Free Players Quick Links
Now, Why We're Learning from History
I love NFL history. I spend embarrassing amounts of time just navigating aimlessly around www.pro-football-reference.com. I'm the Tom Brady of trivia night at Footballguys' annual staff retreat. I rather frequently joke on Twitter that I'm just a Del Shofner trivia account that also happens to do some fantasy analysis. Some like-minded friends and I got together to start our own Hall of Fame just because we believed the NFL was doing such a bad job of it. I love NFL history.
Far from being just a random quirk, I believe this love of history is a big part of what makes me a good fantasy analyst. And there's solid evidence backing me up on that.
In 2011, the Intelligence Advanced Research Projects Activity (IARPA)— the research arm of the US intelligence community— staged a contest to see who was best at predicting uncertain, one-off world events. Would a specific country fall into a civil war within three years? Would a certain resource top a certain price within the next six months?
This contest was dominated by a group called the Good Judgement Project led by scientist Philip E. Tetlock. Tetlock started with a group of a thousand participants and averaged their predictions in a "wisdom of the crowds" approach, but over time he graded the participants, separated out the most accurate predictors, and put them together in a special group of so-called superforecasters.
After the contest, Tetlock paired with journalist Dan Gardner to write up what he learned in a book titled, appropriately, Superforecasting. The eponymous superforecasters had many differences, of course, but they shared a few key traits that helped them achieve such remarkable accuracy at predicting inherently unpredictable events. One of those key traits was that the starting point for every prediction was learning from history.
Superforecasters know that there is nothing new under the sun. Nothing is 100% “unique.” Language purists be damned: uniqueness is a matter of degree. So superforecasters conduct creative searches for comparison classes even for seemingly unique events, such as the outcome of a hunt for a high-profile terrorist (Joseph Kony) or the standoff between a new socialist government in Athens and Greece’s creditors. Superforecasters are in the habit of posing the outside-view question: How often do things of this sort happen in situations of this sort?
I've written in the past about the flaws with historical comparisons. Julio Jones and Sterling Sharpe have been similarly productive through age 29, but this doesn't mean that Julio Jones is suddenly at heightened risk to suffer a career-ending injury just because Sharpe did. Saying Saquon Barkley is similar to LaDainian Tomlinson doesn't mean that Barkley is destined to share Tomlinson's health, and had Tomlinson gotten seriously injured it wouldn't make Barkley's career more likely to go one way or another as a result.
But for all the weaknesses historical comparisons have, they do have one thing going for them: when you do them well, they're demonstrated to improve your predictions. In this series, I'll take a long look at how to do them well, in addition to sharing comparisons for some prominent players and discussing how these comparisons should shape our opinion of the range of possible paths their careers could take.
Historical Comparisons: Details Matter
The three main factors that dictate whether one player is comparable to another are age or experience, draft position, and production to date. Some might quibble with the inclusion of draft position, but as you'll see when we look at individual players, it matters quite a bit even when players have several years in the NFL. The longer a player is in the league, the more weight age and production take on in finding comparisons, to the point where for a 4-time first-team All-Pro like Antonio Brown, the fact that he was drafted in the 6th round becomes a footnote at best.
Production comes in a variety of forms. One could look at total yards or postseason honors (pro bowls, All-Pros, etc). One could use total fantasy points, which is a better measure of the thing we actually care about as fantasy football players.
Personally, I prefer to measure production with a home-brewed metric I call "EVoB", or "Estimated Value over Baseline", which I have calculated for every player back to 1985. Because fantasy players need another acronym almost as much as they need another hole in the head, I'm going to be referring to this metric simply as "value" for the course of the series.
If you really want to know how the sausage is made, here's a link laying it out in detail. But really, all you need to know is that EVoB essentially asks "How many points would you score if you started a player off of waivers at this position every week for a full season? Now, how many more points would you score if you started this player instead?"
(Long-time readers of Footballguys might ask themselves "isn't this just VBD?" The short answer is yes, basically. The slightly longer answer is that this is VBD with a more accurate measurement of true replacement level, and calculated on a per-game basis instead of a per-season basis. But the underlying principle is the same: EVoB or "value" is my best estimate of how many extra points this player would have put into your starting lineup over the course of a season.)
Why do I use this measure of value to estimate production? Because it has a number of really cool advantages. First of all, it automatically adjusts for era. From 1986 to 1991, tight end Steve Jordan averaged 50/676/3.7 receiving for every 16 games, roughly similar to what Vance McDonald or David Njoku produced last year. Jordan was a pro bowler all six seasons and finished 3rd, 8th, 3rd, 6th, 6th, and 6th for fantasy, rather unlike Vance McDonald or David Njoku last year.
