IDP Draft Capital and Hit Rates: 2021 Edition

Posted on by djkelltown
“NFL Draft 2010 Set at Radio City Music Hall” by Marianne O’Leary is licensed under CC BY 2.0

Soon it will be rookie draft season for dynasty leagues. Every year we get a new group of young players to debate, rank and then draft. You can probably tell yourself a story about how any player can succeed and have a long productive fantasy career. None of them have had a chance to disappoint us yet.

Draft capital is unquestionably predictive of future success. There are several places to read about the impacts of draft capital for offensive players. I personally like Peter Howard’s work on his Patreon (free).  However for Individual Defensive Players (IDPs), there’s not so much out there.  This article comes in to fill that void.

This article is intended to help with rookie drafts in dynasty leagues, by showing you the historical hit rates of IDP players with similar NFL draft capital.

Updates from Last Year

If you have read last year’s article you will find that there are several changes this year

  • All the players are now grouped by their “True Position”.  This means that nominal 3-4 Outside Linebackers like T.J. Watt and Von Miller are classified as Defensive Ends, and 3-4 Defensive Ends like Leonard WIlliams become Defensive Tackles.  Sleeper is using true position already, ESPN changed to this format in 2020 and MFL is changing to true position in 2022.  The hit rates are different across each position, so it makes sense to group players with similar roles together, instead of a depth chart position which can have different roles from team to team.  Every fantasy website classifies players a little bit differently, and I’ll discuss how I’ve assigned positional designations further down.
  • In addition to adding the 2021 season, I’ve gone back and added the 2012 and 2011 seasons.
  • In past years I’ve presented the hit rates for a top 24 finish only, but this year I have included multiple thresholds.  There are many variations of starting requirements in IDP – I play in 5 dynasty leagues and none of them are the same. My hope is it gives you, the reader, some flexibility to apply this to your own leagues.
  • Previously I’ve shown hit rates for “3 position” (DL/LB/DB, think Sleeper) and “5 position” (DT/DE/LB/CB/S, think MFL) leagues.  This year I’ll only present “5 Position” hit rates, with some quick rules to help adjust for your “3 position” leagues.
  • I’ve added some info on repeat rates at each position.  We can’t just assume that each position has repeat performers at equal rates (and they don’t).

Comments on Methodology, Scoring, Position Designations

All hit rates shown are for players who hit in their first 3 NFL seasons. Waiting more than 3 years for a player to breakout clogs your roster, and there’s a good chance you have cut the player by that point.

The scoring used here is IDP123, which is a 2 point per solo tackle system, and the current scoring default on Sleeper.  The best place to learn more about that scoring is Jordan Rains’s Dynasty Nerds article here, or his tweet thread on the topic here.  However for the purposes of rookie hit rates, the scoring doesn’t change the analysis unless you are using something completely different than this or the Fantasypros standard scoring.  I’ve done some review on IDP scoring previously here.

Defensive statistics are from Pro Football Reference and scored myself, so there are some small variations between what I have scored and what you might see on your favourite fantasy site if they use different statistic providers.  It also means that I have not captured any offensive statistics, so I don’t have Christian Wilkins’s touchdowns in 2019 or 2021, Jeremy Chinn’s rushing yards in 2020, or anyone else’s offensive stats accounted for.  This is defensive stats only.

Positional designations were primarily determined using free information available PFF – I don’t have the budget for an Elite subscription to get the snap counts by alignment – as well as any other info I could find on old rosters or depth charts.  Since every platform has variations on how they assign positions, there will be some differences between what each site has and what I have, and I’ve chosen to live with those differences.  This is also somewhat of a retroactive continuity for the early years in the data set, because while we have complained about the way pass rushers get classified by their depth chart position for as long as I can remember playing IDP, The true position revolution on fantasy platforms is relatively new (at least as far as I know).

Some players legitimacy change positions over the course of their careers.  For the purposes of rookie hit rates those players are grouped with their initial position, unless it was clear that they were changing positions when you would have drafted them as rookies.

Results

Defensive End

Defensive End is influenced the most by early draft capital.  The top half of the first round is very good, but that very quickly changes and the hit rates drop substantially by the end of the round.  Generally all subsequent rounds have worse hit rates than any other position shown here. I would argue that for Defensive End, only the first round should be considered “good” draft capital.  It is the only position with no top 36 hits in either the 6th or 7th round.  Day 3 Defensive Ends should generally be avoided in rookie drafts.

Moving all the pass rushing Linebackers to this group hasn’t changed the hit rates very much at all – the top 24 hit rate for top 16 picks is exactly the same as last year.  This validates re-classifying all the players to True Position for the purposes of determining hit rates.

