Introduction

I, like many, have an irrational belief in my own drafting ability. I do my research. I don’t actually scout players but I listen to podcasts, I follow scouting twitter and I make a hell of a spreadsheet. Confirmation bias reminds me of all my correct picks. My incorrect picks just need more time or were totally flukes.

I put 10 years of Dynasty Wars rookie drafts into a spreadsheet to see what data I could glean. Dynasty Wars is a 16-team, 1 QB, IDP dynasty league. I used the rookie drafts from 2004-2013 for data. The starting lineup is 1 QB, 2-3 RBs, 2-3 WRs, 1-2 TEs, 1 PK, 3 DT+DEs, 3 LBs and 2 CB+Ss. I wanted this analysis to be useful for all IDP leagues. As scoring can change dramatically from league to league, this purely uses positional rank rather than scoring.

All of my leagues use full IDP starters, so selfishly, that’s what this data assumes, 1 DT, 2 DEs, 3 LBs, 2 CBs and 2 Ss. A cursory review of starting-caliber players shows that about ⅓ of the starting DT+DEs are DTs and about ⅓ of the starting CB+Ss are CBs. So if you’re in a league that lumps DLine and DBs you can adjust accordingly.

Dynasty Wars has an 8-round rookie draft every year. Ten years of 8-round rookie drafts have 1,280 draft picks. The breakdown is as follows:

113	Quarterbacks
219	Running Backs
269	Wide Receivers
86	Tight Ends
42	Defensive Tackles
115	Defensive Ends
240	Linebackers
51	Cornerbacks
103	Safeties
21	Kickers
21	Time Expired/Skipped Pick

Each draft pick was tracked for how often they produced a “startable” year. Starters are assumed as follows: 16 QBs, 32 RBs, 48 WRs, 16 TEs, 16 DTs, 32 DEs, 48 LBs, 32 CBs and 32 Ss. Any player that produced (2) startable years over the course of their career, was considered Startable.

Year	Starters	Startable Rate
2013	18	14.1%
2012	32	25.0%
2011	31	24.2%
2010	38	29.7%
2009	24	18.8%
2008	33	25.8%
2007	29	22.7%
2006	36	28.1%
2005	34	26.6%
2004	39	30.5%
Overall	314	24.5%

This 24.5% overall success rate is fairly useless on its own but can serve a number of purposes in the right context. It can serve as a litmus test. Finding a situation that offers a success rate greater than 25% will offer up an advantage during future drafts.

Continually drafting picks with a better than 25% chance of success will build a stronger team over time. A consistent 25% chance of success will yield 7.5 starters in 5 years with a 6 round rookie draft. Bumping that success rate up to 35% will yield 10.5 starters in the same span. That’s an extra starter every other year.

NFL Draft

Round	Drafted	Starters	Startable Rate
1	255	147	57.6%
2	222	75	33.8%
3	195	39	20.0%
4	188	24	12.8%
5	121	11	9.1%
6	104	8	7.7%
7	89	7	7.9%
UDFA	64	5	7.8%

Buried the headline a bit here. 1st round draft picks perform better. Most of this preliminary information is fairly obvious but we’ll build on it as we go.

 

Rookie draft

Round	Starters	Startable Rate	Offense Drafted	Offense Startable Rate	Defense Drafted	Defense Startable Rate
1	99	61.9%	134	55.2%	25	84.0%
2	59	36.9%	98	30.6%	61	47.5%
3	40	25.0%	94	14.9%	65	40.0%
4	36	22.5%	73	19.2%	84	26.2%
5	27	16.9%	75	12.0%	77	23.4%
6	17	10.6%	76	3.9%	78	17.9%
7	16	10.0%	72	6.9%	82	13.4%
8	20	12.5%	65	3.1%	79	22.8%

Drafting a defensive player in the first round of rookie drafts is rarely a good idea, but that Startable Rate is the highest I’ve seen.

 

A few general observations.

The first thing to learn from looking at a decade of rookie drafts is the peaks and valleys become glaringly obvious. Any draft class has the potential to produce a legit starter at any of the deeper positions, but most classes have one or two positions that produce multiple high-end starters.

The 2007 LB class offered (4) players (Patrick Willis, Paul Posluzny, David Harris and Lawrence Timmons) who each put up at least (7) startable years. Injuries limited Jon Beason to (5) startable years. The class overall produced 42 startable seasons. By comparison, the 2008 and 2009 LB classes have combined for 33 startable seasons.

2008 produced a very strong RB class, featuring Jonathan Stewart, Darren McFadden, Rashard Mendenhall, Kevin Smith, Matt Forte, Ray Rice, Chris Johnson, Jamaal Charles, Steve Slaton and Tim Hightower. All had at least (2) years of RB2 startability. Together, they account for 42 startable years. By contrast, the 2007 RB class which featured Adrian Peterson and Marshawn Lynch, only has 21 startable years.

2010 will be remembered for its TE class. Rob Gronkowski, Jimmy Graham, Aaron Hernandez and Dennis Pitta. 21 startable seasons overall with Hernandez only contributing two.

2011 had one of the best DE classes. Headlined by JJ Watt, it also featured Muhammad Wilkerson, Marcell Dareus, Robert Quinn, Cameron Jordan, Jabaal Sheard and Cameron Heyward. They have combined for 33 startable seasons in 6 years, including (4) DE1 overall seasons and (1) DT1 overall  season. By comparison, 2010, with an extra year, has produced 25 DE startable seasons. The 2009 DE class has only produced 8 startable seasons.

2012 was the year of the LB. Luke Kuechly, Lavonte David, Bobby Wagner and Vontaze Burfict star in a class that has produced 35 startable seasons, not even counting Mark Barron after he switched from S to LB. 2011 has produced 13 LB startable seasons. 2010 has produced 19 LB startable seasons.

The numbers that I will present for each position will bear this all out. Hitting on one or two of these peaks could propel a team to the playoffs on a regular basis.

I think a potentially winning draft strategy is to pinpoint the strongest position and pound it in rookie drafts, regardless of need. Rob Gronkowski, Jimmy Graham, Dennis Pitta and Aaron Hernandez were all drafted in 2010. Someone could’ve easily drafted Graham and Gronk together. Even if you already had a solid TE on the roster, that would create so much more value than drafting a 4th or 5th NFL round WR or RB who never produces a startable season.

Obviously hindsight is 20/20 and you could’ve just as easily drafted Anthony McCoy or Tony Moeaki even if you pinpointed the 2010 TE class. But anything you can do to increase your odds of success, even a little, will pay off over time.

Kuechly and David were both drafted in the 1st round of most fantasy drafts in 2012, but Wagner went in the 2nd round, Zach Brown went in the 4th round and Burfict went in the 5th round. Drafting Wagner, Brown and Burfict would’ve netted you the LB2, LB5 and LB19 in 2016.

I’ve looked for any kind of correlation I can find to forecast what position will blow up every year. The total number of a position drafted doesn’t forecast a productive class. The number drafted in the first round doesn’t forecast a productive class. The total number drafted in the startable range for each position doesn’t forecast a productive class. You’ll still need to scout.

This year, the TE class and S class are both expected to be very good and very deep. I may experiment with a few drafts that only draft TEs and Ss after the 1st round.

Going forward, I’ll put together a post for every position group to look at draft trends and what has worked in the past as a potential starting point for future drafts. The ultimate goal is to create a set of guidelines for drafting to maximize the success rate of every pick.