Posted on Leave a comment

History of WR Changing Teams

In dynasty fantasy football there is a vast spectrum of ways to frame a player’s situation. When it comes to a wide receiver switching teams, the public opinion can be polarizing. For a wide receiver like Robby Anderson or Breshad Perriman, it may be viewed that they are freed from bad situations. On the other hand, elite wide receivers like Deandre Hopkins and Stefon Diggs are severed from their high target shares and quarterbacks connections.  Blow is a  history on these situations to see if there are any observable trends.

Production Dip 1st year vs 2nd year of new Offense

The sample is comprised of all wide receiver seasons dating back to 2000. In a subset of this sample, there are 184 wide receivers under the age of 30, who scored over 5 PPR points per game in a year before changing teams change (5PPG minimum threshold to weed out the truly irrelevant players). The table below shows the results, subset by PPG prior to the move. Each subset progressively isolates the more productive wide receivers in order to see if the effect of the team change is greater or lesser for higher caliber wide receivers. The first row shows the percentage of players that experience a production drop in year 1. The second row shows how many of those players bounced back in year 2. The third row shows how many were still below their pre-move total in the second year of the new offense.

Although the sample size gets smaller (and therefore less reliable) as we isolate the more productive wide receivers, the empirical evidence is obviously not positive for them. At each extra step of isolation, we see a higher rate of production drop in year 1 of the new system. We also see progressively lower bounce back rates (row 2), and progressively lower rates of returning to the former glory they exhibited with their prior team. 

3-Year Average Analysis

To circumvent any outlier seasons from a player the 3-year averages before and after a team change were compared.  Below are the results, this time the rate at which wide receivers experience a drop in 3-year average PPG is shown (row 1), as well as the average size of that drop (row 2).

Again, we see an overall negative outcome for changing teams, and that is amplified with the more productive wide receivers. It is important to note that only 9 wide receivers had over a 15PPG 3-year average so the sample was quite small. Interestingly Brandon Marshall was the only one not to drop in production, he managed to adapt quickly when he moved from Miami to Chicago in 2012.

Takeaways

 Lower Tier WRs- The likes of Breshad Perriman, Robby Anderson and Randall Cobb weren’t staples in starting lineups in 2019, and all have changed teams this off season. Given the findings above and the shortened offseason it’s hard to imagine any of them vaulting into relevance.

Brandin Cooks – Cooks is hard to pin down, his PPG number last season was heavily influenced by health issues, and the three-year sample prior to that consisted of seasons with three different teams. He seems to have an almost comical history of changing team and still producing. Your guess is as good as mine.

Stefon Diggs and Deandre Hopkins – Both of these guys crack the top echelon grouping (Over 15 PPG) and will see their first year in a new system in 2020. History suggests a decline should be expected for both as the average drop in 3-year average for player in that group is 5.08 PPG. Remember only Brandon Marshall has ever maintained or outperformed their prior production from that tier.

Odell Beckham Jr. (Year 2)- As a highly debated player entering his second year on a new team, Odell Beckham Jr. also falls into the top group in terms of productivity prior to the move. Results suggest it is unlikely that OBJ returns to his former ceiling. Only 38% of wide receivers above the 15 PPG threshold managed to rebound following a poor first year in a new offense and none returned to their prior form.

Posted on Leave a comment

Rookie WR ADP Expectation of Production

The previous article explored the ADP movement over a Rookie WR’s first year in the hopes of finding that historically cheap pedigree buys produced a positive return on investment more often than not. Unfortunately, this was not the case.

So what DOES influence movement in ADP? Are you ready for an anticlimax?

Production, production, and production. Yes, groundbreaking stuff, but points per game production in the rookie year can explain a large amount of movement in ADP. This means that even in a “rebuilding” mode team needs to be considering production, insofar as it affects the value of their assets. In fact, points per game ALONE can explain 65% of the variance in ADP following the first year. In the world of data science that is a staggering amount of explanation from a simple regression with one variable. The chart below illustrates this, and with the strong correlation between ADP and trade value, it can be inferred that this production will greatly determine the type of return you may receive on their rookie investment.

ADP post-Rookie Year

So, the advice is to “buy rookies that will perform”?

