How do you compile the rankings and determine where a player is ranked?
In 2008, we launched Fantasy Football Nerd with the goal of not just aggregating fantasy football rankings and projections, but with the added twist of weighting them based upon each source's accuracy.
If you're familiar with the concept behind Google's Page Rank algorithm, the core concept of Nerd Rank should immediately present its value. If the rankings or projections from a particular website have demonstrably proven to be more accurate than others, then we deem the rankings and projections from that site to be more authoritative - thus we give those rankings and projections more weight than those less historically accurate.
In short, we aggregate rankings and projections from a number of sites around the net - most you already know, some you should get to know. We record their rankings & projections and then compare them to what actually happened in the games. We use that data to create a Nerd Rank score for every site based upon how accurate each of them are. A single site has multiple Nerd Rank scores depending upon how accurate they are for each position.
The simple truth is that no single site that we aggregate from is the most accurate across the board, so we weight and rank each by position and type (weekly or draft rankings). The best predictor of QB's will have the most influence on our QB rankings. Best predictor of RB's will be most influential on our RB rankings, and so forth. Weekly rankings are treated independent from draft rankings.
We are the fantasy sports industry's only weighted consensus. More than one million people used our technology last season to win their leagues.
* In 2021, Fantasy Football Nerd became FantasyNerds.com. The Nerd Rank algorithm has been applied to football, baseball, and basketball.
There are three main components to the Nerd's ranking system.
The Nerd Crawler is quite fast and very efficient. Its job is to simply visit each site and report back on the rankings that each site has applied to all of the players. The details from every crawl are recorded for use down the line. FN then caches the results so that the Crawler can remain bandwidth-friendly to all of our sources.
The original algorithm had more than a dozen factors influencing the quality score for a specific website. As you might expect, the algorithm has been modified over time to help us determine what the most accurate experts are predicting. Based on past performance, how accurate was each source? How did each source trend during the season (ie, draft rankings vs season performance, weekly projections vs that week's performance, etc.) and how did each expert do overall across all positions as well as by individual position? Our research shows that some experts are consistently better by specific player position. That's important when calculating their Quality Score and applying that QS to their projections. It allows us to really factor in their specific accuracy.
The higher the Quality Score, the more "votes" that site gets during the final stage of calculation and the more influence it has over the final outcome. Every site gets a chance to "vote", but the sites that are accurate more often get more "votes" in the final tally. Nerd Rank simply considers them to be more authoritative.
Additionally, if a source site performs poorly and their projections are either radical or detrimental to the consensus, a Quality Score limits their impact on the final outcome.
Nerd Rank is the process that combines all the data from the previous steps factoring in each source rank (r) and the associated Quality Score (QS).
Trending is also very important. Remember earlier when we mentioned how the Nerd Crawler keeps a record of each player from each source from each visit? Well, we can use that data to see how the sources feel about a certain player over a period of time and we can compare that to other players in the same tier. The QS for each source changes as more data is obtained and is recalcuated every week for each source.