Defensive Tackle

I’ve chosen to not include a Top 36 hit rate for Defensive Tackle because after the 24th Defensive Tackle, they score significantly fewer points than any of the other positions.

Defensive Tackle is otherwise very similar to Defensive End, except the hit rates look marginally better  after round 1. That difference is 1 or 2 extra players from each draft round having a hit season.

Whether you are required to start 1 or 2 Defensive tackles will make a big difference in terms of how you apply this table to your league.  There is only 1 player out of the 134 day three picks to have a top 12 season this early in their career, whereas there are 10 players in that group with top 24 finishes in their first 3 season.

The gap in hit rate from the top to the bottom of the 1st round appears to be unique to Defensive Ends and Tackles.

Linebacker

Linebackers have similar hit rates the in the top and bottom halves of the first round, and the sample size is much smaller than Defensive End and Defensive Tackle, so the split in the round isn’t presented here.

The big difference from this year to last year is once you remove the Linebackers who primarily pass rush (they have become Defensive Ends), the hit rates rise dramatically, especially in the first round.  A 76% hit rate for a top 24 season is the best of any position and round, though they still lag behind top 16 pick Defensive Ends and Defensive Tackles for hit rates for top 6 and 12 finishes.

The third round has a 22% hit rate for a top 24 finish, which is about where I would draw the line of having good draft capital.  That said there are many more players who have fantasy relevant seasons from rounds 4 through 7 than there is on the defensive line.

Cornerback

I’m inclined to stream Cornerbacks as much as possible, but some dynasty leagues don’t have good options on waivers so you have to keep several on your roster.  Unfortunately for our rookie drafts the hit rates on Cornerbacks with 1st or 2nd round draft capital is the worst of any of the defensive positions, with the exception of Defensive Ends (the 1st round for Defensive Ends as a whole is worse, but that position has a noticeable difference between the top and bottom halves of the round while Cornerback does not). 

Late round picks don’t make up the difference either.  More Cornerbacks are drafted than any of the other defensive positions, which dilutes their hit rates.

Safety

Safety is the flattest position.  The hit rate for elite seasons (top 6 finishes) is about the same for each of the first three rounds.  This is a red flag to me.  Safety is so dependent on deployment for a good fantasy season, and NFL teams are looking for different things from that position in real life than we are as fantasy players (we are looking for a bunch of tackles, mostly).

Adjustments for 3 Position IDP Leagues (DL/LB/DB)

I chose to only show only the 5 position tables for clarity, however you may play in a 3 position league where DT/DE and CB/S are combined.  I looked at the scoring results since 2011 to see how the top 12, 24, 36 and 48 split between each of the positions.  I said above that the scoring used doesn’t matter much for hit rates, but that is less true here. In particular, the scoring for passes defended (PD) relative to tackles will affect the number of Cornerbacks and Safetys with high finishes at the DB position (generally Cornerbacks get a few more PD’s).  It is important to check your own league scoring and past results.

Defensive Line

  • On average 2.2 Defensive Tackles finish in the top 12 DLs (range 0 to 5)
  • On average 4.3 Defensive Tackles finish in the top 24 DLs (range 1 to 8)
  • On average 8.7 Defensive Tackles finish in the top 36 DLs (range 5 to 13)
  • On average 13 Defensive Tackles finish in the top 48 DLs (range 7 to 18)

Defensive Back

  • On average 2.5 Cornerbacks finish in the top 12 DBs (range 1 to 6)
  • On average 6.8 Cornerbacks finish in the top 24 DBs (range 5 to 9)
  • On average 13 Cornerbacks finish in the top 36 DBs (range 7 to 15)
  • On average 18 Cornerbacks finish in the top 48 DBs (range 10 to 20)

If you are playing in a league with a combined DL or DB position and you only start 2 of each, you are really looking for elite outcomes from Defensive Tackles or Cornerbacks. If you start more than 2 of each, then those positions are more viable, but they look very much like floor plays to me as most of those players appear to slot in your 3rd and 4th DL/DB spots.

Repeat Rates and Late Hits

We want players that hit, but it’s even better if they hit for multiple seasons. The following table shows the average number of top 24 seasons per player, split by players who had at least one top 24 season within their first 3 NFL seasons, and players who had their first top 24 season later.  I restricted this to the 2011 to 2018 draft classes to only include players who have had the opportunity to have repeat top 24 seasons.

Defensive Linemen are more likely to repeat that performance than other positions.  Linebacker and Safety share the same repeat rate, while Cornerbacks have the lowest frequency of repeating.

