Those of us who play daily fantasy sports usually have a pretty extensive history with season-long fantasy sports, and that’s true of myself and fantasy football. I’ve been playing season-long leagues since 1988 and have enjoyed that pastime so much that I’ve written a book about fantasy football draft strategies for five years running now.
This will be the fifth straight year that I’ve written a fantasy football draft guide. Like previous iterations, this year’s version still relies on a data-driven approach to outlining various draft strategies to gain an advantage on your opponents. Also like past editions, it contains plenty of charts, graphs, links to other sites, and detailed explanations of game theory and analytics and how these concepts can lead to better fantasy football drafts.
I wanted to start off by writing a section that is friendly to fantasy football newcomers while still giving veteran players something to think about and maybe question any longstanding beliefs about fantasy football drafting that they may hold. To do this, I’m going to show the results of a recent mock draft for a 12-team PPR (point per reception) league that I recently took part in on fantasyfootballcalculator.com. I’m going to discuss the results of this mock draft and share my thoughts on it using the experience I’ve accumulated from playing fantasy football for the past couple of decades. I’m also going to introduce some of the concepts and statistics that you’ll find in the draft guide.
Mock drafts are a tremendous asset to aid you in your quest to win fantasy football leagues. Whether they are human mock drafts or automated ones, they give you insight into where players are currently being picked for the upcoming season, what positions are stronger and weaker than in previous years, and which players are being drafted higher or lower than what you expect. The results of mock drafts can then be used to adjust your player rankings and draft strategies to put you in position to pick the players that you desire in the rounds that you want to draft them to maximize value.
Below is an image of the mock draft results:
First, a few things about the mock draft. I picked my position of drafting eighth overall for this PPR league and cleverly named my team “2019 Mock.” There were two other human drafters in the mock: someone who chose the name of “Liberal MEDIA” picked sixth overall and someone who just went by “H” (and you thought my name was boring) picked from the tenth spot. The draft was a snake draft. Both of the other human mock drafters left after about six rounds or so, whereupon the “computer” finished their mock draft by making the picks. I used “computer control” throughout my mock draft to see how the automated draft would turn out for me.
The main reason I’m doing this is because this mock draft looks very similar to real drafts that will be taking place this year. The first two rounds are dominated by running backs and wide receivers with a couple of elite tight ends also being selected. The top drafted quarterback doesn’t go until round three and you’ll see no hint of a defense or kicker taken for quite a few rounds. Like I said, it’s a typical fantasy football draft that looks like hundreds of others that I’ve taken part in over the years. The question is, though, why is this such a typical looking draft? Why are certain positions selected high in the draft and why are others left until much later? It’s conventional wisdom that you draft this way, but is this typical-looking draft really the best way it should be done? Are the assumptions we’ve made for years indeed correct or is there a mathematically proven better way to draft? That’s the main issue this guide will tackle.
Now let’s get back to the mock draft, specifically the picks the computer made for my team because they’re quite illustrative to some of the points I’d like to make.
First off, it took two running backs in the first two rounds for me. Specifically, Melvin Gordon in the first round and Todd Gurley in the second round and both have question marks. If I was drafting, I would’ve went with Le’Veon Bell with that pick in the first round, but I have no real problem with Gordon provided he gets his contract situation sorted out. If he holds out long into the preseason, that will lower his ranking on my board, and if he sits out he’s obviously not going to be drafted by me, but I assume that’s pretty obvious. Anyway, I would’ve taken a runner in the first like I almost always do and Gordon is a proven commodity on a high-powered offense. In the second round, I would’ve been on the lookout for a top wide receiver or, if one that I liked wasn’t there, then go back and get my RB2. Gurley was the top back in fantasy football last year, but we all know about his arthritic knee condition (or you should if you don’t) but getting the top back from the year before as the eleventh drafted back overall could be a huge steal if he’s even eighty percent of last year’s form. I probably would’ve gone with Antonio Brown there myself because the cardinal rule of drafting for me is to do everything you can to not lose your league in the first several rounds, and taking a running back with the potential for an injury-riddled year with your second round pick qualifies as a big-time risk. Obviously we’ll learn more about Gurley’s situation as training camp and the preseason plays out, but right now there are too many questions to take him with that pick in my mind. Arthritis is common in knees with previously torn ACLs, which is what Gurley had while at Georgia. Aside from the first few games of his career and the last two games of 2018, Gurley hasn’t missed a game due to injury in the NFL, but I expect he will have a reduced workload this year at the minimum.
