Explained: Team Talent Ratings

As we embark on our second season running this website, we did some reflecting in the off-season about how we would be different than other information sites (of which there are many, and most are great). I’m going to do my best to explain in short-order what the team talent ratings and rankings ARE and ARE NOT. If you don’t read anything beyond this next paragraph, you should have an idea of what you are looking at when viewing our LIVE, dynamic team talent ratings and rankings:

This is a data-driven approach to analyzing the talent level of players in the rotation, by positional group, for each of the 134 FBS schools. The factors include individual player ratings (more info on that coming in the next “Explained” article) and their projected playing time. This is NOT a ranking of how well the team will perform on the field. In order to get to that, we’ll need to take into account other factors, including opposition, coaching (player development and scheming), depth, and more.

The purpose of this approach is to show in real-time any glaring strengths, weaknesses or gaps in a positional unit. During the season, as injuries pile up, these ratings will adjust and show how they affect each team and hopefully help us identify some “edges” to be exploited in your wagering, fantasy or just general knowledge of the game.

VIEW: Live Team Talent Rankings >>

Because this is a data-driven approach, the more data we have to work with — the better the picture that is drawn for each positional group and eventually the entire team. Example time. Using our early “State of the Chart” 2024 updates, we see one thing that immediately jumps off the page – and something that got a lot of buzz around this X post – in that Indiana is rated higher than Michigan. Don’t freak out. Again, we are not saying that Indiana is a better team than Michigan. We are simply saying that by our ratings system and model, the re-build Hoosiers are coming out ahead of the somewhat decimated Champion Wolverines. As the games start rolling and more data is collected, you can expect to see Michigan pass up the Hoosiers – or will they? Bottom line: we need to see more from Michigan before we put all of our chips in the repeat basket and that there is a real potential buying opportunity on Indiana.

So, not only are we data-driven, we aren’t driven by the same data that drives a lot of the ratings/rankings models in the industry. If you are looking for a top 25 ranking website…well, keep looking. In the future, we will be developing a “potential” model for each play to show a “ceiling” if they reach their recruiting / social hype factors in their time amongst the college ranks. Stay tuned for that.

THE SAUSAGE MAKING

Okay, so now that we got the high-level out of the way, we’ll dive a bit more in to the ‘how’ we come up with the talent ratings and subsequent rankings amongst the 134 FBS schools. 

  • It all starts with the player-level ratings. These drive the ratings/rankings and they are a whole animal in and of themselves. Which we will be divulging more on that algorithm here shortly. For now, take our word for it. These ratings are a proprietary algorithm taking into account several data factors including: recruiting rankings, on-field performance, level of competition, player-level grades via Pro Football Focus and more. We combine these sources to put out a real-time rating out of 20, including an overall and speciality skills ratings.

Before you ask, here is the general rule of thumb on our 20-point player ratings scale:

  • Over 17 = Star player destined to play on Sundays
  • 15-17 = Very high-level player, a starter on just about any P4/5 school
  • 13-15 = Good player; starter at most G5s; rotation-level player on most P4/5s depending on depth at position. Top recruits start in this range based on their consensus recruiting grade from 247/Rivals.
  • 11-13 = Solid player; playing time depends on situation at each school / unknown 3/4-star recruits with no collegiate playing history
  • 9-11 = Replacement level player & 3-star recruits with no playing experience
  • 7-9 = Below average player. If getting a significant amount of reps, you should be slightly worried
  • Below 7 = Poor players and/or those with no recruiting history and/or walk-ons
  •  Next, it goes to our playing time projections. We project out the amount of snaps each player is expected to see on the field and thus affecting the rating of each category. You can see our playing time projections and/or target/volume shares on each depth chart page for each player.
  • Finally, each player contributes their skills to the ratings for each of the categories we are tracking and then weighted based on our own weighting system to determine the “score” for each of the areas, which feeds into the overall offense, defense and specialists ratings and then finally to the overall team rating (also weighted).
Here is a quick (high-level) look at how these all feed into each other:
 

The weights are key here as everything builds up to an overall rating. As for the offense / defense / specialist weights, those shift each year based on the state of our game. Right now, a really good offense will only be slowed by a really good defense – hence the lean towards the offense contributing a bit more to the overall rating. 

USE CASES

Finally, what do we use these ratings for? What good are they? 

Pre-Season Level Set: In our current age of college football, with crazy turnover and moving of players, this data-driven approach gives us a baseline to work from as we start a season to give us a glimpse in the pure talent level (that we know of) heading in. 

In-Season Trending: The model then ingests the data we receive as the season rolls along and gives us an idea of who is trending up or down to aid in future prognostications. We’ll keep an eye on the teams, and each of the units listed above, to see the direction they are moving and include that in our game-by-game projections.

Injury Affect: When the injuries start piling up – and they will – how does that actually effect the talent level of a full unit or a speciality within? For those teams with good depth, the affect will be lessened. If this star player is at a G5 and we don’t have data on backups, we could see quite a drop in production.

Gap/Differential Analysis: A fancy way of saying that we can do some quick comparisons of units to see where there are the largest talent differentials show up and draw some insights on what that means from the “gaining an edge” stand point. The focus is in the trenches and we’ll be putting out a “Trench Report” each week to show you the biggest differentials compared to the Vegas lines to see where there may be an opportunity.

Questions? Comments? Suggestions? We’ll take it all.

Hit us up on X or e-mail us directly to share your thoughts.