Home | Around the NFL | Fantasy Football Predict-Ability: Pairing Consistency with Upside

Fantasy Football Predict-Ability: Pairing Consistency with Upside

LeSean-McCoy
Team Stake

Fantasy Football is a weekly game. This is a notion that can often be overlooked during the offseason as talent evaluation and season long projections become the focus of fantasy football prognosticators. During the NFL season, though, the focus quickly switches to weekly match ups and start/sit decisions. How we make start/sit decisions on a weekly basis can ultimately make or break our fantasy teams. Aside from examining a players individual match up in season; how can we approach drafts in a way that supports us in making easier start/sit decisions once the season begins?

When approaching start/sit decisions the terms “floor” and “ceiling” often are thrown around as a means to describe a player’s weekly potential.  Ultimately, these two terms are generally accepted to describe how predictable a players weekly production is (their “floor”) and how much upside a player has to post a large amount of fantasy points on a given week (their “ceiling”). In episode 29 of The Late-Round Podcast, JJ Zachariason explains the use of coefficient of variation as a means to determine a player’s floor or more specifically their weekly predictability: “(coefficient of variation) tells us how much a set of data varies, or the standard deviation, divided by the set of data’s mean, or average”. In fantasy football evaluation a player’s points per game is often referenced as a means of evaluation.  Coefficient of variation allows us to go a step further, from just looking at a player’s average to looking at how consistently they stay near or exceed that average. This becomes important because we don’t just want players who score consistently but who score a large number of points consistently. For instance, a player who scores five points every game is more consistent than one that scores 10 points one week, 14 points the next, and 19 points the week after; but that consistency isn’t valuable to us in fantasy football because if a player only scores five points in our starting line up each week our heads might explode out of frustration.

Having players that are more predictable allows us to have the floor that makes our teams competitive each week, but the thing that seems to excite fantasy players the most is a players ceiling. Ceiling and upside are terms that get attached to different players constantly, but how often do these upside players really post weeks at the top of their position? Tod Burros examined this concept in his Heatmap series for Rotoviz, in which he illustrates how often different players hit certain point thresholds. Being able to identify players who are more predictable and understanding a players weekly ceiling allows us to make better start/sit decisions and build our fantasy teams to have a mix of consistency and upside.

In this article, the sample size used for all players is their last 15 games played (excluding week 17). All games in which a player played less than 25% of the game were excluded from the sample. The scoring used for this sample is point per reception scoring, one point per 20 yards passing, and four points per passing touchdowns.

QB Elite Tier (Consistency and Upside)

Quarterbacks that fall into the Elite tier possess a coefficient of variation under 35 while being in the top 12 of their position greater than 40% of the time, the top 6 greater than 25%, and posted at least one top 3 performance.

The Elite tier comprises some of the best of the best of the quarterback position and is the ideal archetype of the reliable quarterback in your weekly redraft league.  These quarterbacks poss the unique ability to be consistent while maintaining elite weekly upside. Aaron Rodgers sets the standard for this type of production and it’s the reason why he is consistently drafted as the first quarterback in fantasy drafts. His ability to be consistent while producing elite performances is unmatched by anyone at the position. Even in Matt Ryan’s MVP season, he was unable to rival Rodgers weekly upside as Rodgers exceeded all other quarterbacks in both top 3 and top 6 performances. Kirk Cousins exceeded many of the presumed elite performers at the position, having more top 6 performances than anyone outside of Rodgers and Drew Brees. According to Fantasy Football Calculator’s average draft position, four of these quarterbacks are being drafted amongst the top six of the position. Meanwhile, Kirk Cousins, Cam Newton, and Matthew Stafford are all being drafted between the 8th and 10th rounds and each possess the consistency and upside of a quarterback you can rely on weekly.

 

 

QB Boom Tier (Inconsistent with Upside)

Quarterbacks that fall into the Boom tier possess a coefficient of variation above 35 while being in the top 6 greater than 20% of the time and posted at least one top 3 performance.

