Inflection Points: Another Measure of Consistency

Scott Spratt looks at which players have been more and less consistent than expectations over the last few years.

| 4 years ago

Scott Spratt looks at which players have been more and less consistent than expectations over the last few years.

Inflection Points: Another Measure of Consistency


I’ve devoted a lot of time and energy to valuing the fantasy production of players, but that value only means something if you can capture it. How many owners left Bryce Brown on their waiver wires in Week 12 and then on their benches in Week 13, missing out on 347 yards and four touchdowns, only to play him the next two weeks when he ran for 40 yards and zero scores? Chances are that, no matter how unlucky you were with injuries last season, between your roster and the waiver wire, you had access to enough points to win your league. You either won or lost because of your decision-making.

You already make decisions based on the volatility of players, even if you’ve never thought of it in those terms. Consider Vincent Jackson. When he was in San Diego, he was as likely to give you a 0 as he was a 20. The only score he never seemed to get was his actual average, but that was ok. If you owned Jackson, then you trained yourself to insert him into your lineup every week because you knew that, over the course of the season, he would reward you with a top-ten receiver.

That is why consistency is not enough to drive your decision-making. When Jackson scores a handful of points over a couple of weeks, you know not to worry about it. The same may not be true for a traditionally-consistent player or for one with a lesser track record. The likelihood that you will either absorb poor production or miss out on outstanding production depends on the shortfall of a player’s actual consistency versus the expectation you have for his consistency.

I’ve tried to measure that shortfall in a statistic I call inflection points. Inflection points are times when a player’s fantasy production either falls well short of his average or greatly exceeds his average in consecutive weeks. Those are the times when you are at the greatest risk of losing out on potential points by playing a player when you shouldn’t, not playing a player when you should, or trading a player at less than his full value.

The calculation I used for inflection points carries a lot of complexity. The heart of it is a comparison of a player’s fantasy production in consecutive weeks to his average production. If he has less than his average minus one half of a standard deviation or has more than his average plus one half of a standard deviation for two weeks in a row, then I count that as a point.

From there, I tried to build in a typical decision-making framework. I used a player’s previous season’s standard deviation as an approximate measure of the perception of his volatility. For most of the season, I considered a player’s actual average for that season, but for the first four weeks, I used his average from the previous season, which is the greater part of his actual perceived value for the first month. Finally, I filtered out players that did not have a sufficient track record of fantasy success, since we would not have expectations for their volatility that they could fail to meet.

Here are the results:

