Writers of Pro Football Prospectus 2008

22 Aug 2017

2017 KUBIAK vs. ADP: The Underrated

by Vincent Verhei

Last week we looked at this year's overrated players, those who have been going higher in fantasy drafts than their KUBIAK projections say they should. Today we'll flip the script and look at those players we like who have been falling into the later rounds. Once again, we'll be comparing each player's rank in Fantasy Points Over Baseline (FPOB) to their Average Draft Position (ADP). These numbers will show where each player ranks at their position, not overall. This is a change we made in 2015 compared to the way we wrote these pieces in the past, but we think it's a good one -- players at different positions can change in overall value greatly depending on league settings and waiver wire rules, but their rankings at their position shouldn't change much at all. Obviously, which players are underrated or overrated by ADP may change if your league has non-standard rules.

How should you use this information? Suppose we have a third-round grade on a player, but his average draft position has been in the sixth round or so. You could take that guy in the third round and expect good production. Or you could pass on him, knowing he'd probably still be available in rounds four or five, when he'd be an even better value pick. As always, fantasy drafting is an art, not a science.

We should note that KUBIAK is generally better at picking out players who are overrated than it is at picking out players who are underrated, and the difference between KUBIAK and ADP is often bigger for those extremely overrated players than it is for the most underrated players.

(Ed. Note: It's important to mention that this year, ESPN switched its default leagues to PPR, so their ADP tends to reflect values based on PPR. Because of this, we also switched the default scoring when you first download KUBIAK to PPR scoring, and the KUBIAK ranks listed below reflect that.)

If you're interested in history, you can check out what we said in 2011, 2012 (overrated and underrated), 2013 (overrated and underrated), 2014 (overrated and underrated), 2015 (overrated and underrated), and 2016 (overrated and underrated.

* * * * * * *

This year, we have partnered with FanDuel to bring our readers a very special offer to reduce the cost of the KUBIAK projections. If you have never played a daily fantasy contest at FanDuel.com before, just deposit $10 into a brand new account and you will get $20 off your KUBIAK purchase. You can take $20 off of package deals that feature KUBIAK along with FOA 2017, or you can get KUBIAK on its own absolutely free.

Click here to go to FanDuel so you can make your deposit! Make sure after making your deposit to wait 8-10 minutes before coming back here to use your e-mail address as your code to get KUBIAK for free. Please note that this offer is only available to those users who are making a first-time deposit at FanDuel.com.

* * * * * * *

Ordinarily at this point, we break things down one position at a time. This year, however, there is a common theme among fantasy football's most underrated players: many of them come from the same team.

Detroit Lions

QB: Matthew Stafford
FPOB Rank: 11
ADP Rank: 14

WR: Golden Tate
FPOB Rank: 15
ADP Rank: 25

TE: Eric Ebron
FPOB Rank: 12
ADP Rank: 15

K: Matt Prater
FPOB Rank: 4
ADP Rank: 13

This is all quite strange. Our projections are actually a little pessimistic about Matthew Stafford, based on his prior history. He was seventh among quarterbacks in fantasy scoring last year, and has ranked 11th or better five times in the past six seasons. Our projected statline for Golden Tate (92-1,001-5.5) is virtually identical to what he posted last season (91-1,077-4), when he was 17th in ESPN's PPR scoring. Eric Ebron was 14th among tight ends in PPR scoring last season despite missing three games and only scoring one touchdown -- he's healthy now, and a virtual lock to score more touchdowns in 2017, if only because he can hardly score fewer. Matt Prater has ranked seventh or higher among fantasy kickers in three of his last four healthy seasons.

And that's just the players on Detroit's roster who we think should be fantasy starters. Running back Theo Riddick (FPOB Rank: 36; ADP Rank: 43) and wide receiver Marvin Jones (FPOB Rank: 39; ADP Rank: 56) may be nothing more than bench players in deeper leagues, but we think they're being underrated too.

We're not quite certain why the general public are down on the Lions as a group, but we think it may have something to do with how few drives Detroit got per game last season -- and how likely they are to get more drives in 2017. The Lions used a heavy dosage of small-ball on both offense and defense last season, favoring long marches down the field no matter which team had the ball. The offense led the league with a devilish 6.66 plays per drive, while the defense allowed opponents to run 6.34 plays per drive, third-most in the league. As a result, the Detroit offense only got the ball for 152 drives last season. That's 13 fewer than any other team. In fact, only two teams in the past 20 years have had the ball less frequently: Peyton Manning's Colts in 2008 (144 drives) and 2006 (148, but still second in the NFL in points scored that season, because Peyton Manning was ridiculous).

