Unlocking the Secret Twist That Will Change Everything You Thought You Knew About This Story

Ever find yourself mid-game, sneaking a peek at Myles Garrett with a “Where does he really rank among pass rushers?” kind of curiosity? Or maybe you’re sitting courtside (or, okay, on your couch) watching Jonathan Taylor bulldoze through defenses, scratching your head at how one guy can be so dang effective. Well, buckle up — because Opta Analyst’s coming in hot with a fresh take on NFL player ratings unlike anything you’ve seen before. This isn’t just a numbers game; it’s a deep dive powered by 1-on-1 ELO ratings that slice through the fluff and hype, adjusting for who’s truly worth beating. Offense, defense, and every gritty matchup in between — these ratings pull back the curtain on players’ skills with a predictive punch. So, if you’re wondering who’s best at pass rushing, run blocking, or just making those clutch plays nobody talks about… your answer might just be a click away.

LEARN MORE

These new NFL player ratings come from our ELO scores for each player’s skill set. Our ELO rating system uses 1-on-1 data and runs it through a model that is designed to adjust for the strength of competition.

Do you ever sit back and wonder where Myles Garrett stacks up as a pass rusher in the league?

Or maybe you’re a Colts fan watching Jonathan Taylor run over defenses and want to know why he’s so effective this year. Maybe it’s getting close to the trade deadline, free agency, or the draft and you just want to figure out where your team needs to improve.

Introducing Opta Analyst’s NFL player ratings.

Our player ratings are designed to describe a player’s skill set. In the initial release, each side of the ball has three sets of ratings. On defense, we have pass rush, run defense, and pass coverage. On offense, we have pass protection, run blocking, and route-running ability.

Each player has a rating of 1 to 100 in each of the sets of skills. In most cases, a player will only have two applicable skills. A wide receiver is not often asked to participate in pass protection, but they are used as run blockers and route runners.

The ratings are also designed to be predictive. Getting a sack is a good result for a defense, but getting a sack is not necessarily a good indicator of getting future sacks. A player with a high pass-rush rating is more likely to get future pass-rush wins.

What data goes into the ratings?

The ratings are all produced from our 1-on-1 matchups data. Throughout a game, players are asked to do a whole array of tasks, and the evaluation of performance is often linked to a teammate’s performance. A receiver running a crossing route must rely on his quarterback to deliver the ball accurately and on time, but the QB must also rely on the pass protection to hold up long enough to deliver the ball.

Evaluation using simple play-by-play data is not good enough. There is too little information in the play-by-play to draw meaningful conclusions about the player’s skills.

Our 1-on-1 matchup data aims to solve this problem. It isolates situations in which a player is tasked with winning a matchup against another player. We’ve written about pass rush 1-on-1s before but have only briefly discussed run blocking and pass coverage.

As the data type suggests, we’re looking for 1-on-1 opportunities in both route running versus pass coverage and run blocking against run defending. For a deeper look at both, check out the section “A deeper look at 1-on-1 data collection” to follow.

How are the ratings calculated?

At the beginning of the data set, each player is assigned the same value: 1200. For each matchup, the model predicts an expected outcome based on the two players’ scores. For the first matchup, it would be 1200 vs. 1200, or a 50% chance of a win for both the offense and defense. The winner sees their score increase (in this case by about 10 points), while the loser sees their score decrease (by the same amount).

As more matchups occur, the expected results become more varied. If a 1400 player faces a 1200 player and wins, that player’s score increases by about five points. If the 1400 player loses, the change in score is over 15 points. This way, we’re properly rewarding a player for beating worthy opponents.

Tampa Bay Buccaneers left tackle Tristan Wirfs is our highest-rated pass blocker with an ELO of 1778 (as of Week 9, 2025). Let’s say the Bucs are set to play the New England Patriots, with Wirfs likely to see a lot of K’Lavon Chaisson, who has a rating of 1360. If Wirfs wins a matchup, the point change would only be about 1.7 points. Wirfs would be expected to win almost every time, so him actually winning wouldn’t hurt Chaisson much.

But if Chaisson were to win a matchup, the change would be over 18 points. This system helps us to better contextualize 1-on-1 performances.

This does create some huge swings when a player has a good or bad game, so we use an average of a player’s last 50 ELO scores for a current rating.

What do the ratings mean?

Our 1-100 scale is more about an intuitive understanding than anything else, but let’s dig into some more player examples.

Wirfs has a rating of 100 and a 1-on-1 win rate of 94%. Seattle Seahawks left tackle Charles Cross has a rating of 82.3 and a win rate of 78%. A higher win rate usually means a higher rating, but not always. As we’ve already discussed, the strength of competition can play a huge factor in a player’s rating.

Being a Cincinnati Bengals tackle is a thankless job. The Bengals pass a zillion times per game, and one quarter of your season is spent blocking the pass-rush units from the Pittsburgh Steelers and Cleveland Browns. Orlando Brown, the Bengals’ left tackle, sports a win rate of 70%, but he has a rating only slightly below Cross at 81.6.

His win rate is lower than Miami Dolphins LT Patrick Paul, Carolina Panthers LT Ikem Ekwonu and Indianapolis Colts LT Bernhard Raimann, but those players have ratings of 77.5, 76.0, and 75.5, respectively.