The QB Question
Now we get to the interesting part. The computer picked Pat Mahomes for me with my third round pick — the first QB drafted overall. Mahomes set a fantasy football record with 417 points scored last season (using four points for TDs and one point per 25 yards passing). It was a remarkable year and Mahomes scored the most fantasy points in all of fantasy football in 2018 regardless of position. So he must be a pretty decent pick in the third round, right?
Well, not so fast, I’d say. There are a few things we need to consider first. The biggest one of these is how do quarterbacks compare to other positions in fantasy football in terms of VALUE. That’s what you’ll read throughout this guide — value is king. So let’s do a little research to determine if it’s a good idea to draft Pat Mahomes in the third round because it seems like a no-brainer at first glance.
As I mentioned earlier, Mahomes scored a fantasy football all-time high of 417 points last season. If he duplicates that year, then he would unquestionably be worthy of being picked in the third round of any fantasy football draft. But here’s the rub: of the top ten highest scores in fantasy football last season, nine of those came from quarterbacks. In fact, 14 of the top 16 highest scoring players in standard scoring (non-PPR) leagues last season were QBs. If you use PPR scoring, then four of the top 10 highest scorers were QBs and the 14 of 16 highest in standard scoring leagues becomes 14 of the top 29 in PPR leagues. Still, that’s almost half. Therein is the reason why QBs aren’t generally drafted in the first two rounds of a fantasy football draft: there’s so many of them that score a lot of points when compared to other fantasy football positions.
Also, you have to factor in how likely it is that Mahomes has a season somewhat comparable to the one he had last year. Well, here’s some information on that.
Year ~~ Top QB (FPTS)~~~~~~~~ Finished the Next Season at QB(x) (FPTS)
2010 ~~ M. Vick (316) ~~~~~~~~ QB11 (241)
2011 ~~ A. Rodgers (399) ~~~~~~ QB12 (344)
2012 ~~ D. Brees (346) ~~~~~~~~ QB2 (358)
2013 ~~ P. Manning (414) ~~~~~~ QB4 (313)
2014 ~~ A. Rodgers (358) ~~~~~~~ QB7 (305)
2015 ~~ C. Newton (391) ~~~~~~~ QB18 (260)
2016 ~~ A. Rodgers (382) ~~~~~~~ QB29 (130) INJ
2017 ~~ R. Wilson (348) ~~~~~~~~ QB9 (298)
2018 ~~ P. Mahomes (417) ~~~~~~ QB?
The above list shows the top scoring fantasy quarterback in each season from 2010 through 2018. The first thing you should notice is that not a single QB ever repeated as top fantasy QB the following year. Aaron Rodgers does appear on the list three times, but the following seasons have not been good for him. If I throw out Rodgers’s 2016 season because he missed the majority of it due to injury, the net points lost by each top fantasy QB from the season they led to their following season is a whopping -453 fantasy points in seven seasons! That’s an average of -64.7 fantasy points a year. That’s a little over four fewer fantasy points a game and is the equivalent of the difference between the top QB and one who finishes around QB5 or QB6 historically.Considering that QB1 is typically drafted late in round 2 or early in round 3 and QB6 is typically drafted in round 6 or 7, it’s a huge loss of value.
Also, the impact on my team by drafting Mahomes in the third round was such that it resulted in a somewhat weak wide receiver corps, in my opinion. Personally, I would’ve picked A.J. Green, Cooks or Diggs with that third round pick and then instead of having Julian Edelman as my WR1, he would’ve been my WR2 when selected in the following round. If you look at the mock draft, you’ll see that I could’ve drafted solid QBs like Matt Ryan, Philip Rivers and Russell Wilson in rounds 7 and 8. This is where I typically take my starting QB most seasons and I follow that up with a pretty strong backup maybe two or three rounds later.
Let’s add a graphic to better show quarterback value when compared to other positions.