It's very likely that the Lions will get more chances with the ball next year, and maybe a lot more chances. Only eight teams from 1996 to 2015 collected fewer than 160 drives. Six of those teams had more drives the next year, by 6.4 drives per team on average. If we up are threshold to teams with fewer than 176 drives (or 11 per game) than we find 146 teams, 106 of which (73 percent) had more drives the following season, by an average of 6.7 drives. More drives means more opportunities to pick up touchdowns (and field goals, for that matter), which should mean better numbers for the Lions across the board.

And now, back to our usual look at underrated players by position.

Quarterbacks

Andy Dalton, CIN
FPOB Rank: 13
ADP Rank: 18

Ordinarily we limit the players in this piece to starters in typical leagues, which for quarterbacks usually means the top 12. We're making an exception here, however, because Andy Dalton is only one slot out, and because he is by far the most underrated passer of any prominence this season. The funny thing is that fantasy drafters are reasonably high on the Bengals as a team. They have A.J. Green as the fifth-ranked wide receiver and Tyler Eifert seventh among tight ends, and they think Joe Mixon will be a midrange RB2 (though, as we covered last week, we think they are wrong on that last one). Well, if the Bengals are going to have productive players at running back, wideout, and tight end, doesn't that mean they will have a productive quarterback too? Who do you think is going to be throwing all those touchdowns to Green and Eifert? It's important to note that we are not saying you should grab Dalton as your starter. We are saying, however, that you can probably wait a while to grab your QB2, knowing that Dalton will likely still be there late in the draft, after four or five inferior quarterbacks have been taken.

Ben Roethlisberger, PIT
FPOB Rank: 9
ADP Rank: 12

Ben Roethlisberger's situation is similar to Dalton's. The Steelers have fantasy football's highest-rated wide receiver in Antonio Brown, and its second highest running back in Le'Veon Bell -- who is ranked so high due in large part to his receiving value. The Steelers are going to score plenty of points this year, which means Roethlisberger should score plenty of points for his fantasy owners. Surprisingly, for a quarterback with a long track record of success and some serious Hall of Fame credentials, Roethlisberger has never been a fantasy juggernaut. Only three times in his career has he been a top-ten quarterback in fantasy production. With Brown, Bell, and Martavis Bryant ready to go, however, he should produce one of his best statistical seasons in 2017.

Running Backs

Bilal Powell, NYJ
FPOB Rank: 18
ADP Rank: 27

When we saw that Bilal Powell was the most underrated running back in fantasy football, we assumed that the fans would be overrating Matt Forte, but that's not the case. In fact, we think Forte (FPOB Rank: 37; ADP Rank: 41) is underrated too, though we still don't think he should be on your draft board. What's really going on here is that the Jets offense, though quite terrible, is not going to be as terrible as you may have been led to believe. Don't misunderstand, they will probably finish last in points, yards, and touchdowns this year, but they aren't going to get shut out 16 times. They will score some points, they will gain some yards, they will manage some touchdowns. The player most likely to produce those points, yards, and touchdowns is Powell, the only player on the roster who should be on your fantasy radar. In the last four games of 2016, Powell had 552 yards from scrimmage and three touchdowns. He won't match that kind of production over 16 games, but he'll do enough that he should be a starter in most leagues.

Mark Ingram, NO
FPOB Rank: 14
ADP Rank: 23

Now here is a case where we think fantasy players are backing the wrong horse. Adrian Peterson is 24th among running backs in ADP, but just 46th in FPOB. (With those numbers, Peterson would have been far and away the top running back in last week's piece on overrated players, but at the time even his ADP wasn't high enough to consider him a viable starter. That changed this week.) Peterson arrives in New Orleans as a 32-year-old running back who missed 13 games last year, and when he did play, averaged 24.0 yards per game and 1.9 yards per carry (not a typo!). This is not a recipe for success, even when talking about arguably the NFL's best running back this century. Meanwhile, only seven running backs have more yards from scrimmage than Mark Ingram in the past three years, despite the fact that Ingram has had his own injury struggles, missing three games in 2014 and four more in 2015.