Teams simply didn't use their tight ends as much in the late '80s and early '90s. Jordan's fantasy finishes tell a much more compelling story about how good and productive of a player he was than his raw statistics might.
The second big advantage over using something like "positional ranking" is that value shows a much finer level of detail. Finishing as the 10th-best back in 2018 was actually more impressive than finishing as the 2nd-best back in 2015. A #1 overall finish in 2006 was more than twice as valuable as a #1 overall finish in 2008. A bunch of RB18 finishes are no substitute for one or two top-3 finishes. "Value" quantifies that and allows us to make comparisons between players whose careers took on a variety of different shapes.
How Much Value is "Value"?
Since "value" is how many extra points a player was worth to you, it helps to have some general sense of scale.
- 200+ points: this player is a dominant, league-winning force. In any given year there are usually 1-2 running backs that meet this threshold. Last year, in PPR leagues, there were four: Todd Gurley, Saquon Barkley, Christian McCaffrey, and Alvin Kamara.
- 80+ points: this is typically where solid RB1s wind up falling. There have been between 11 and 13 running backs to top 80 points of value in seven of the last eight seasons, (the exception was 2015 when the position was beset by injuries and only six backs met this mark). In 2018 terms, when you see 80-100 points of value, think of someone like Phillip Lindsay (85.35 points).
- 30+ points: this is usually around where you see the cutoff between weekly starters and flex options. In 2018, 29 running backs added at least 30 points of value. In 2017, only 22 backs did so. In 2018 terms, when you see 30-40 points of value, think of someone like Matt Breida (34.3 points) or Jordan Howard (33.4 points).
In terms of total career value, here are the benchmarks to consider.
- 1500+ points: A first-ballot Hall of Famer for both NFL and fantasy purposes, the kind of player where a decade from now you'll still be telling people you owned them in your dynasty league. Since 1985, the only running backs to meet this threshold in PPR scoring were Marshall Faulk, LaDainian Tomlinson, Emmitt Smith, and Barry Sanders.
- 1000+ points: This player will probably have a half-decent argument for the Hall of Fame, though he usually won't be a lock. Hall-of-Famers Curtis Martin and Thurman Thomas fall between 1000 and 1500 points of value, as does presumptive Hall-of-Famer Adrian Peterson, but so do Edgerrin James, Ricky Watters, Tiki Barber, Priest Holmes, Matt Forte, and LeSean McCoy.
- 600+ points: 45 backs in total have topped this mark since 1985. These players usually either had long careers as quality starters (Frank Gore, Warrick Dunn, Clinton Portis, Corey Dillon, Jerome Bettis), or short careers as dominant league-winners (Arian Foster, Terrell Davis, Jamaal Charles, Larry Johnson). Todd Gurley and Le'Veon Bell both already belong to this club, and Ezekiel Elliott is one good season away; how they fare in the next few years will dictate whether they finish their careers here or ascend to a higher tier.
- 300+ points: Roughly corresponding to either two big years or four to five startable seasons, this is around the level a player needs to reach before you'll start to remember them a few years down the road. In total, 95 running backs have reached this threshold since 1985, including disappointments like Darren McFadden (306 points), Knowshon Moreno (304 points), and Ronnie Brown (332 points) and pleasant surprises like Rudi Johnson (367 points), Michael Turner (362 points), and Ahmad Bradshaw (333 points).
One last thing: in case it's not obvious, the thing we truly care about is not how much value a player will have over his entire career, but how much value he'll have over his remaining career. Le'Veon Bell only needs about 208 more points of value to top 1,000 for his entire career. 1000+ points of value is a lot, but 208 points of value is not very much at all; it's what both Danny Woodhead and Giovani Bernard have provided in their careers to date. If you trade for Bell because you think he's a borderline Hall-of-Famer and he returns value equivalent to Danny Woodhead, you will likely be dissatisfied with the deal. Points from 2017 no longer count.
For the most part, I just wanted to lay all of this out today so I wouldn't have to devote a wall of text at the beginning of each article to explain what everything means-- in future installments I'll get into the nitty-gritty and take an in-depth look at actual players and the paths their careers might take. I'm excited to share my passion for NFL history with you and hope that it helps you become better fantasy football players in the process.
More articles from Adam HarstadSee all
More articles on: ForecastSee all
Prop Talk Week 7 - Knotts
Offensive Line Rankings and Notes: Week 7 - Bitonti
Week 7 Offensive Sleepers - Bloom