The Boom tier contains a few of the biggest names at the position in Drew Brees, Russell Wilson, and Ben Roethlisberger. These quarterbacks definitely have the potential to be every week starters on your fantasy team but might be a bit more inconsistent than is ideal. The benefit is that in good match ups, these quarterbacks still posses the ability to post huge fantasy performances. This group is ideal for best ball where weekly predictability matters less and upside is rewarded. Falling just outside of the Elite tier of quarterbacks, Drew Brees trailed only Aaron Rodgers in top 3 and top 6 weekly performances. Meanwhile, two of the most active rushing quarterbacks in the league, Tyrod Taylor, and Blake Bortles, fell just under a coefficient of variation of 35 due to their high weekly floor created by their rushing but also showed the ability to produce a handful of top 6 and top 3 performances. The takeaway from this is that rushing production for quarterbacks supports them with having a higher weekly floor, but it also makes it easier for them to produce elite weekly performances when they are successful both as a rusher and a passer.

QB Steady Tier (Consistent but Lacking Upside)

Quarterbacks that fall into the Steady tier possess a coefficient of variation below 35 while failing to perform in the top 6 greater than 20% of the time.  

Fantasy Football T-Shirts
Quarterbacks in this tier are great targets to pair with a Boom tier quarterback in redraft leagues. This tier of quarterbacks are going to be the most predictable weekly, but inside of the tier, that predictability has a lot of variance in regards to their weekly ceiling. For instance, Carson Wentz was an extremely consistent QB2 for fantasy, posting a coefficient of variation similar to Andrew Luck, but had the lowest volume of top 12 performances of any of the 28 quarterbacks I charted. On the opposite end of the spectrum, Carson Palmer was one top 6 performance short from being in the Elite tier right next to Cam Newton who had a very similar top weekly performance profile.  Andy Dalton has been more consistent than any other fantasy quarterback over their last 15 games. With analysts like TJ Hernandez outlining his expected positive touchdown regression that should come in 2017, Dalton is primed to be a fantasy quarterback that should have more high scoring weekly performances to pair with his elite consistency.

QB Unproven/Unreliable Tier (inconsistent and Lacking Upside)

Quarterbacks that fall into this tier possess a coefficient of variation above 35 and failed to perform in the top 6 greater than 20% of the time or failed to perform in the top 12 greater than 25% of the time.

This tier is a mix and match of quarterbacks that for one reason or another aren’t the most predictable and lack consistent weekly upside. Oddly, Brian Hoyer has a lower coefficient of variation than all but four of the 28 NFL quarterbacks included in this piece but he still somehow performed as a top 12 quarterback in over 50% of his appearances something only 10 other quarterbacks have been able to accomplish.  Alex Smith is a quarterback that tends to get the reputation of being very consistent with a high floor and a low weekly ceiling but he has not been nearly as consistent as perceived; ranking 19th in weekly consistency.  Trevor Siemian might be miscast in this tier as he has performed as the definition of boom/bust in his starts with the Broncos; ranking 13th amongst quarterbacks in top 6 performances while falling outside of the top 24 in nearly half of his starts.

RB Elite Tier (Consistency and Upside)

Running backs that fall into the Elite tier possess a coefficient of variation under 55 while being in the top 12 of their position greater than 33.3%, the top 6 greater than 20% of the time, and posted at least one top 3 performance.

The first thing that should jump out when looking at the running back position is the difference in consistency in comparison to quarterbacks. As we’ll continue to see, quarterbacks are by far the most predictable position. The top four of this Elite tier should jump out due to their extreme consistency amongst the elite of the position. The surprising omission from that group has to be Le’Veon Bell, who was held back due to his last three games in 2015 being included in the sample due to his three game suspension in 2016. If those games are removed from the sample, Bell’s coefficient of variation reduces to 38, just behind Ezekiel Elliot. Latavius Murray was one of the largest surprises in this tier, but his move from rushing behind Oakland’s elite offensive line to competing with rookie Dalvin Cook for touches in Minnesota should hold him back from repeating that performance. Eight of the nine running backs that currently have the highest average draft position in 2017 drafts are included in this tier, the three others are often being drafted in the 5th-8th rounds of fantasy drafts. Theo Riddick, Carlos Hyde, and Mark Ingram are all going to be competing for touches with additions to their respective backfields in 2017. However, what we’ve seen from them in their ability to produce elite performances on a consistent basis, make them all interesting targets in the middle rounds of fantasy drafts.

