Unfortunately, it’s difficult to convey this process without experiencing it firsthand. Much of the information I gather while working through my projections doesn’t make it into the final numbers. That’s why I always remind readers that projections are just one part of a larger decision-making process. Avoid becoming too attached to any projections, including mine. They are just best guesses that sometimes lack context and don’t fully address a player’s range of outcomes, potential downsides or upsides, or the nuances of his offense.

To address some of the shortcomings of projections, like those I mentioned earlier, I decided to take my baseline projections for 2024 and adjust them in various ways. The goal is to better capture and illustrate the different ways a player’s season could unfold. Before we discuss that process, let’s first review how I build my projections.

I should note that we have a tool here at RotoViz called the Projection Machine, which walks you through the entire process and makes building projections much easier. I strongly recommend spending a few hours with it this summer. While doing so, keep in mind that you might not account for unforeseeable challenges or hiccups during a player’s season, which could lead to overly optimistic results.

The RotoViz Projection Building Process

At RotoViz, our projection-building process uses a layered but holistic approach and consists of three distinct phases.

• Phase 1: We estimate the number of plays a team’s offense will execute. As I previously discussed, wins and losses are a major driver of a team’s overall play volume. This total is then divided between rushing and passing attempts.
• Phase 2: We assign percentages of passing and rushing attempts to specific players, determining the workload they are expected to handle.
• Phase 3: We factor in player-specific assumptions such as catch rate and yards per attempt. When developing efficiency assumptions, we consider each player’s historical results, changes in their situation, positional benchmarks, and other predictive factors.

When building my projections, I create assumptions for every skill player expected to garner more than 1% of rushing or passing attempts. This is important because the results of all passing attempts aggregate to the quarterback. Additionally, it helps ensure that the percentages of plays assigned to each player are realistic. In the projections presented on the site, only players projected to score more than 75 PPR points are included.

The full set of projections and the underlying stats can be found here!

Making The Projections More Useful

As we discussed, for most players, the opportunity they earn in a year is the biggest driver of their fantasy scoring. Therefore, if we want to better understand the outcomes they could produce in a given season, we need to understand how their overall opportunity could change and assess the impact of this. This involves considering changes to team play volume or the player’s specific market share.

To account for this, I went through the following process for the RB’s, WR’s, and TE’s for every team:

• Increased the team’s passing volume by 5%, holding all other variables the same
• Increased the team’s passing volume by 10%, holding all other variables the same
• Increased the team’s rushing volume by 5%, holding all other variables the same
• Increased the team’s rushing volume by 10%, holding all other variables the same
• Decreased the team’s passing volume by 5%, holding all other variables the same
• Decreased the team’s passing volume by 10%, holding all other variables the same
• Decreased the team’s rushing volume by 5%, holding all other variables the same
• Decreased the team’s rushing volume by 10%, holding all other variables the same

*Quarterbacks were not included in this process as more nuance would be required and this exercise felt like one more suitable for the other positions. If time allows, I will hone in on something similar for passers.

**Before walking through this process, I looked at historical play volume data from the last 20 seasons. In cases where adjusting a team’s volume up or down by the noted percentages would move them outside of historical boundaries, I lowered the adjustment percentages. This allowed me to capture outcomes where my volume expectations were off without moving teams into likely unreasonable ranges. This was not a perfect process, but it is probably better than skipping this step altogether.

Next, I did the following for each player:

• I increased his rushing market share to what I believed to be the highest percentage that could be reasonably achieved without a major injury, trade, or other unforeseeable event affecting his team.
• I decreased his rushing market share to what I believed to be the lowest percentage that could be reasonably recorded without a major injury, trade, or other unforeseeable event affecting his team.
• I increased his target share to what I believed to be the highest percentage that could be reasonably achieved without a major injury, trade, or other unforeseeable event affecting his team.
• I decreased his target share to what I believed to be the lowest percentage that could be reasonably recorded without a major injury, trade, or other unforeseeable event affecting his team.

For all the scenarios outlined above, I recorded each player’s fantasy points. The results of this process have been aggregated and will be shared in a future article. In the coming weeks, we will work through these projections while also considering the output of the Range of Outcomes tool. By analyzing players from all these different angles, we will be much better equipped to build our 2024 fantasy squads.

Example: Atlanta Falcons

The bottom table shows the impacts of changes in team play volume. Since Atlanta’s projected volume