<
>

Introducing OFP: Opportunity-adjusted fantasy points

play
Was Will Fuller lucky last year? (2:00)

Matthew Berry, Field Yates, and Stephania Bell evaluate how Will Fuller took advantage of big-play opportunities last season and he'll be able to repeat any of the magic this year. (2:00)

Volume is king in fantasy football.

Those are six words I've used often throughout my time in the industry. Football analysts (yours truly included) spend a lot of time studying player efficiency and trying to determine the most accurate ways to project the short- and long-term effectiveness of each and every player. Though that's an important piece of the puzzle -- and one that, if done correctly, can supply an edge -- nothing correlates better with fantasy production than opportunity.

And that's why today, I'm introducing opportunity-adjusted fantasy points, or OFP for short. OFP strips away player talent and efficiency and focuses solely on opportunity. It's a natural extension of OTD (opportunity-adjusted touchdowns), which I introduced back in 2013 and update weekly here on the site.

Attached at the hip with OFP will be FORP -- fantasy points over replacement player -- the fantasy football equivalent of VORP, which was made popular in MLB analysis many years ago. In this case, FORP will be the difference between a player's OFP and actual fantasy point total.

You may be wondering, "If volume is what's important, is this even necessary? Shouldn't I just look at pass attempts, rushes and targets?"

Though those three categories will help fuel this exercise, OFP goes even deeper. Volume isn't as simple as just having more touches, as all opportunities are not created equal. That's why categories like line of scrimmage, average depth of target and in-box defenders are factored into OFP. This process takes each of the categories that generate a player's fantasy point total in ESPN PPR leagues, adjusts them based on opportunity and spits out both the player's adjusted stats and fantasy point total.

Those adjustments are made as follows:

Pass attempts: No change

Pass completions: Pass attempts multiplied by a passer's expected completion percentage based on depth of throw and location

Passing yardage: Pass attempts multiplied by a passer's expected yardage based on depth of throw and location

Passing touchdowns: Passing OTD

Interceptions: Pass attempts multiplied by a passer's expected interceptions based on depth of throw and location

Rushing attempts: No change

Rushing yardage: Rush attempts multiplied by a rusher's expected yardage based on in-box defenders.

Rushing touchdowns: Rushing OTD

Targets: No change

Receptions: Targets multiplied by a receiver's expected catch rate based on depth of target and location

Receiving yardage: Targets multiplied by a receiver's expected yardage based on depth of target and location

Receiving touchdowns: Receiving OTD

Two-point conversions: OTD multiplied by league-average two-point rate

Fumbles lost: Touches multiplied by league-average rate

Return touchdowns: Excluded due to extreme rareness and randomness

To get an idea of what this looks like, click on the links below for the OFP leaderboards from the 2017 season:

If you scan through the leaderboards, several things should pop out, and the purpose of this introductory piece is to address them. Note that the OFP leaderboards will be updated weekly during the 2018 regular season.

Wait a second ...

Something I battled with during the creation of OFP -- a project I've been visualizing and thinking about since before I created OTD -- was exactly how to adjust for specific categories. I had to continually remind myself that this is all about opportunity, not skill. And that's important for you to remember as well, as it's easy to get spun off course.

An example of this is a player's interception total. Alex Smith, for example, is very protective of the football, yet his 2017 adjusted interception total works out to 12, which both matches his career high and is well above the five he threw in reality. It may seem silly to calculate an adjusted total using 12, but the fact is, an average quarterback who matched Smith's 505 passes and delivered them to the same location on the field would've thrown 12 interceptions.

Again, it's opportunity, not skill. Smith has been around long enough that we will input a better interception rate for our player projections, but the adjusted total can give us better information if there's a quarterback or coaching change and for players with a shorter NFL résumé. As we'll get to later, there are always exceptions, but players who post an extremely high or low FORP (again, the gap between their OFP and actual fantasy point total) tend to regress to the mean (i.e., crash back to earth) in a hurry.

Why is this useful?

Though I am posting only the 2017 OFP results here, I did the obvious and went back a decade to determine if this was a relevant stat. And by "relevant" I mean that it's not just here for show and actually has some value in projecting the future.

Though we're still at the beginning stages of studying OFP, it proved to be extremely valuable in terms of trying to nail down players likely to regress to the mean (not unlike OTD). "Regression" is a term thrown around often in the industry, but it's important to understand that, when applied correctly, regression to the mean isn't necessarily a negative. It can be either good or bad for a player's future production.

Consider that, from 2008 to '16, 93 players posted a FORP of at least 50. Of those 93, 66 appeared in at least 12 games the following season, and of those, 50 (or 75.8 percent) scored fewer fantasy points. Only eight (or 12.1 percent) of the 66 managed a larger FORP the following season. The average dip in actual fantasy points for the 66 players was 36.5 points.