We choose this rating system rather than raw win rates because the ratings are designed to account for the strength of competition, while the raw win rates are not. The rating system gives us a quick snapshot of a player’s skill.

Is there a relationship between play success and 1-on-1 matchups?

Yes, there is! Let’s look at pass-rush matchups first. On plays in which there was at least one 1-on-1 pass-rush matchup, the offensive team averages 7.3 yards per play (YPP) when there are no pass rush wins, 5.7 YPP when there is one win, 4.9 YPP when there are two wins, and minus-0.5 YPP when there are three wins.

More pass-rush wins equals fewer yards for the offense.

Routes run vs. pass coverage matchups are broken up into man and zone categories. All man coverage does not necessarily fall into the man matchups category, however. Mesh concepts (think pick plays) rely on “picking” the coverage defender, rather than relying on the route-running ability of the intended receiver.

We restrict man-coverage matchups to situations in which it is truly man vs. man. A fade route against man coverage is a good, easy example.

On plays in which there was at least one man-matchup win for the route runner, the offense averages 10.3 yards per play. On plays in which there was at least one man-matchup win for the defender, the offense averages 3.4 YPP. This does not mean the receiver who won was targeted, but it is a significant difference in offensive output.

For zone coverage, the difference is 9.6 yards if there is at least one offensive win and 4.7 yards if there is at least one defensive win. Man coverage is a higher-risk, higher-reward coverage. For this reason, we weigh man coverage ability slightly higher than zone coverage ability when it comes to ratings for route running and pass coverage.

Do different positions get more or fewer 1-on-1s?

Yes! Here is a breakdown of the different offensive positions and what percentage of total 1-on-1s they get in pass pro and run blocking

Pass Protection

LT – 28.34%
LG – 19.53%
C – 7.85%
RG – 19.98%
RT – 24.29%

Run Blocking

LT – 13.08%
LG – 14.97%
C – 14.12%
RG – 14.76%
RT – 12.16%
TE – 22.09%
RB – 2.41%
WR – 6.41%

One interesting note here is that tight ends get the highest percentage of 1-on-1s on the offense. A good run-blocking tight end can be key to a good rushing offense.

Check out the leaderboards to see which tight ends are currently the best run blockers.

Are these ratings all-encompassing?

The ratings are designed to describe certain skills. Run-blocking double teams are not included in the data set, largely because there is a huge communication element between linemen to properly execute a double team.

In fact, we don’t include any crossups in the data set. We can’t say for certain if a crossup is the fault of the player, the quarterback, the coach, or someone else. Because we’re trying to isolate situations in which a player has a match against another player, we want to make sure both sides are aware of what’s happening to get clean results.

Awareness and understanding are huge parts of being a good player, but they aren’t something that is easily quantified using data.

Pass-rush and pass-pro ratings describe a player’s ability to block 1-on-1. We don’t use stunts in the data, though we collect those and may release something on stunt protection in the future. Run blocking and run defense focus on executing or beating blocks. It does not describe a player’s ability to tackle, which is a part of defending the run.

Route running is one part of being a good pass catcher, but the act of catching the pass itself is a different skill, and it is another we’re working on releasing. Route ratings are best thought of as “players most difficult to defend in pass coverage.”

Tackling as a whole is a huge part of the game. You can have a zone-matchup win by driving on a flat route, but if you miss the tackle and the receiver gets another 8 yards, is it really good coverage? We’d argue that it is good coverage, but bad tackling.

We’re truly trying to isolate skill sets with these ratings, but stay tuned for tackling ratings! The data is there, the modeling is there, we’re just working on getting them into ratings form.

A deeper look at 1-on-1 data collection

With run blocking and pass coverage, data collection starts with the question: What is the purpose of the play?

For a quick slant against press-man, the receiver’s purpose is to get inside leverage on the defender and get enough separation that the ball can be delivered without the defender getting his hand in to break it up.

For a fade route, the defender’s purpose is to use the sideline to squeeze the receiver, then make a play on the ball. Once we have “purpose” in mind, we can evaluate whether or not that purpose was fulfilled.

If the receiver catches a jump ball on the fade route, the receiver has won. If the Cover-2 corner doesn’t get enough depth against an out route, the receiver has won. If a defensive lineman cuts off an intended rushing gap, the defender has won.

Each play has its own definition of success.

Once we understand the purpose of the play, we can see how the different players are being used to fulfill that purpose. On a classic power run to the right, the play-side guard and tackle are responsible for sealing the edge and working up to the back-side linebacker.

The lead blocker clears out the end on the line of scrimmage and the puller works up to the play-side linebacker. In this play design, we typically find three 1-on-1 opportunities: lead vs. end-man, play-side G/T vs. back-side linebacker, and puller vs. play-side linebacker.

If the end-man blows up the lead blocker and forces the ball carrier to the back-side A-gap, there will probably be a defender waiting to make the tackle. This defender might get a stuff or tackle for loss, but it was the 1-on-1 between the lead blocker and end-man that created the short gain.

Follow along on Instagram, Bluesky, Facebook and X for more.

The post Introducing Opta Analyst’s NFL Player Ratings appeared first on Opta Analyst.