I haven’t presented this by draft round here, but generally the earlier rounds have better repeat rates than late rounds. That said, the group sizes outside of the early rounds are all small. After round 3 the biggest group is 4th round linebackers with 6 players who have early top 24 hits.

Tom Kislingbury of Dynasty League Football (DLF) has looked at this through the lens of year-to-year repeats and found similar results.

Depending on the position, the number of players who have breakout seasons later in their career is about 35-55% of the number of players who have breakout seasons earlier. There also isn’t a distinct pattern of what rounds these players are coming from, except for Defensive End, where 9 of the 17 late breakouts are first round draft picks.

The average number of top 24 finishes from late breakout players varies between 1.4 and 1.7 for each position, suggesting that late breakout players don’t have the same longevity as their early breakout peers, but there are exceptions. Jordan Poyer changed positions from Cornerback to Safety and has had five consecutive top 12 Safety finishes, starting with his 5th NFL season. There is also a little bit of skewing in this sample, in particular with the 2018 class who’ve had a chance to breakout in their 4th year, but haven’t had a chance to repeat it yet.

Conclusions and Next Steps

If you have read my past work on IDP rookie hit rates, the biggest change is that once you filter out edge rushers from the Linebacker group, the remaining Linebackers have excellent hit rates compared to the other positions. Before you filter out edge rushers, the best hit rates come from Defensive Ends with early 1st round draft capital. Tripp Brebner of Dynasty Football Factory made the following comment when I posted my first study of IDP draft capital in 2020.

I think the higher repeat rates at Defensive End are significant enough that a Defensive End drafted with a very early 1st round pick should still be considered ahead of a 1st round Linebacker, although that is going to depend on your league lineup requirements, scoring and whether the league is using True Position designations or not.

A couple of things I haven’t addressed here that I think will make for good future articles:

  • More work on repeat rates of players who have hit.
  • A closer look at successful and unsuccessful players with 1st round draft capital to see if the middle of the round is an appropriate place to split players, or if other spots are better (I’ve shown here that Defensive Linemen have a break point in the first round. It is worth looking at again for all IDP positions).
  • The choice to use the first 3 years as a cut off for hit rates was somewhat arbitrary, and necessary for me when I first started this with the computer knowledge I had at the time. I think that is probably the correct spot but checking whether the 3rd and 4th year breakout players have similar career trajectories to 1st and 2nd year breakouts or to late breakout players is worthwhile.

Thank you for reading this study on draft capital and its impact on fantasy IDP players. Feel free to reach out to me with any questions or comments. You can find me on Twitter @djkelltown.

The Cornerback Streaming Experiment: 2021 Edition

Posted on by djkelltown
“WR Edgar Poe 1” by West Point – The U.S. Military Academy is licensed under CC BY-NC-ND 2.0

NOTE: This is a article that will be updated throughout the season.

Introduction

Welcome to the first edition of the Cornerback Streaming Experiment. The Streaming Cornerback Experiment is a living document that will be updated throughout the season, and will detail my Cornerback plays in the 2021 IDP Invitational run by IDPguys.org.

Cornerbacks can be unpredictable, at least season to season as noted in the tweet below from Tom Kislingbury of Dynasty League Football.

The IDP Invitational uses 22 total starters and 8 bench spots, so bench allocation is important. In fact, there are less bench spots than unique positions (9) in this league. Only 1 Cornerback is required in the starting lineup (2 if used in the flex), so this is an ideal league to try streaming, as there are plentiful options on waivers.

The benefit of being able to stream the position is the draft capital freed up to spend on another position. For example, here are all the Cornerbacks taken during my draft.

By spending a pick in 12th to 14th round on a Cornerback, teams were passing up on drafting their 4th WR, or their first edge rusher

The idea of streaming Cornerbacks is not new. Johnny The Greek wrote a weekly series last season detailing his start and sit recommendation at the position for IDP Guys. He described his process as follows:

We’re going to be looking at things like expected game script, opposing offensive tendencies, snap counts, role and historical production to inform our weekly decisions on what corners are an “ideal” streaming option. (From: http://www.idpguys.org/johny-the-greeks-cornerbacks-corner-week-sixteen-16/)

I will be generally following that method, but also adding some of my own ideas, which may evolve throughout the season. For example, I plotted team CB points allowed by pass attempts, interceptions, etc., and found that the strongest correlation was to pass attempts to Wide Receivers, as opposed to total attempts.

