Value of Things: QB Similarity Scores

· Yahoo Sports

By far, the most influential person in the history of sports statistics is Bill James. He focused mainly on baseball, but there are some concepts that are portable to other sports. One of those concepts is the concept of similarity scores. The idea was to look at players across the different category and find the most similar players to them. It included the basic counting numbers like home runs, runs, RBI, hits, and stolen bases.

The use of this was actually pretty ingenious. When you are a fan of a player you have an emotional connection to them. You will naturally inflate or deflate his abilities. Similarity scores allow you to compare your favorite player to other players they are similar to. They probably will be players you do not have an emotional attachment to, so it will be much easier for you to peg your player’s place in the history of the game.

Visit asg-reflektory.pl for more information.

Before I introduce my formula I should throw out a few caveats. One player having a higher score than another player does not mean they are better. These scores are so close that the difference is negligible. Plus, people will correctly point out that some are in better situations than others and that the three year average (or two) is skewed based on a very good last season or a very bad last season. All of these things will impact how a player is perceived in and out of the sport.

What I like to do is use multiple rating systems and combine them into one number. I have done that in the Hall of Fame Index and Hall of Fame Index Part II in baseball. The idea was to get a cross-section of analytical thought on a player and combine those to see who players were similar to. That formula is much more complex than this one. In this edition, we have added quarterback rating (standard), QBR, and PFF scores to come up with one total sum. In each case, we took the average over the last three years (two if the player has only been active for two years). Again, this is not a rank ordering, but a cross-section of who is most similar to C.J. Stroud over the last three seasons.

RatingQBRPFFTotalTrevor Lawrence88.257.180.1225.4Sam Darnold95.654.375.2225.1C.J. Stroud93.555.576.1225.1Bo Nix90.556.077.7224.2Kirk Cousins92.452.978.8224.1Tua Tagvailoa97.052.073.1222.1

The number one rule of analysis is to never ski over your skis. In other words, never stretch your numbers beyond what they are able to tell you. These numbers DO NOT tell us which of these quarterbacks is actually the best quarterback. There are some serious math issues that any good professional statistician could pick apart. The primary one is that each of these numbers is scaled differently, so combining them creates what we might label as fidelity issues.

All that being said, there is some good news and bad news when looking at this table. Unless you are one of the elite quarterbacks, I am assuming you would love to be compared to Sam Darnold. Like Stroud, Bo Nix has been a fixture in the playoffs his first two seasons and Trevor Lawrence is coming off of arguably his best season as and NFL quarterback.

People also should misunderstand the point here. The point is not to say that the Texans should have jettisoned Stroud to sign Kirk Cousins to Tua Tagvailoa. While cheaper, there is serious rot in their game, so you likely would not have gotten the results you see above. Plus, there are leadership questions for Tua at the very least. What these numbers demonstrate is that Stroud at his best is probably similar to all of these guys when they were at their best.

The Dolphins, Vikings (twice), and Jaguars all had to look at these quarterbacks and decide if they wanted to commit to them long-term. I would probably argue that the Dolphins and Jaguars probably wish they hadn’t. The Vikings probably wish they had in Darnold’s case. Cousins was the classic “good enough to get you there, but not good enough to win it” quarterback. He has yet to sign and could be a fringe starter at this point for a bad football team or a backup for a good football team.

No one can deny that Stroud is in the same boat. He is different that Lawrence, Cousins, and Tua, but he is similar in that the Texans aren’t quite sure they can get over the hump with him. That’s not an indictment on him necessarily. There aren’t many quarterbacks that can win the Super Bowl in any given season. Like Darnold, most need everything to break right for them. The history of the sport is littered with enough Jim Plunketts, Doug Williams, and Joe Flaccos to discount the idea that a quarterback of this level could possibly win it. Naturally, those quarterbacks weren’t breaking the bank though. They had to have a championship roster around them and just a little bit of luck along the way. That is where the Texans are with C.J. Stroud.

Read full story at source