Tevin Coleman, ATL
FPOB Rank: 22
ADP Rank: 30

You can take Kyle Shanahan out of Atlanta, but you can't stop Atlanta from throwing passes to its running backs. We're expecting modest upgrades in Tevin Coleman's rushing totals, and more catches too. His receiving yardage will go down, because he's not getting a trio of 40-yard receptions again like he did last year (two of them in one game against Denver). But even if Atlanta's offense takes a big step backwards, it should still be among the NFL's best, and we see Coleman taking just a little bit of playing time away from Devonta Freeman.

Wide Receivers

Larry Fitzgerald, ARI
FPOB Rank: 14
ADP Rank: 26

Emmanuel Sanders, DEN
FPOB Rank: 20
ADP Rank: 27

When we saw Larry Fitzgerald and Emmanuel Sanders, along with Golden Tate, as the most underrated wide receivers in fantasy football, our first thought was that PPR scoring was skewing our results and making all slot receivers look underrated. However, that is not the case. Our average draft position rankings are based on drafts from three websites, and only one (ESPN.com) uses PPR scoring. With ESPN.com data, Tate, Fitzgerald, and Sanders, in that order, are clustered 25-26-27 in ADP. Remove those ESPN numbers, however, and Fitzgerald and Tate are tied at 28, with Sanders one spot back at 29. That's hardly any difference at all, and does nothing to explain why the general public disagrees with us when we say that each of the three should be a fantasy starter. Even stranger, we're not expecting banner years from any of the three, we're just saying they'll do this year more or less what they did last year. We mentioned that Tate is a victim of Detroit's entire team being underrated. Perhaps fantasy players are equally skeptical of the situations in Arizona (where we're expecting a mild bounceback from Carson Palmer) and Denver (where Trevor Siemian has been named the starting quarterback, which might cause a change in Sanders' ADP a week from now).

Tight Ends

Kyle Rudolph, MIN
FPOB Rank: 6
ADP Rank: 9

We're expecting a slight decline for Kyle Rudolph, who was third among tight ends in fantasy scoring last season. That's largely because he had 840 yards in 2016, but never even hit 500 in his first five NFL seasons, and he figures to regress towards his baseline production level. The general public, meanwhile, is predicting a big fall. We're expecting a modest uptick in Sam Bradford's numbers next year, and while Stefon Diggs and Adam Thielen will probably be the primary beneficiaries, Rudolph will still get his catches too. He's not just a fantasy starter, he'll outscore his opposition at the position more often than not, and should be a plus-player on the boards in late draft rounds.

Greg Olsen, CAR
FPOB Rank: 2
ADP Rank: 3

We have Greg Olsen and Travis Kelce second and third among tight ends behind Rob Gronkowski, while the consensus draft order has been Gronkowski-Kelce-Olsen. That's a subtle difference, but an important one when you're talking about the first players at one position to go off the board. It comes down Olsen's longer, more sustained track record of excellent production. He is coming off his third straight thousand-yard season, and has eight total seasons with at least 500 yards. Kelce has only one thousand-yard season, and only three with 500-plus yards. In fact, though he won't get close to Gronkowski's per-game numbers, Olsen has been much more reliable. Olsen hasn't missed a start in five years. Gronkowski has missed 24 games in that same timeframe, at least one game every season. In that light, Olsen is the safer pick, and you could make a reasonable argument that he should go ahead of Gronkowski as the first tight end off the board.

Kicker

Graham Gano, CAR
FPOB Rank: 7
ADP Rank: 18

Ryan Succop, TEN
FPOB Rank: 10
ADP Rank: 19

Yes, we are listing two kickers here, in addition to Prater in the Detroit section. There are just a lot of kickers who are being underrated this year. It's a little funny that we're listing the kickers in Carolina and Tennessee as underrated after saying the quarterbacks on those same two teams were overrated last week. In both of those cases, though, it's not so much that we are skeptical about the offenses' ability to score points, just that we expect the running backs to get most of the production. We actually have the Panthers and Titans as favorites to win their respective divisions, which would mean plenty of points for Graham Gano and Ryan Succop.

However, there is a caveat here, which is that Gano is in a camp battle with seventh-round pick Harrison Butker. The underrated player here is really "Carolina kicker," whoever that is. If Butker wins the job, he becomes the underrated kicker.

