Visualizing alternative outcomes in fantasy football
Book quiz data geekery, plus another free book

Just one change evokes an entirely new world

Before I get to normal programming, please note that today (Friday) is the last day to enter the contest to win my new book. Only three easy questions, and you may get a nice summer read, with my autograph. Enter here.

New: the sample pages are now on Slideshare as well, so no need to download PDF.

***

Yesterday, I showed a chart of alternative outcomes for a fantasy football team for each week of the season. I looked at whether the team owner activated the right set of 9 players, from a roster of 14. The area of the histogram to the right of the vertical line is the probability of scoring more points than the squad that was activated. The smaller the area, the better was the owner's performance.

In the chart today, I switched one thing... what the vertical lines represent. In the following chart, the (pink) line represents the score compiled by one's opponent during each week. (The opponent changes each week but the schedule is fixed at the start of the season.)

Ad_range_opponent

This is a completely different chart. This chart tells you a little about the luck of the draw.

Look at Week 4. The opponent activated a really wretched squad. No matter what this team owner does, he/she will score more points than the opponent.

Now look at Week 3. No matter what this owner does, he/she is bound to lose because the opponent scored more points than his/her best possible squad.

The area to the right of the line is the probability of beating your opponent.

More to come.

Comments

Matt Hudson

Interesting stuff. What fantasy league did you use? Simple enough to extract the data? Planning to share the R code anytime?

Kaiser

Matt: in the chapter, I analyzed my friend's league that is hosted at ESPN. I explained how I assembled the data set in one section of the chapter. It's not simple but can be done with a little guidance. Once I'm done with my series of charts, I will put up a dataset for those interested in visualizing the data. I can put up the R code but I'm just using basic ggplot2 here.

The comments to this entry are closed.