One issue with the chart above is that it is using team CB points allowed. It doesn’t adjust for the number of total Cornerback snaps being played against a team. Pittsburgh and Cincinnati among other ran a lot of 3WR sets last year. It would make sense that they also play against more nickle or dime than other teams, reducing the effect.

Benchmark

As this is a start 1 Cornerback league, we are aiming for streaming to cover at least the CB12 score, but ideally higher than CB6. Based on last year’s scoring, we are aiming for about 12.2 ppg for a top 12 finish and 15.2 ppg for a top 6 finish.

I want to note that past work I’ve done on rookie hit rates shows that the CB6 finish generally finishes around DB24. So if you can successfully stream the CB6 in a league, it’s potentially valuable even in leagues that start Defensive Backs as a group instead of Cornerback & Safety

Picks and Results

Week 1

Joe Haden, PIT vs BUF – Based on ESPN’s projections from July 14, 2021, Buffalo was projected to have the 2nd most WR targets this season. While Buffalo is the favourite in this game (-6.5 per The Score app), they passed on 57% of plays when up 7 or more points (via sharpfootballstats.com).

Pittsburg is also a good match up for cornerbacks, and project to have a more favourable game script as the underdog, however Tre’Davious White and Taron Johnson were both drafted earlier in the draft, and I didn’t feel comfortable that Levi Wallace will play 100% of snaps.

Result – 12.5 points (3 solo tackles, 2 assists, 1 forced fumble). The matchup ended up being better than anticipated as Buffalo trailed late and attempted 51 passes. Haden made the best of only 4 targets. Cam Sutton ended up being the better play as he was targeted 7 times and had 19.5 points. Regardless, Haden’s 12.5 points was right around the expected CB12 season average.

Week 2

Chris Harris Jr., LAC vs DAL – Dallas is expected to be leading the league in pass attempts in 2021, and attempted 58 passes (41 to WRs) in week 1. Dallas is a 3 point underdog (via The Score app). Both Dallas and Los Angeles were in the top 12 in offensive pace (via sharpfootballstats). Numberfire projects this game to be one of the fastest in week 2, and projects Dallas to have a 77% pass rate.

Chris Harris played 98% of snaps in week 1 and – as far as I could tell – played primarily in the slot .

Saturday night revision: Michael Davis, LAC vs Dallas – same as above. Harris was declared out for this game. I lost having the player who is playing in slot, but I’d rather Davis, who just about played every snap last week. Asante Samuel Jr. played about 85% of snaps in week 1, but I feel less comfortable projecting his usage.

Result – 12 points (4 solo tackles, 2 assists). Davis played 100% of snaps, but was only targeted twice. Samuel got the interception, so in the end he was the better play.

Week 3

Byron Murphy, ARI vs JAX – Through 2 weeks Murphy has been the only Cardinals CB to play near 100% of snaps in both games. Both Arizona and Jacksonville are near the top of the league in pace of play. Jacksonville is about a touchdown underdog, and are leading the league is pass rate so far. As a bonus, we get a rookie QB who hasn’t played well so far this year.

Result – 31.5 points (3 solo tackles, 1 assist, 2 passes defended, 2 interceptions, 1 touchdown) – So this obviously was a good result, but I think it’s noting that this wasn’t actually a good game to stream off of in the end. Jacksonville had 34 passing attempts, which is right around the league median through 3 weeks. Jacksonville had 29 rushing attempts in this game, including 6 from Trevor Lawrence), something that is happening a lot to Arizona this year.

As I’m updating this late in the week, I’ll note that Jacksonville only had 24 attempts on Thursday Night Football. It is looking like their 51 attempts in week 1 against Houston might be noise, and they aren’t going to be as pass happy as I thought after 2 weeks.

Week 4

Vernon Hargreaves, HOU vs BUF – So I have going back to streaming against Buffalo this week, a team near the top of the league in pace and passing rate in all situations. Houston is a 15 point underdog here, so the hope is that Buffalo passes even while winning and/or Houston keeps this game closer than expected for a while.

I could have used Byron Murphy again against the Rams, who also pass at a high rate a run a lot of plays. I didn’t want to partly because I don’t want to re-use someone here in back-to-back weeks, and also because Arizona has been run on a lot… a lot more than you would think…

It is entirely possible that I’m putting too much weight into this, because the Rams are otherwise a good match-up for streaming the Arizona Cornerbacks.

Result – 2.5 points. Hahahaha this was a horrible pick. Hargreaves lost his starting spot and played only 14 snaps in the game

Week 5

Charvarius Ward, KC vs BUF – Taking advantage of a game with two very pass heavy offenses, this pick is going back to playing against Buffalo (see week 1 & 4). Not much to say here. Buffalo one of the fastest and highest neutral-script passing teams. They re a slight underdog in this game.