In 2017, seven players fit into this category: Alvin Kamara (101 FORP), Tyreek Hill (69), Alex Smith (65), JuJu Smith-Schuster (64), Todd Gurley (63), Kareem Hunt (53) and Dion Lewis (52). If recent history holds, three or four of the seven players will score fewer fantasy points in 2018 ... likely by a significant margin.

On the flip side, 78 players posted a FORP of minus-40 or worse during the same span. Of those 78, 41 appeared in at least 12 games the following season, and of those, 25 (or 61.2 percent) scored more fantasy points. Only two (or 4.9 percent) of the 41 players posted a worse FORP the next season (it was ESPN's Matt Hasselbeck both times, oddly enough). The average increase in actual fantasy points for the 41 players was 29.2.

In 2017, five players fell into this category: DeShone Kizer (minus-79), Carlos Hyde (minus-50), Zay Jones (minus-50), Tom Savage (minus-47) and Mike Evans (minus-41). If recent history holds, two or three of these players will appear in fewer than 12 games and one or two will score more fantasy points in 2018.

"One or two" may not seem like a particularly optimistic number, but -- and this makes sense if you think about it -- most players actually score fewer points the following season. This is primarily a product of a high injury rate and the league's absurd turnover rate. Nonetheless, one-third of this sample of players is actually pretty solid, especially because you can increase your odds of hitting by process of elimination. For example, Kizer and Savage are strong bets to see a drastic decrease in playing time in 2018, and Hyde is very likely to score fewer fantasy points in Cleveland. That eliminates three names and suggests Jones and Evans are good bets to score more points in 2018, thus making them post-hype or buy-low targets.

That's a lot to process, but the point is that NFL players are simply unable to use great (or poor) efficiency to consistently overcome (or fall short of) their volume-fueled expected production. This is the same conclusion reached by OTD and other similar studies and it clearly applies to other statistical categories, including fantasy points.

What about the cream of the crop?

Though it's become clear even the league's best players have trouble consistently outperforming expected production by massive margins, common sense tells us that they should be, at least, beating out expectations to some extent.

I mean, if they aren't, everyone is just replacement level, right?

Of course that's right, though something important to remember is that you earn volume by being a good player. That sounds obvious, but it's notable here because sometimes inferior players post elite efficiency numbers over small samples. This is often a product of variance or circumstance and doesn't necessarily mean those players are better than the players getting more volume. Regardless of whether or not that player is promoted, his stats will almost always regress to the mean.

To help prove the value of volume, I took the 20 running backs who have scored the most total fantasy points over the past five seasons. If we remove seasons in which the backs appeared in fewer than five games, we're left with a sample of 90 seasons from some of the best backs of the past half-decade. Of the 90, 44 fell within 10 fantasy points of their OFP (or expected total), 62 were within 20 points and 72 were within 30 points. The backs averaged a FORP of 7.1.

The same study done on the other three positions shows similar results, with wide receivers (13.9 VORP) exceeding expectations by the largest margin, tight ends by the smallest (3.6) and quarterbacks (9.0) in the middle. The studies also did well to locate the league's clear superstars (players who consistently posted high FORP marks). These include Aaron Rodgers, Russell Wilson, Drew Brees, Antonio Brown, Odell Beckham Jr., Doug Baldwin, Rob Gronkowski and Travis Kelce.

So, what have we learned? The league's best players can consistently post a high FORP, but it's a relatively small group, and even those players tend to work their way back into the vicinity of their expected total. Once again, if you're familiar with OTD, this should sound familiar.

Notable names and application

As I worked my way through the research, some names of interest popped up along the way.

• The first is Smith, who sports the fourth-largest FORP (1,388 actual fantasy points, 1,182 OFP) among quarterbacks over the past six seasons. Only Rodgers (285), Wilson (268) and Brees (239) top Smith's mark of 206. Smith played an infamously conservative style while rejuvenating his game in Kansas City and consistently posted a high FORP by protecting the football, completing a high percentage of his passes and adding value with his legs. Smith's best season was 2017 (65), but he was much better than expected in 2014 (34) and 2015 (54), as well. I think it's fair to say Smith is an underrated player. Now in Washington, Smith already had the looks of a good value in 2018 fantasy leagues, and this only helps.

• Is Antonio Brown returning to earth? After posting a FORP of at least 60 in both 2014 and 2015, he fell to 38 in 2016 and 39 in 2017. Those are still terrific marks, but it's a trend to monitor as he's now 30 years old.