Defense

Philadelphia Eagles
FPOB Rank: 4
ADP Rank: 16

Fantasy defenses are very volatile and can change wildly from one year to the next, but this is still bizarre. The Eagles defense was sixth in fantasy scoring last year, so why would we expect them to suddenly fall so far and no longer be worthy of starting? They had three defensive touchdowns last year, an average amount and not a suggestion of future regression. The schedule offers plenty of matchups that could lead to big numbers, including games against the 49ers, Broncos, Bears, and Rams. One of their toughest matchups is Dallas, but their second game against the Cowboys is in Week 17, which means it probably doesn't matter if your league has playoffs that end in Week 16. However, do be aware that the Eagles have a Christmas night game against Oakland in Week 16, so you may have to stream a new defense for your fantasy championship.

Posted by: Vincent Verhei on 22 Aug 2017

19 comments, Last at 22 Sep 2017, 6:37am by psncodes1

Comments

1
by ALauff :: Tue, 08/22/2017 - 5:17pm

Regarding the Detroit players, I think one thing to consider is weekly consistency. I began tracking such things as median weekly positional rank, hierarchical-pt. games (i.e., top-3, top-6, top-12, etc.), start percentage, and so on.

For instance, Golden Tate finished 17th cumulatively, but he only posted 4 games within the position's top-24 scorers in a given week, and had only 6 games within the position's top-36 all season. He had the same number of top-24 performances the season before that. Such volatility is odd for a receiver who works so close to the line of scrimmage. Compare with Edelman or Landry, similar types of players, and you'll see what I mean.

Something similar for Stafford. End-of-year numbers look good, but it's a weekly roller-coaster: less than half his performances ranked in the weekly top-12 for QBs over the last 2 years, while he had 7 games outside the top-18 last season. It gets kind of hard to differentiate him from the mass of guys from, oh, QB12-20, encompassing Palmer, Dalton, Rivers, etc.

I'm with you on pretty much everyone else, though, even Ebron. It's really just Stafford and Tate who stick out for me as being
counterintuitively over-underrated (or something).

2
by Aaron Schatz :: Tue, 08/22/2017 - 6:25pm

Admittedly, predicting weekly volatility is something we've never worked on and probably would be a good addition to our projections. I wonder how much that is predictable from season to season.

7
by nkowal :: Tue, 08/22/2017 - 10:42pm

Not sure on the r squared here, but Tristan Cockroft does consistency rankings at ESPN:

http://www.espn.com/fantasy/football/story/_/page/consistency201416/fant...

11
by ALauff :: Wed, 08/23/2017 - 10:57am

It's hard to say, and I'm not entirely confident considering I've only tabulated 2015-2016 for QB, RB, WR, and TE. Thirty games (excl. fantasy-irrelevant week 17s) isn't a huge sample size. Some players, like DeSean Jackson, seem to validate conventional wisdom that they're highly boom-bust, while others, the lower-ADoT/ypr guys like Edelman, have a steadier production baseline. But then Tate comes along with a similar profile and pushes against against establishing a definitive archetype along those lines.

I try to treat it as one more tool in assessing player value. There is probably a way to determine the predictive qualities of the data, but this is only a hobby for me, my time is short, and I have no background in statistical analysis. Fun, though!

12
by MilkmanDanimal :: Wed, 08/23/2017 - 11:11am

It does seem to dovetail nicely into some of FO's standard wisdom, in the sense that DVOA prioritizes consistent production over typical boom-and-bust guys (hence the standard "except for Barry Sanders" codicil in every version of FOA). I don't play fantasy football anymore, but do think this sort of thing would really be highly valuable and having some kind of expected standard deviation for week-by-week performance would add a lot of value. As someone who (A) doesn't have any background in statistical analysis but (B) loves reading statistical analysis, it's an interesting idea.

3
by bingo762 :: Tue, 08/22/2017 - 6:48pm

I'd like to subscribe to your newsletter. Seriously. If you have a fantasy blog, I'd like to check it out

13
by ALauff :: Wed, 08/23/2017 - 11:39am

Thanks--you're too kind. This is a hobby for me at the moment. I've been playing fantasy football for about 10 years and find it more fun to do my own legwork.

Two writers I recommend who have helped shape my thinking are JJ Zachariason of numberfire and Rich Hribar of Rotoworld. I highly recommend reading their work.

14
by Raiderjoe :: Wed, 08/23/2017 - 1:01pm

I do a 2-qb league. wel, one QB slot but thenalso it has a Qb/RB./WR.TE slot which moist people ise for a 2nd qb. tremendous league. have run many leagues in past but find this one most enjoyable. (second place was a scoring only league I did. some cool scores like 34-1 and 13-13 happened).

10-team league so that there are at least a few real life starting QBs available on waivers. Stafford still valuable in that league even with inconsistency and stuff.