This table shows the top 18 scoring positions for QB, TE, DEF and PK from the 2018 season and the top 24 scoring positions at RB and WR. I’ve done the top 18 for the positions that usually only require one fantasy starter at that position per week (QB, TE, DEF, PK) and 24 for those that require two starters at the position (RB and WR). At QB, in the first column, first row you have “1:15”. This means that the top scoring QB last season (Mahomes — QB1) was picked at QB15 (around the end of round 10 in 12-team PPR drafts) from an average of drafts that can be found at fantasyfootballcalculator. Any player or position labeled with “ND” means that they were “Not Drafted” on average in 12-team PPR drafts in 2018.
QB2 in 2018 was Matt Ryan (354 fantasy points) and he was drafted, on average, as QB13 (round 9). Any player whose cell is marked with yellow, regardless of position, is one who was drafted as a backup at his position in 2018 but finished among the top players at his position. As you can see, four of the top ten quarterbacks were guys who were drafted as backups in a 12-team draft (drafted as QB13 or lower). If you look at RB and WR, the very top highest-scoring players at the end of the season were ones who were drafted higher, especially at the WR position where not a single one of the top-10 WRs was picked as a backup (WR25 or lower). So this graphic is yet another case for why taking a QB high is riskier than RBs and WRs.
Winning Teams:
In the mock draft, the computer selected only four players that finished on losing teams last season for my team (two from Green Bay and one each from the Giants and Broncos). Since the goal of winning a fantasy football game is the same as that of winning an NFL game — scoring more points than your opponent — are there data to suggest that drafting from winning teams is better? At least among the top fantasy football scorers, there clearly is. Here’s a bar chart that shows it:
The above chart shows the number of players who finished among the top-20 scorers ranked by VBD number in PPR leagues from 2014 through 2018 that were on losing teams (in red). The VBD number is one that you will certainly learn more about later in this book, but it basically compares players between positions and assigns a value to their overall worth. Anyway, the most players on losing teams in that time period is 6 out of 20, while the least has been 3 out of 20. That means that between 14 and 17 of the top-20 fantasy scorers per year came from teams that finished with a record of .500 or above. The top-20 players that were on losing teams last season were Saquon Barkley, Christian McCaffrey, Davante Adams, Julio Jones, George Kittle, and Mike Evans. Among your top players, you’ll find that about 70-85% a year will come from non-losing teams. Of course, we can’t predict what teams that finished with losing seasons last year will be surprise winners this year and what winning teams will take a tumble this upcoming season, but generally your perennial playoff teams with good offenses are a good and safe bet to take your top players from.
Byes and Backups:
One thing that the computer did when it drafted my team was select the following wide receivers (player name followed by team and bye week):
Julian Edelman (NE) 10, Tyler Lockett (SEA) 11, N’Keal Harry (NE) 10, Geronimo Allison (GB) 11, Marquez Valdes-Scantling (GB) 11, DaeSean Hamilton (DEN) 10
So what stands out when looking at these players? First of all, look at the bye weeks. All six of my wide receivers have bye weeks during weeks 10 and 11. That’s completely unacceptable and you should never have that many players at a position with only two different bye weeks between them. Clearly the algorithm that did the mock drafting didn’t take into account bye weeks and that’s something that you absolutely must monitor during your draft.
The other clearly noticeable aspect that stands out is that the computer drafted two wide receivers from two teams — New England and Green Bay. Moreover, the two Green Bay WRs are probably the WR2 and slot receiver (WR3) for the Packers heading into the season while N’Keal Harry, knowing Bill Belichick’s general aversion to playing rookies in the past, might not see many snaps early in the season despite being the final pick in the first round of this year’s NFL Draft. I absolutely do not like taking two WRs from the same team even if the second receiver is a top one like JuJu Smith-Schuster was last season. He was drafted around WR18 in PPR leagues last year while teammate Antonio Brown was usually WR1 in most drafts. Even as a Steelers fan, I wouldn’t have drafted both of them on my team. The chance of the team getting shut down and both putting up bad fantasy games far outweighs the benefit of them both having good games. You want to mitigate risk during a draft, and taking two wide receivers from the same team is the opposite of mitigating risk, you’re exacerbating it.
