What in the hell is the NFL EloRater? It’s “a community-based project with the goal of rating the best players in NFL history,” according to Pro-FootballReference.com, who is running the project.
Fans who participate will be shown two players’ names, and are given applicable statistics of each player. They then choose which player they believe to be better after examining the data presented. A rating system with “its roots in chess” is then applied to analyze and compile the player ratings and rankings. Anybody can participate in the EloRater project.
As you might expect, alumni of the Miami Hurricanes are well-represented in the rankings. Ray Lewis is currently eighth all-time, and first among active players. Ed Reed is seventh among active players, and Reggie Wayne is ninth. Lewis is second all-time among defenders, and Reed sits 30th. Former Hurricanes quarterback Jim Kelly is ranked in the top 25 all-time among offensive players.
Other Canes in the top 100 all-time rankings include Edgerrin James on the all-time offense list, and Ted Hendricks, Cortez Kennedy, and Warren Sapp on the all-time defense list.
The ratings change by the minute, so rankings reported above may no longer be entirely accurate. The project began on August 22.
Here is more specific information about how the EloRater works. *Warning: Unless you’re the guy from A Beautiful Mind, you might get a migraine.
"The Elo rating system is a method for calculating the relative skill levels of players in two-player games. The creator of the system, Arpad Elo, was a professor of physics at Marquette University who wanted an improved chess rating system. Although the system has its roots in chess, today it is used in many other games.We decided to take Elo’s work and apply it to the ranking of football players. Our player pool consists of 1500 NFL players: the top 1000 in AV post-1975 and the top 500 in AV pre-1975.All players have an initial rating of 1500 points. These ratings are then updated by randomly selecting pairs of players and having them “play” each other.We start by computing the win probabilities for each player (let’s call them A and B):P(A wins) = 1 / (1 + 10^((RB – RA) / 400))P(B wins) = 1 / (1 + 10^((RA – RB) / 400))where RA = rating for A RB = rating for BAfter the winner has been determined, the ratings of the two players are adjusted. If A wins the match then the new ratings are:RA_new = RA + K * P(B wins)RB_new = RB – K * P(B wins)The “K” above stands for the “K-factor” and has a value of 10.While if B wins the match then the new ratings are:RA_new = RA – K * P(A wins)RB_new = RB + K * P(A wins)For example, suppose Tom Brady (“TB”) has a rating of 2500 and Peyton Manning (“PM”) has a rating of 2450. The win probabilities for each player are:P(TB wins) = 1 / (1 + 10^((2450 – 2500) / 400)) = 0.571P(PM wins) = 1 / (1 + 10^((2500 – 2450) / 400)) = 0.429If TB wins then the new ratings are:TB_new = 2500 + 24 * 0.429 = 2510PM_new = 2450 – 24 * 0.429 = 2440While if PM wins then the new ratings are:TB_new = 2500 – 24 * 0.571 = 2486PM_new = 2450 + 24 * 0.571 = 2464Here are a few more notes about the way we have chosen to implement the Elo rating system:This is a community-based project with the goal of rating the best players in NFL history. In each matchup, the user should choose the player who they believe was the better player. It is up to the user to determine how much weight to give to offense versus defense, peak value versus career value, regular season versus playoffs, etc.Pairs are not chosen completely at random. The first player is randomly selected to begin the process. Following that, a second player with a rating within 250 points of the first player is randomly selected to complete the pair. We did this in order to prevent bizarre choices (e.g., Wayne Chrebet over Jerry Rice) from distorting the ratings."