RB Boom Tier (Inconsistent with Upside)

Running backs that fall into the Boom tier possess a coefficient of variation above 55 while being in the top 12 greater than 20% of the time and posted at least one top 6 performance.

The first four names on the list should jump out due to their age and consistently elite fantasy production over the course of their careers. Matt Forte is the only one amongst the four that didn’t miss significant time in 2016 due to injury, yet he finds himself being drafted the latest of the four according to Fantasy Football Calculator’s average draft position. His production in 2016 was definitely inconsistent, but with four 20+ point fantasy performances he was definitely a positive asset on best ball rosters and could have similar boom/bust potential this season. It might surprise some that Jay Ajayi, who is currently being drafted as a top nine running back had the same number of 20+ performances as Forte a season ago. However, Ajayi posted those numbers while only being a featured running back in his last 10 games. Another running back that only had 10 games as a featured back might not even be able to be considered a feature running back at all in 2016 because he was only a running back for 10 games. Ty Montgomery played as many games at running back last season as Jay Ajayi was featured as the Dolphin’s top back, but during that span, he posted more top 12 performances than Ajayi.

RB Steady Tier (Consistent but Lacking Upside)

Running backs that fall into the Steady tier possess a coefficient of variation below 60 while failing to perform in the top 6 greater than 15% or the top 12 greater than 30% of the time.

This tier seems to contain each and every one of the limited usage pass catching running backs that maintain fantasy relevance. The ability to catch passes out of the backfield has a lot of similarities to the ability to rush the ball as a quarterback. In both situations, these attributes allow a player to increase their weekly floor, and thus their consistency while creating an easier path to elite weeks when these players find the end zone. Amongst the running backs in the Steady tier, Frank Gore is as steady as the elite of the position.  Gore has not missed a game in five seasons, which is just unheard of for an NFL running back, and with his ability to give consistent reliability he’s the ideal player to accompany a redraft roster that has a plethora of boom/bust players. Dion Lewis is probably the most interesting player to find himself in this particular tier as he didn’t exactly fit cleanly in any. Lewis has the unique status of maintaining a lower coefficient of variation than Jamaal Charles while posting a higher percentage of top 3 performances. When Lewis has been good, he’s been elite, but it’ll be interesting to see if he can recapture his elite production while on the low end of a running back by committee in New England. He’s a unique player that seems difficult to project as his average draft position ranges from the 13th round in Fantasy Football Calculator drafts down to the 18th in My Fantasy League best ball drafts. Ironically, Lewis might be better suited to be an addition to best ball rosters than redraft.

RB Unproven/Unreliable Tier (inconsistent and Lacking Upside)

Running backs that fall into this tier possess a coefficient of variation above 65 and failed to perform in the top 12 greater than 20% of the time.

For the running back position, this tier is full of unproven talents that definitely have upside if their situations break right. C.J. Prosise, Ameer Abdullah, and Derrick Henry are all young running backs who have been touted as having elite potential at the position but who have yet to show that production due to lack of touches or injury. Shane Vereen is a surprise omission amongst the pass catching backs in the Steady tier as he is much more inconsistent than his peers. However, the elite upside that comes with being a pass catching back has shown up for Vereen on a fair amount of occasions; as he is tied for 13th in percentage of top 6 performances at the position.

WR Elite Tier (Consistency and Upside)

Wide receivers that fall into the Elite tier possess a coefficient of variation under 60 while being in the top 12 of their position greater than 33.3% of the time and posted at least one top 3 performance.

Fantasy Football T-Shirts
The first thing that should jump out about this tier is that while the wide receiver position had the largest number of players included in this study, only nine fell into this Elite tier. The other thing that should jump out is that all nine of these wide receivers are currently being drafted amongst the top 12 of the position; often in the first two rounds of fantasy drafts. T.Y. Hilton surprisingly has led all wide receivers in percentage of top 6 performances over their last 15 games. This group had few surprises amongst it’s ranks, but one surprise was the lack of top 12 performances from Odell Beckham Jr. Odell tied for 9th in top 12 performances with three other receivers. Which isn’t a glaring issue but when deciding amongst the elite of the elite at the position it is definitely something that can give pause before immediately assigning him a top three wide receiver ranking.

WR Boom Tier (Inconsistent with Upside)

Wide receivers that fall into the Boom tier possess a coefficient of variation above 60 or produced a top 12 rate between 13.3% and 30%.