There are very good second receivers heading into this season, though. In a few cases some teams don’t have a clear-cut top WR like you have in Minnesota with Diggs and Thielen or Cooks, Kupp and Woods from the Rams. But you also have clear WR2s who happen to play along a star WR but are very good in their own right like Tyler Boyd, Calvin Ridley and Mike Williams from the Chargers. Good WR2s on their own team are very much worthy of being picked in the late-teens or just a bit later — just don’t pick two from the same team.
Summary:
To wrap things up, I’ll outline my general fantasy football draft philosophy, and it’s a pretty simple one. In the first four rounds of any standard-type draft (non-two-QB, non-superflex, etc.) I’m usually going with two RBs and two WRs. In what rounds they get selected depends on the type of scoring format used and where I think the value clusters for those two positions fall on that given year. Some years RB is thin after the top dozen or so backs while other years it’s deeper. When it’s deeper, I like to get a top WR in round two. This year there are about 20 running backs that I’m comfortable with having as either an RB1 or RB2. That’s thinner than usual, so I will probably go two RBs in the first three rounds of most drafts.
Round 5 is usually the wildcard position for me. Sometimes I’ll draft a QB that I think is going to have a very good season there and sometimes I’ll go after a good TE. If no player at either of those positions really pops for me during a given year, then I’ll add my third RB or third WR in round 5. I don’t think round 5 is an overpay for a top fantasy QB, but I have to have a pretty good feeling about that QB for the coming season. Last year I took Deshaun Watson in a few leagues around round five or six. Sometimes, many of the QBs I like are bounce-back candidates. I liked drafting Russell Wilson the year before (2017) when he was coming off a subpar 2016 season. Matt Ryan was a guy I liked taking as a bounce-back QB last season around the 10th round — an area of the draft where I have been very comfortable taking undervalued veteran QBs like Roethlisberger and Rivers year after year.
For my bench players, I gravitate toward higher-upside guys in the middle and later rounds of the draft. You’ll get more busts that way but you’ll get your fair share of breakthrough players. Last season I hit on mid-to-late-round guys like Robert Woods, Nick Chubb, Kenny Golladay, Calvin Ridley, George Kittle and Eric Ebron but missed on players like Pierre Garcon, Jamison Crowder, Keelan Cole and Ronald Jones.
Also, since I play in multiple leagues, I’m a firm believer in not putting all your eggs in one basket. I try to spread my exposure to a bunch of different players aside from ones that I basically have a “do not draft anywhere near their average draft position” grade on. It’s something you have to do to win in daily fantasy sports when playing in GPP tournaments since taking only chalk players won’t differentiate your team enough from the rest of the field. For regular season-long fantasy football, you should try and do the same if your goal is to win leagues. For instance, last year I think I had every one of the top-20 running backs by ADP on at least one of my dozen or so paid-league fantasy teams aside from LeSean McCoy, Alex Collins and Jerick McKinnon. I either didn’t like the player or the situation they were in for those three. On the other hand, I really liked Joe Mixon heading into last season but I made sure to have him on only like three of my fantasy football teams because if I was wrong with my evaluation or if he got injured, then having him on half of my fantasy teams was going to hurt a lot. I’m not saying to draft players you don’t like, but if you think a player is worthy of his ADP but he’s not already on one of your fantasy teams, I’d probably take him over a player who you like a bit more but you already have on multiple teams. A player like that for me last season was Lamar Miller from the Texans. He was a guy who had burned me in the past and is someone I think of as merely an above-average back, but he was the clear top RB on what I thought would be a good offense so I made sure I got him on at least one of my fantasy teams and he produced right around what he should have for where he was drafted. I had similarly ambivalent feelings toward James White, who had a much better year than where he was drafted. This is a guy who was drafted around the 40th overall RB taken in PPR leagues but finished 7th in that scoring format. I just haven’t been a fan of taking New England running backs because of Belichick’s committee approach, but I wanted exposure to White in a PPR league at least and made sure to get him in one. I also wanted exposure to Sony Michel, especially in a standard-scoring league, because I really liked him as a player coming out of college at Georgia, so I also got exposure to him on a couple of teams despite him being a New England back. Both guys had very good seasons but wouldn’t have been on my team if I didn’t employ a “player exposure” technique when drafting. Don’t let your biases force you into fixating on certain players while ignoring others.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
If you’re interested in reading the new 2019 version of the Fantasy Football Draft Strategy Guide, here is the link.
