In the Trenches - Week 8
Each week I’ll be using data from PFF and around the web to give some insight into IDP fantasy football, and how using these metrics can lead to finding undervalued players, potential breakouts and sleepers.
One of the joys of writing for PFF is the wealth of statistics we have access to, and how segmented they are. That allows fantasy analysts such as myself the opportunity to study trends and data patterns in a bid to uncover fantasy value and indicators of future performance. One of the most notorious defensive positions for fluctuating performance and the transiency of their value is defensive back, so this week I wanted to take some time to look at how players have been performing in 2014 and see if we can find some key indicators of performance.
Not all fantasy football leagues are created equal, and there are a multitude of scoring systems and league setups that can complicate the waters somewhat when talking about IDP targets. Take this week’s column’s featured position, defensive back, which can be a catch-all position, or separated into cornerback and safeties depending on league rules.
That’s an important distinction to begin with, and IDP gamers should be aware that safeties are the more reliable point scorers with their typically higher tackle totals, while playmaking cornerbacks can be expected to have a higher scoring ceiling in any given week due to a higher opportunity snag a pick-six. All that is well and good, but knowing one position is consistent and another is boom-or-bust doesn’t help owners plug the best options into their line-ups.
One statistic that we collect at PFF and has potential to be an indicator of fantasy success is the number of targets a defensive back faced. The next question though is, how dependable are targets in forecasting fantasy success? To determine this we need to look at the correlation coefficient between targets and fantasy points. Correlation coefficient (r) can be summarized as follows;
A correlation coefficient is a number between -1 and 1 which measures the degree to which two variables are linearly related. If there is perfect linear relationship with positive slope between the two variables, we have a correlation coefficient of 1; if there is positive correlation, whenever one variable has a high (low) value, so does the other. If there is a perfect linear relationship with negative slope between the two variables, we have a correlation coefficient of -1; if there is negative correlation, whenever one variable has a high (low) value, the other has a low (high) value. A correlation coefficient of 0 means that there is no linear relationship between the variables.
For example, last week we looked at linebackers and tackle frequencies in order to identify fantasy value. Although tackle frequency (r = 0.5263) is an excellent metric for showing how efficient a player can be at producing fantasy points, it’s not a highly correlated to having a high fantasy points total because you can be a run-stuffing two down linebacker on a limited snap count, or get injured, and therefore not have enough opportunity to rack up fantasy points., while still maintaining a high tackle frequency.
Cases in point would be Koa Misi (17.74 percent TakFreq and 15.5 fantasy points), Jasper Brinkley (16.44 percent and 48 Fpts) or Mason Foster (15.57 percent and 25 Fpts). A better way of identifying fantasy success would be looking at a linebacker’s number of snaps played (r = 0.7320), or total tackles (r = 0.8740) whereas forced fumbles (r = 0.3176) is obviously not a good predictive statistic.
Ross Miles is a Lead Writer for Pro Football Focus Fantasy. Follow him on Twitter – @RossMilesNFL
Need more specific advice, start/sit help or expert opinions on trades? Let us know on Facebook by joining the conversation