This tier, by far, contains the most players of any position and they greatly vary in both ability to produce elite performances and predictability. The top two of this tier, Julio Jones, and Keenan Allen, are in a six-way tie for the highest rate of top 3 performances and both have actually had more top 6 performances than Antonio Brown over their past 15 starts. The issue with both is their weekly consistency and when each fall off they have disastrous fantasy outputs. Over the 15 game sample used for this study, Jones had five occurrences of scoring 10 fantasy points or less and Allen had four. Both of these players have struggled with injuries, which likely have impacted both their ability to stay on the field and their ability to perform up to their elite talent when on the field. One of the surprises amongst this group has to be Golden Tate who is tied with seven other players behind only T.Y. Hilton for most top 6 performances. Tate is a player that often has the perception of being more of a floor play for fantasy, but in 2016 he was quite the opposite; posting four 20+ point performances and five performances under 10 fantasy points. Over his last 10 games, Tate averaged greater than 17 points and had a coefficient of variation under 52. Rookie, Cameron Meredith surprised many last season when he exploded on to the scene with the Chicago Bears. Meredith posted five performances of greater than 19 points in his debut season and had the same percentage of top 6 and top 12 performances as Doug Baldwin and Odell Beckham Jr.

WR Steady Tier (Consistent but Lacking Upside)

Wide receivers that fall into the Steady tier possess a coefficient of variation below 60 while failing to perform in the top 6 greater than 15% or the top 12 greater than 30% of the time.

Fantasy Football T-Shirts
Eric Decker’s coefficient of variation jumps off the page when examining, not just the wide receiver position, but every position.  Over their last 15 games, no player has been more consistent for fantasy owners than Eric Decker. Decker missed the majority of the 2016 season due to injury but he caught a touchdown in the each of the two games he finished. In addition to that, throughout the entirety of 2015 Decker only went scoreless on three occasions and during those three games, he produced 15.4, 12.1, and 18.1 fantasy points. Now Decker is teamed up with a young quarterback who has posted a historic red zone touchdown rate of 33.7% through his first two seasons and is in prime position to continue his outrageous touchdown consistency. Similar to how the pass catching back dominates the Steady tier for running backs; possession and slot receivers tend to make up the majority of this tier. The sheer volume of targets that they receive tends to boost their weekly floor. Two interesting receivers in this tier that seem to have a reputation for being more boom/bust but actually carry a lower coefficient of variation, are Mike Wallace and DeSean Jackson. Wallace has actually been the 8th most consistent wide receiver over his past 15 games.

WR Unproven/Unreliable Tier (inconsistent and Lacking Upside)

Wide Receivers that fall into this tier possess a coefficient of variation above 60 while failing to perform in the top 12 greater than 12% and the top 24 greater than 33.3% of the time.

The biggest name on this tier has to be Brandon Marshall who a year after being 4th in scoring amongst wide receivers in fantasy, failed to produce as a top 36 wide receiver in over 70% of his performances. Hopefully, his transition to New York and playing opposite Odell Beckham reinvigorates his career and his 2016 season will be remembered as an extreme outlier. Along with Marshall, two other Jets pass catchers appear on this list, so I think it’s fair to assume that situation played a large factor in their inconsistency and inability to produce boom weeks. DaVante Parker tops this list, with many hoping for a breakout season from the third year wide receiver; he’ll have to show both an increase in consistency and an increase in elite performances to be a reliable fantasy starter.  Hopefully, the loss of Ryan Tannehill becomes a blessing in disguise for him with the addition of Jay Cutler who has supported multiple receivers in producing top 12 years throughout his career.

TE Elite Tier (Consistency and Upside)

Tight Ends that fall into the Elite tier possess a coefficient of variation under 60 while being in the top 12 of their position greater than 50% of the time, the top 6 33.3% and posted at least one top 3 performance.

