Switching Teams; Fantasy Wide Receiver Style

Dave Cherney / Guest; Dr. John Bush

The question of players moving between teams has been debated endlessly in the fantasy football universe. It keeps the pundits busy during the off season. We decided to investigate this question to determine if there is a difference on average for wide receivers moving between teams and the following performance changes, if any, that occurred.

Fantasy just got a whole lot better thanks to Monkey Knife Fight. With fast-paced games like Rapid Fire and Either/Or, it’s never been easier to play fantasy and win. New to MKF? Get Exclusive $100 Deposit Match + Free $5 Game >

This question leads to an understanding that the Planning fallacy is at work.

The planning fallacy is a phenomenon of most decision makers that is composed of three characteristics. They will have an unrealistic understanding of how much time will be needed to complete a future task; they display an optimism bias and will also underestimate the time and resources needed to complete the task.

We believe your drafts can be considered a plan in a sense and thus we relate fantasy football drafting to the concepts associated by the planning fallacy and human optimism/pessimistic bias. Note when you consider your drafting process you are very susceptible to the optimism bias. When other fantasy football owners judge your draft, they will show an opposite; the pessimistic bias! The definition includes both risk and expected benefits to be included.

These benefits are your expectations that a player will improve with a new team. You draft a “hot player” with that thinking in mind.  In answering the main question, we must turn to one method to cut through the optimism by using Reference Class Forecasting for an outside viewpoint to balance our thinking.

Reference class forecasting or comparison class forecasting is a method of predicting the future by looking at similar past situations and their outcomes.

Based on these established methods we turned to finding our own reference class data for wide receivers switching teams vs. those remaining with their current squad. Previously we have developed a player performance numerical score to compare all players in all positions. Our PSN (Positional Standardized Number) score goes from 100 (more on occasion) to 0. This scale annotates a player from the highest to lowest player performances of the week or season

Materials and Methods

We have an established a database of PSN for all QBs. RBs, TEs, and WRs from 2013 to 2016. We assembled the WR data from that database and divided the PSN achieved between years noting if a WR had switched or not switched teams the preceding year. We then used yearly player movements from 2013 to 2014, 2014 to 2015, and 2015 to 2016.

Another aspect of our research methodology was that we separated the WRs into 3 groups based on their PSN from the preceding year. WR that were between 40 and 100 got assigned the “high performance label” These would be the best players; all of a team’s WR1s maybe some WR2s, and rarely WR3s.  WR’s between 11 to 39 are labeled as “mid performance” while WRs below 11 were considered the “low performance” players

Next, after we divided the players into the High, Mid and Low PSN levels. We then determined if they switched teams or did not switched teams between the seasons. The team-switch PSN averages were calculated as the amount of difference between the beginning year (old team) and the following year (new team). The data calculated for each player could be either positive (improvement) or negative (decreased performance). Those differences were graphed and plotted in the below figures (Low, Mid and High)

Data Presentation and Analysis

Swithing Teams 01

Swithing Teams 02

Swithing Teams 03

Group Averages, T Test Stat, Probability Scores

We now turn to a statistical method to validate these initial conclusions. This is the first time that we are aware of that statistical analysis will used in testing the effects of team switching instead of opinions of pundits.

The Student’s T Test for unpaired data is the first stat test used in this kind of comparison. The t-test assesses whether the means of two groups (Switching Team WRs vs Not Switching Team WRs) are statistically different from each other. This analysis is appropriate whenever you want to compare the means of two groups.

In most research, the “rule of thumb” is to set the alpha (risk level) at .05. This means that five times out of a hundred you would find a statistically significant difference between the means even if there was none (i.e., by “chance”). In other words, it’s a false positive. The T Test gives a probability score and we must have that score to be equal to or below 0.05 to believe the 2 groups are really different.

Below we present a chart of data containing our group’s averages for each WR performance level, Numbers of WRs, The T Test stat, the Probability Score and a conclusion.  Yellow highlights mean no difference in player performance from team switching. Green annotation highlights a positive player performance from switching teams!

Group Averages, T Test Stat, Probability, Conclusion

Swithing Teams 04

Discussion and Prediction for 2017

Given the clear results of High PSN WRs switching team, we determined and present a list of the 2017 receivers who changed teams this season. We have identified five players who meet the High Probability category (Green Labels) along with several others which come with the ‘buyer beware’ label (Red Labels). Yellow labeled are players whose 2017 may be average or so.

Swithing Teams 05

Swithing Teams 06


  • WRs moving between teams that performed at Mid to Low levels in PSN did not improve in their new surroundings! Of the 23 candidates listed, we are recommending five players. Any difference in the group average was by chance and/or exceptional opportunity and not significant!
  • WRs moving between teams that had preformed the preceding year at 40 PSN or better can be expected on average to improve 14.3%. We are expecting Terrell Pryor to increase from 13.15 points per game to 15.03. Now your research can focus on the only WRs that particular group moving to new teams.
  • If you can predict what players improve by more than 14.3% then you are ahead of the average of public opinion. This is your personal benchmark.
  • In years past if your analysis focused on a WR that was moving to another team and you were right but only by 14.3% or less. You really have done no better than average and maybe even worse.
  • More research is needed for conformation of these conclusions.

Dr. John Bush (@prof_fantasy1)

Full Professor of Biology

In fantasy football, John writes for his blog the Fantasy Sports Professors and has published along with Dave Cherney a fantasy football textbook

2017 Winning Dynasty and Redraft Fantasy Football Drafts (Amazon eBook).


Leave a Reply

Your email address will not be published. Required fields are marked *