Fantasy Football 2016 - Building a Data-Driven Draft

Author: | Category: BI Office, Fantasy Football | Tags: Fantasy Football 2016, data-driven, draft strategy, data analytics | Published: 8/31/2016

Fantasy Football 2016 - Building a Data-Driven Draft

Fantasy Football Arbitrage

Like most businesses, fantasy football coaches are drowning in data. They show up to their draft after countless hours of research with the “perfect” list of players ranked by scoring projection, mock draft ranking, and personal preferences. Other coaches show up similarly armed. They draft as close to their script as they can, then spend the season convinced they made the best decision they could, based on the data available. However, many coaches are using the right data in the wrong way.

Fantasy football is a data-driven activity. Fantasy coaches, like most businesses, have access to massive volumes of data. Every fantasy football website, magazine, blog, and newspaper publishes scoring projections, rankings, and draft lists based on historically structured and unstructured data. Experts use this data to perform countless mock drafts to gather intelligence on each player’s average draft position and popularity. Coaches use the mock draft lists to make the best possible picks.

Fantasy football is one of the largest, most successful, and best funded Open Source Decision Support Systems (DSS) in the world, with a nearly perfect adoption rate and thousands onboarding every season. In other words, everyone that plays fantasy football uses data to make decisions.

The problem is, as with any DSS, you have to contextualize the data to make it relevant to your environment. The data gathered, generated, and published around fantasy football is based on a specific scoring system. There is a “Standard” fantasy football scoring system that details how many points a quarterback or running back receives for things such as touchdowns, yards gained, and points per reception (PPR). The scoring system is very intricate and awards or revokes points based on dozens of criteria. The number of points given/taken for each scoring category directly impacts the projected scoring potential of each player.

Figure 1. Pyramid Analytics Fantasy Football Advisory dashboard

fantasy football dashboard

This “Standard” scoring system is the basis for most fantasy football projections and rankings. However, most fantasy football leagues use a customized scoring system. Most coaches show up to their draft with a list that does not reflect their league scoring rules at all, and draft the wrong players.

This is where arbitrage comes in: “the simultaneous buying and selling of securities, currency, or commodities in different markets or in derivative forms in order to take advantage of differing prices for the same asset.” Coaches can use arbitrage, and exploit the differences in the scoring systems, to draft smarter.

To solve for this, we combine transactional data from Fantasy Data with expert mock draft data from Fantasy Football Calculator to get an Average Draft Position (ADP). We then add in analytics from BI Office to assemble a ranking list based on the scoring projections from multiple fantasy football providers (Fan Duel, Draft Kings, Yahoo, and others) that use unique scoring systems. Players are ranked based on the score projection of the selected system. Then a simple calculation shows us the difference between our ranking and the mock draft ranking to provide a Fantasy Draft Advantage metric.

As we slice and dice this data, we see many players drafted very high in mock drafts, but who are actually ranked much lower based on the rules of the selected scoring system.

Antonio Brown is a perfect example of this. His average mock draft ranking across all providers is #1. In the Draft Kings scoring system, he is the #2 ranked player, with only Cam Newton ahead of him. Draft Kings is a WR-friendly scoring system, with three WRs in the top 10. But if we look at the same list based on the Yahoo scoring system, we see him ranked #13 overall, putting him into the second round in most leagues.

Figure 2. Draft Kings scoring dashboard view

fantasy football scoring dashboard

Figure 3. Standard scoring dashboard view

fantasy football scoring dashboard

We see this pay off by when we let others draft overrated players, while you choose undervalued players. Ryan Tannehill is a great example of this. Based on Standard System mock drafts, Tannehill is ranked #141, which means he will be available late in most drafts. However, he is a top 20 pick in the Fan Duel scoring system, and will reward those who select him sooner in these leagues.

While fantasy football decisions are extremely data-driven, and coaches have unlimited access to a massive, mature, well-curated DSS, they still make decisions without the necessary context and relevance to ensure success.

Pyramid Analytics provides that context and relevance to any data-driven decision making, putting your team in a better position to win, week in and week out.

Disclaimer: The Fantasy Football-related content on this website is for general information purposes only. Any reliance you place on such information is therefore strictly at your own risk.


Related resources

  1. Follow us @PyramidAnalytics on Twitter for more Fantasy Football insights with #BIOffice
  2. Learn about BI Office analytics platform
  3. Read our blog post Microsoft has named Pyramid Analytics the 2016 U.S. Public Sector Health & Life Sciences Solution Partner of the Year

Add your comment below

 

Be the first to comment


Stay up-to-date with product announcements

Popup Content