Result – 0 points. Hughes played 32% of snaps this week, before going back to 80% the following week.

Week 6

Anthony Averett, BAL vs LAC – I didn’t do a proper update before this game, but LAC passes a lot and plays fast. The betting line was also close so a decent script was expected.

Result – 16 points

Week 7

Donte Jackson, CAR vs NYG – Some of my favourites are on bye (LAC, BUF). The giants actually play fairly fast. They throw a lot, because they don’t win a lot. I didn’t want to play a Lions (vs Rams) or Bears (vs TB) player because I don’t expect those games to stay close.

Result 12.5 points (YTD 87 points (12.4 ppg; tied with Avonte Maddox for CB4 on the year (not accounting for bye weeks) <- That’s wrong. Not sure how I got mixed up there.

Week 8

Taron Johnson, BUF vs MIA – both teams play quickly. Miami is expected to trail, and they’ve thrown a lot more with Tua back (47 pass attempts vs Jacksonville)

Result 12.5 points

Week 9

Darius Slay, PHI vs LAC – 7.5 points

Week 10

Jamel Dean – 11 points

Week 11 –

Xavian Howard – 12.5 points

Week 12 –

Fabian Moreau – ATL vs JAC. Jacksonville has given up a lot of points to CBs this year

Result – 7.5 points

Week 13

Nevin Lawson (JAX vs LAR) – 17.5 points

Week 14

Sean Murphy-Bunting (TB vs BUF) – 18.5 points

Week 15

I’m finally doing this early enough in the week to write a proper write up.

Xavian Howard (MIA vs NYJ) – The Jets are a sneaky 4th in WR targets this year and 8th in point allowed to CB’s (via idpguru.com). The Jets are also about a 10 point underdog on the road, so it’s a good script for Howard.

Result – 7 points

Week 16

Desmond King (HOU vs LAC) – At least the third time I picked a chargers opponent this season. Same story as the previous weeks

Result – 27.5 points – With Houston surprisingly winning this game, it led to a better script for King.

Week 17

I was fortunate enough to make it through to the final week of the tournament, and got to select a full week of picks this year.
Byron Murphy (ARI vs DAL) – I’ve used Murphy once before. Since week 9, both Dallas and Arizona are top half of the league in pace, and Dallas is close to the league lead in team attempts and WR attempts.

Result – 20 points – 7 solo tackles and 1 pass defended is a nice result to end the year. Dallas attempted 38 passes in this game.

Season Summary

Total points scored – 228 (13.4 ppg), I beat the benchmark of 12.2 ppg which was projected to be a top 12 CB finish, but was below the 15.2 ppg benchmark for a top 6 finish.

Having said that, here are the top 12 Cornerbacks on the season, after week 17

Source: IDP Invitational League on Fantrax

It’s notable that the 13.4 ppg I achieved this season was beat by every player on the list. 13.4 ppg ends up around 14th or 15th at the position this year (varies because I’m sorting by total points). Of the top 12 players shown above, Moore, Ryan, Diggs, Lattimore and Terrell were drafted in my division, with the other 7 players undrafted.

Lastly, inspired by reading Adam Harstad’s Dynasty, in Theory: You’ll Never Stand in the Same Stream Twice (yes it is 6 years old now, yes, I read it for the first time 4 weeks ago, and also yes, it’s a better read than what you are currently reading so you can just stop and go read that), I compared the points I scored with the players I started vs their seasonal average. Here are the results:

Harstad looked at Quarterback and Team Defence Streaming and found that players did a bad job of picking matchups throughout the season. Here I scored a little more than 1.8 points per game higher for the players I started their combined season averages. So streaming Cornerback may be more viable than Quarterback or Team Defense, but this is also a sample size of 1 team, where Harstad looked at 42 teams.

Conclusion

So does streaming the Cornerback position work? I’m inclined to say that you can get pretty to your baseline starter production, in this case near CB12 production, without spending any draft capital. I’m not sure whether you can get a season of one of the elite players, but the challenge there is knowing exactly who those are and then you have to potentially spend the draft capital to get them (like a 14th round pick on Kenny Moore).

As the season went on I focused more on players with good matchups and had scored a bunch of points throughout the season, as opposed to just any corner getting full time snaps with a good matchup. I think that helped achieve better scores in the 2nd half of the year.

Find me on twitter @djkelltown. When I’m not talking about fantasy football, I’m probably talking about my NHL team (Winnipeg Jets) or CFL team (Winnipeg Blue Bombers)