While Kyle Rudolph has the lowest coefficient of variation amongst the Elite tier, Rob Gronkowski is the tight end by which all others are judged. Only David Johnson, in his historic 2016 season, finished in the top 6 of his position at a greater rate than Gronkowski. When Gronkowski is on the field he more often than not is going to produce amongst the elite of the position and that is why he is consistently drafted as the first tight end in fantasy drafts year after year. What is interesting though is that Gronkowski isn’t nearly as dominant in producing elite weeks at the position as one might assume.  Both Jordan Reed and Tyler Eifert have posted an equal percentage of top 3 performances when they have been healthy and on the field. Health is obviously the biggest factor with consistency amongst these elite tight ends, but as we’ll see throughout the position the lack of stability from a health perspective leaves an opening for many different players to post elite weeks.

TE Boom Tier (Inconsistent with Upside)

Tight Ends that fall into the Boom tier possess a coefficient of variation above 60 and produced a top 12 rate below 55% and top 6 rate below 33.3%.

 


As we’ll see throughout the tight end position, only three of the 25 tight ends included in this study failed to produce a top 3 weekly performance. That is by far the highest percentage of the four positions. In a very limited role with the New England Patriots, last season Martellus Bennett was able to post a top 3 performance as often as Greg Olsen and a top 6 performance as often as Delanie Walker. Bennett is moving from great quarterback situation to potentially greater quarterback situation in making his transition from New England to Green Bay. Aaron Rodgers has never produced an elite fantasy tight end during his career, but Bennett should show a similar ability to find the end zone in Green Bay and produce these highly variant boom weeks. Charles Clay came on extremely hot during the fantasy playoffs last season, performing in the top 6 in three straight weeks and finishing first and second at the position in the final two weeks. Buffalo only had one major addition to it’s receiving core, being rookie Zay Jones, and Robert Woods departed for Los Angeles which leaves Clay in a prime position to pick up where last season left off and be a surprise performer at the tight end position.

TE Steady Tier (Consistent but Lacking Upside)

Tight Ends that fall into the Steady tier possess a coefficient of variation below 60 while failing to perform in the top 6 greater than 30% of the time.

The most consistent tight end over his last 15 games is Eric Ebron and it’s not particularly close.  If Ebron can continue to progress at the position in his age 24 season and start scoring touchdowns, something he’s failed to do consistently, he could easily move into the Elite tier of tight ends.  In his only game with a touchdown catch, Ebron finished fourth at the position. Ebron also went only one game all last season with less than four catches, so he has definitely gained at least some trust from Matthew Stafford.  Similar to Martellus Bennett, Cameron Brate showed potential to post elite weeks due to his touchdown ability.  At the tight end position is doesn’t take much beyond a touchdown catch to have a big week at the position which can often create a lot of weekly variances and lends itself to being a position that can be played well through streaming and targeting tight ends who find themselves in prime match ups for touchdown potential.

TE Unproven/Unreliable Tier (inconsistent and Lacking Upside)

Tight Ends that fall into this tier possess a coefficient of variation above 65 while failing to perform in the top 12 greater than 40% and the top 6 greater than 25% of the time.

As addressed earlier, the tight end position had the largest percentage of players having elite weeks. The position has a lot of inherent weekly variance due to players missing more time at tight end due to injury due to injury than any other position and tight end being the position that regularly scores the least so it is most impacted by the six point swings of touchdowns. In this tier, the player that has gained the most traction in fantasy drafts is Hunter Henry whose Fantasy Football Calculator average draft position is currently 10th amongst all tight ends. The struggle with this is that outside of a three game stretch in which Antonio Gates missed a game due to injury, Henry posted only two top 12 performances and in those, he was 12th and 11th at the position. With Gates still around for another season, Henry will greatly struggle to produce the elite performances that owners are hoping for drafting him 10th at the position.

 

What Do You Think? Facebook Comments
Thanks for Reading: Please Share

About the author

Adam Cahill

Adam Cahill

2 Comments

Click here to post a comment

  • Great article Adam!

    Would it be possible for you to redo this research with a larger sample size? I understand that by doing so, all rookies from last year will be eliminated but by looking at just last 15 games we are also falling victim to the recency bias.

    When looking at players with “consistency” shouldn’t we be targeting players who have been consistent for longer than just a season?

Fantasy Pros Approved
Follow Us: Twitter: @GridironExperts Facebook Google Plus RSS Feed Soundcloud

Fantasy Blog Network

Get This <a href="https://fantasyknuckleheads.com" title="Fantasy Football Rankings"><b>Fantasy Football for all you knuckleheads</b></a>