  Fantasy Points Inflection Points
Player Mean StdDev CoeffVar Opps Negative Positive Total Rate
Pierre Garcon 8.2 5.6 0.69 21 7 3 10 0.48
Drew Brees 24.8 4.8 0.19 28 6 7 13 0.46
Jimmy Graham 10.6 4.5 0.43 13 5 1 6 0.46
Vernon Davis 7.4 5.7 0.77 26 9 3 12 0.46
Fred Jackson 14.5 8.3 0.57 14 2 4 6 0.43
Marques Colston 11.6 5.9 0.50 25 7 3 10 0.40
DeAngelo Williams 8.7 3.5 0.41 28 7 4 11 0.39
Jordy Nelson 11.1 8.8 0.79 8 2 1 3 0.38
Calvin Johnson 15.7 7.2 0.46 28 6 4 10 0.36
Ronnie Brown 3.5 3.6 1.06 20 5 2 7 0.35
Adrian L. Peterson 15.8 9.0 0.57 23 4 4 8 0.35
Mike A. Williams 7.3 4.9 0.67 26 5 4 9 0.35
Aaron Hernandez 8.8 6.0 0.68 6 1 1 2 0.33
Aaron Rodgers 26.1 8.1 0.31 27 6 3 9 0.33
Brandon Marshall 9.6 5.4 0.56 28 5 4 9 0.32
LeSean McCoy 17.9 6.7 0.37 22 5 2 7 0.32
Sidney Rice 7.6 4.4 0.58 13 2 2 4 0.31
Tom Brady 23.5 6.5 0.28 28 4 4 8 0.29
Chris D. Johnson 11.5 8.1 0.70 28 6 2 8 0.29
Tony Gonzalez 7.7 4.6 0.59 28 5 3 8 0.29
Larry Fitzgerald 11.3 4.8 0.43 28 6 2 8 0.29
Percy Harvin 10.9 5.7 0.53 22 2 4 6 0.27
Dez Bryant 9.6 5.5 0.57 26 4 3 7 0.27
Brandon Lloyd 8.8 7.2 0.81 26 4 3 7 0.27
Ahmad Bradshaw 13.1 6.1 0.46 19 3 2 5 0.26
Antonio Gates 9.6 6.8 0.70 23 4 2 6 0.26
Rob Gronkowski 13.9 8.3 0.60 8 1 1 2 0.25
Dallas Clark 6.8 5.0 0.75 20 2 3 5 0.25
Steven Jackson 12.3 4.2 0.34 26 5 1 6 0.23
Andre Johnson 12.1 7.3 0.61 18 3 1 4 0.22
Cam Newton 22.7 7.5 0.33 14 1 2 3 0.21
Marshawn Lynch 15.4 7.0 0.45 14 2 1 3 0.21
Torrey Smith 8.5 9.1 1.07 14 2 1 3 0.21
Darren Sproles 11.0 5.0 0.45 10 1 1 2 0.20
Mario Manningham 7.3 6.9 0.95 15 1 2 3 0.20
Antonio Brown 8.4 4.6 0.54 10 1 1 2 0.20
Malcom Floyd 9.7 7.8 0.81 21 2 2 4 0.19
DeSean Jackson 9.0 8.7 0.96 21 3 1 4 0.19
Miles Austin 10.0 6.8 0.67 21 2 2 4 0.19
Ryan Mathews 12.7 7.4 0.58 21 2 2 4 0.19
Greg Jennings 11.7 7.7 0.65 16 2 1 3 0.19
Matt Forte 14.1 8.5 0.60 22 4 0 4 0.18
Steve Johnson 9.2 7.6 0.82 28 2 3 5 0.18
Wes Welker 12.3 5.7 0.47 28 2 3 5 0.18
Michael Vick 22.7 10.7 0.47 17 2 1 3 0.18
Santana Moss 7.6 5.5 0.72 23 2 2 4 0.17
Maurice Jones-Drew 16.1 7.2 0.45 18 2 1 3 0.17
Nate Burleson 6.7 6.0 0.91 18 3 0 3 0.17
Mike Tolbert 10.2 6.2 0.61 26 4 0 4 0.15
Pierre Thomas 8.7 4.7 0.54 27 3 1 4 0.15
Michael Crabtree 9.6 4.4 0.45 14 2 0 2 0.14
Reggie Bush 11.2 6.5 0.58 14 2 0 2 0.14
Frank Gore 12.0 6.5 0.54 28 1 3 4 0.14
Roddy White 11.4 7.0 0.61 28 3 1 4 0.14
Donald Driver 5.1 4.4 0.86 14 1 1 2 0.14
Tony Romo 19.7 7.0 0.36 14 1 1 2 0.14
Steve Breaston 6.0 4.7 0.78 14 1 1 2 0.14
Steve L. Smith 10.2 3.8 0.38 28 1 3 4 0.14
A.J. Green 12.2 4.5 0.36 14 1 1 2 0.14
Victor Cruz 11.1 8.8 0.79 14 1 1 2 0.14
Jason Witten 8.2 5.7 0.70 28 4 0 4 0.14
Demaryius Thomas 11.7 7.6 0.65 14 1 1 2 0.14
Reggie Wayne 8.2 6.4 0.78 28 3 1 4 0.14
Anquan Boldin 7.7 8.3 1.08 25 2 1 3 0.12
Lance Moore 8.0 6.1 0.77 25 3 0 3 0.12
Mike Wallace 11.0 7.0 0.64 27 2 1 3 0.11
Knowshon Moreno 4.9 7.2 1.48 10 0 1 1 0.10
Hakeem Nicks 11.4 7.5 0.66 21 1 1 2 0.10
Jeremy Maclin 9.2 7.2 0.78 22 2 0 2 0.09
Ben Roethlisberger 17.4 7.5 0.43 22 0 2 2 0.09
Dwayne Bowe 10.0 10.8 1.08 25 1 1 2 0.08
Philip Rivers 19.0 5.6 0.30 28 2 0 2 0.07
Vincent Jackson 11.4 9.5 0.83 14 0 1 1 0.07
Davone Bess 5.1 4.2 0.82 14 1 0 1 0.07
Matt Ryan 20.6 4.6 0.22 14 0 1 1 0.07
Michael Turner 13.6 8.0 0.59 28 2 0 2 0.07
Rashard Mendenhall 11.6 5.8 0.50 15 1 0 1 0.07
Darren McFadden 16.3 11.3 0.69 15 0 1 1 0.07
Deion Branch 7.4 6.9 0.93 16 1 0 1 0.06
Ray Rice 17.6 8.2 0.47 28 0 0 0 0.00
Arian Foster 19.8 10.0 0.51 24 0 0 0 0.00
Julio Jones 11.7 8.4 0.72 14 0 0 0 0.00

 

I was surprised by some of the names on either end of the list, but that was the point of the exercise. The Saints have three of the top-six names, highlighted by Drew Brees and Jimmy Graham, both of whom are considered to be among the most consistent players in fantasy. Since they are both near the top of their positions, there is little real risk of incorrectly benching either one of them, but notice how Brees has a near-even split of positive and negative inflection points. That indicates that Brees has been as likely to go on hot streaks as cold streaks in recent seasons, which, apart from the poor timing of his terrible run leading into the fantasy playoffs in 2012, you can live with.

Meanwhile, Graham has five negative inflection points against only one positive one. The owners who drafted Graham in the first couple of rounds in 2012 because of his consistent excellent play in 2011 failed to get it. Injuries were likely a factor, but I contend that Graham is a lot more tight end-like than most owners believe. He has produced inflection points at nearly double the rate of Rob Gronkowski, albeit it in a small sample of opportunities for both.

For me, the most interesting names are the borderline starters, since they are the players you are the most likely to start and sit at the wrong times. Pierre Garcon’s resume was built in a Colts offense that is thankfully behind him, but Vernon Davis and Chris Johnson are players I would devalue because of their declining consistency.

In contrast, Ray Rice and Arian Foster demonstrate their worthiness for consideration for the top spot in fantasy with zero inflection points in 52 combined opportunities over the last two seasons. There has never been a reason to doubt either of them in-season. In addition, Matt Ryan, Steve Smith, Roddy White, Frank Gore, and Reggie Bush have all been tremendously consistent.

Be careful in how you interpret the bottom of the list, especially. Players like Darren McFadden, Michael Turner, and Vincent Jackson have few inflection points because they’ve already established themselves as volatile players. I included both approximate player averages and standard deviations to give you a basis for comparison. The most meaningful players are those with many inflection points despite a high coefficient of variation—standard deviation as a percentage of mean—and those with few inflection points despite a low coefficient of variation.

Scott Spratt also works for Baseball Info Solutions and writes for The Hardball Times.  Feel free to ask him questions on Twitter – @PFF_ScottSpratt

Find more of our Dynasty Content here.

Comments are closed.