15
by Raiderjoe :: Wed, 08/23/2017 - 1:03pm

yeah but so liek anyway in 12-team, one-QB league, m. Stafford possibly want to avoid as your main guy.

17
by ALauff :: Wed, 08/23/2017 - 2:30pm

Yeah, that sounds about right. Not that he won't have good games, but you might want to be selective about when you use him; probably not a guy you leave in all season.

9
by The Ninjalectual :: Wed, 08/23/2017 - 12:20am

This is a really interesting comment. Based on your research, I wouldn't want Stafford on my team either! Is that bias irrational? Is greater week-to-week consistency more or less important than year end totals with regard to winning your weekly matchup?

10
by ALauff :: Wed, 08/23/2017 - 10:52am

I believe it's a balancing act between baseline consistency (or "floor") sprinkled in with spiked weeks ("ceiling"). One of the valuable things about tracking this data (which you can find conveniently tabled, going all the way back a decade or so, at fantasy data: https://fantasydata.com/nfl-stats/nfl-fantasy-football-stats.aspx) is you can see which QBs are high-floor and which are boom-bust, and tier accordingly without falling into the trap of looking at year-end stats context-free. It also helps you see outliers--that, and tabulating rate stats (ypa) vs. career averages to see who will regress and rebound.

A quick snapshot of two players to illustrate my meaning re: floors and ceilings using this data:

Tyrod Taylor - med. rank 12 / top-3: 2 gms /top-6: 4 gms / top-12: 8 gms (53% start) / outside top-18: 2 gms

Russell Wilson - med. rank 14 / top-3: 2 gms / top-6: 6 gms / top-12: 7 gms (47% start) / outside top-18: 7 gms

*What this tells me is that, despite finishing back to back in scoring, Taylor was a much more consistent starter than Wilson *last year*, because he had a lesser chance of running into a week-losing game AND a more-stable point floor. With that said, 2016 was an anomaly for Wilson, as he was a strong starter in the preceding two years. But this does show how conventional wisdom can be misleading in regard to how fantasy points are valued.

-Andy

16
by ChrisS :: Wed, 08/23/2017 - 1:28pm

Here is a Lions fan comments about Stafford. In 2015 the team started the season with a new OC who only lasted part of the season as the offense stunk. Once they replaced him with Jim-Bob Cooter at OC, Stafford and the offense improved. So I can see a lot of variation possible in the two parts of the season. In 2016 he played fairly well through most the season then he had a finger injury (week 10 or so) and had trouble gripping/throwing the ball and his play turned to crap. Again resulting in a lot of variance. In general I think Stafford maybe be a bit more volatile than average, but you need more data before feeling comfortable with any definitve statements. In general I think fanatasy people under-rate Lions players is because as a team they suck and have for quite awhile.

18
by ALauff :: Wed, 08/23/2017 - 2:45pm

That's fair. I would need to cut 2016 into "before" and "after" injury to see the extent of the dropoff. For what it's worth, the pattern extends into 2014 as well, with a similar rate of good/acceptable/subpar games. The comment was purely numbers-based, not an assessment of his quality as a player.

The larger point: it is often wise to wait on your fantasy QB, unless you can get a real value on the stars of the position (Rodgers, Brady), or that next tier (imo, Brees and Wilson this year). Otherwise, once you get to QB10+, there just isn't enough difference to merit reaching for a guy who doesn't supply a competitive edge. You are often better off waiting till late, getting a guy with a great start to his schedule, and then mixing and matching based on matchups. That the NFL is a passing league simply makes the position more of a commodity, raising the level of production across the board.

4
by jwkelly :: Tue, 08/22/2017 - 7:34pm

The WR "Antonio Bryant" referenced in Roethlisberger's section actually did play for Pittsburgh, but it was the Panthers, not the Steelers.

6
by Vincent Verhei :: Tue, 08/22/2017 - 8:06pm

Thank you. Fixed.

5
by jwkelly :: Tue, 08/22/2017 - 7:39pm

.

8
by The Ninjalectual :: Wed, 08/23/2017 - 12:13am

In his four seasons, he has missed one, two, eight, and 15 games

So he's on pace to miss 29 games this season alone! Wow! If he keeps this pace up, we could be witnessing him missing 219 games a year by 2020

19
by psncodes1 :: Fri, 09/22/2017 - 6:37am

Yes this is true that you can play words with friends online with your friends and this http://freewordswithfriendscheat.com/ makes you cheat in this game without any efforts you can use words finder.