• DeShone Kizer's rookie season was a disaster and that shows up in a big way in OFP. Kizer ranked 25th at the position with 176 fantasy points, but finished -- cue the drumroll -- seventh with a 255 OFP. Seriously. Kizer's propensity for the long ball and hefty 77 rushing attempts positioned him for massive fantasy output, but he struggled badly with accuracy and turnovers. This is notable for Tyrod Taylor, who replaces Kizer as Cleveland's starter. Taylor, by the way, finished eighth in OFP last season, but posted a minus-23 FORP after easily managing positive outputs during each of his first two seasons in Buffalo.

Cam Newton has posted a negative FORP during three of the past four seasons, with the only exception his historic 2015 campaign. Incredibly, he managed "only" a 36 FORP during that 2015 season (389 fantasy points). Turnovers and inefficient passing are the culprits here.

• Carlos Hyde was an enigma in 2017, posting his best fantasy season, but also his least efficient body of work. This shows up here, as he posted the position's fourth-highest OFP (284), but a minus-50 FORP led to an eighth-place finish in actual fantasy points. This, of course, bodes well for Jerick McKinnon, who replaces Hyde in Kyle Shanahan's running back-friendly offense.

• Believe it or not, DeAndre Hopkins sports a minus-1 FORP during his five-year career. That includes a 2017 campaign in which Hopkins scored 310 fantasy points, but his 21 FORP barely exceeded his position-high 289 OFP. On the plus side, Hopkins was overloaded with volume, something unlikely to change in 2018. On the other hand, he isn't showing an ability to consistently show well in FORP like other star players. That should register as a red -- or at least yellow -- flag for a player being drafted in the first round.

• Over the past eight seasons, Matt Ryan's OFP has risen no higher than 277 (2012) and no lower than 238 (2017), but his fantasy point totals have stretched as high as 347 in 2016 and as low as 228 the season after that. Despite that precipitous drop in fantasy points from '16 to '17, his OFP fell only from 254 to 238. He's a prime candidate for a step forward in 2018.

• Mike Evans has been about as inefficient as possible over the past three seasons, managing a minus-102 FORP (820 OFP). That includes two seasons with a minus-41 or worse FORP (2015, 2017). Two seasons with fewer than six touchdowns and a 53 percent catch rate will cause that.

• Rob Gronkowski has posted a FORP of at least 24 each of the past four seasons, but hasn't surpassed 34 during the span. Gronk had posted a FORP of at least 31 each of his first three NFL seasons. Similarly, Travis Kelce posted a FORP of at least 20 each of the past four seasons, but hasn't topped 41 during the span.

• Sammy Watkins' FORP is 84 over the past three seasons. He's pretty good.

• Tyreek Hill has managed a FORP of 61 and 69, respectively, during each of his first two NFL seasons. Hill ranked 22nd at the position in OFP, but eighth in actual fantasy points last season. With our earlier research considered, he's about as close to a lock for a return to earth as anyone in recent years.

• Alvin Kamara's historic rookie season saw him post a ridiculous 101 FORP. That's the largest single-season FORP for any player since 2011. Kamara's unsustainable efficiency allowed him the third-most fantasy points at the position, but he was ninth in OFP, just behind teammate Mark Ingram.

• JuJu Smith-Schuster's 198 fantasy points were 20th-most during his impressive rookie season, but the second-round pick finished 47th in OFP. Deep-ball efficiency, touchdown fortune and, oh yeah, that 97-yard touchdown will do that. Smith-Schuster will need more volume to sustain WR2 production in 2018.

• During his first five seasons in the league, Dez Bryant posted a 212 FORP. He sits at minus-77 in the category over the past three seasons.

Jameis Winston has posted a minus-16 or worse FORP each of his three NFL seasons. Winston played better in 2017, but has had struggles with accuracy, turnovers and rushing efficiency.

Julio Jones posted a FORP of at least 17 during each of his first six NFL seasons before posting a minus-8 mark in 2018. Unlucky touchdown production was the obvious culprit. He's a strong bounce-back candidate.

Larry Fitzgerald has posted a negative FORP five of the past six seasons, including minus-4 in 2017.

Final thoughts

Because yards per reception (YPR) and yards per attempt (YPA) weren't enough, I generated average depth of target (aDOT). Because red zone data is badly flawed, OTD was invented. Now, because simply looking at touches or raw fantasy point totals is not enough, we have OFP and FORP. No single statistic can give you all the answers when it comes to football prognostication, so the best way to maximize your chances of winning your league title is to gather the best possible information, while also ignoring noise that won't help your decision-making.

Volume isn't the only piece of the puzzle, but it is certainly the largest. As the research laid out in this piece shows, OFP can help you make the best decisions on draft day and when evaluating lineup, waiver and trade decisions during the season.

Be sure to check back each week during the 2018 regular season for an updated leaderboard.