Short answer, yes. But there is more to be gained here. Let us examine the effect of production when controlling for rookie ADP and Draft Pedigree. When we combine all of these factors into a model, we can calculate the amount of production required for a rookie to expect to maintain current ADP. To be clear, hitting that mark exactly does not guarantee the players ADP will not move, it just means they are equally likely to drop or rise, and what actually happens will depend on the other factors that affect perceived value (Note that not all ADP variance is explained by the model, column 5 of the table below).

The following table shows the calculated average PPG expectation (column 4) for WRs in terms of their respective draft pedigree. Column 3 contains the ADP/PPG slope values. These tell you how many ADP positions these players typically rise or fall for each PPG exceeding or falling short of the expectation.

PPG Effect on ADP

What about individual player expectations?

Below is a table showing the market’s PPG expectations based on the ADP and pedigree of each 2020 rookie.

Rookie Expectations

What does all this mean?

To help with interpretation, we can see that CeeDee Lamb, as a 1st round draft pick with an ADP of 35, carries a market expectation of 11.17 points per game. This means if he hits this mark exactly, we can expect his ADP to stay where it is, ALL ELSE HELD CONSTANT. This doesn’t incorporate exogenous variables such as QB change, injuries/trades concerning himself or teammates, or even just offseason press hype. We can also expect his ADP will rise/fall by 6.63 positions for every PPG he exceeds or falls short of this expectation.

But how likely is it that he reaches his expectation? Well, the last column shows the percentage of 1st round rookies that reached that mark, in this case we may say that if Lamb reaches expectations that would put him in the top 38% of rookie seasons for first rounders historically. So, if you “CeeDee catch radius on Lamb” (you like that one?) and believe his talent and situation is conducive to reaching such a benchmark, then he is appropriately priced. Furthermore, with the 6.63 ADP/PPG slope, we can say that even with the median season for a 1st rounder (10.08 PPG), his ADP will only fall approximately half a round.

Bryan Edwards has a low PPG expectation. Does this mean good value?

Yes and no. Edwards stood out as a “buy” in the building of this table and this seems to be a popular sentiment within the industry. I saw that his PPG expectation was lower than some of the rookies being taken behind him and his upside seemed large with every PPG above expectation correlating to an expected ADP improvement of 11.53 spots. While this is true, only 27% of 3rd round rookies met even this seemingly low expectation of 6.51 points. There is an argument to be made that Edwards is in a unique position to be immediately impactful as a 3rd rounder but that is certainly priced in.

Expensive 2nd Rounders

Another stand out for this study was the risk associated with the highly touted 2nd rounders. Higgins, Pittman and Mims would all need seasons that put them in the top 19% of 2nd rounders historically in order to reach their market expectations. Even Shenault would require a top 23% historical season. Now is a good time for a brief disclaimer; many fantasy players aren’t buying the aforementioned players for their rookie year production. I hear many cite the eventual departure of AJ Green, TY Hilton and Adam Gase (please Lord, free the Jets) as potential tipping points in the ascension of these rookies, and therefore we may expect strong 2nd or 3rd year breakouts. While those narratives make sense, I think the thing to take away from this is that it is highly likely that some or all of the 2nd round rookies will be more or as attainable next year and carry tremendous downside if they don’t see the field in year one; If Higgins has an average season for his draft pedigree(6.84PPG) his ADP is expected to plummet almost 4 full rounds! With a rookie QB and two incumbent WRs ahead of him on the depth chart (3 depending on which version of John Ross shows up this year), perhaps expecting such an exceptional rookie season is a bit bold. We can stroll down narrative street and talk about AJ Green potentially leaving or the rookie connection that may be formed with Burrow, but the fact is you’re making a bet and you aren’t getting fair odds. That is the result of the high price; the average ADP for 2nd rounders is 125 so drafting these guys in the 60s, 70s and 80s is quite rich. In contrast, at a 143 ADP, KJ Hamler only needs a measly 5.24 PPG to meet expectation with a 10.81 ADP improvement for every PPG that he exceeds that expectation.

Conclusion

These are just some of the insights that can be gleaned from this information, there is plenty of data in the table that may surprise you and change your sentiment. The thing to take away is that there are risks and expectations baked into these players’ costs. By all means make your bets, but make sure you know the odds. I hope this article serves less as an action plan but more of a framework in which you can make decisions based on your own speculative expectations. Finally, the importance of production cannot be understated. Even if you are not contending, production will impact the value of your assets.

 

Posted on Leave a comment

Rookie Wide Receiver ADP Movement

This article is written by Carlo Surace.  You can find him on Twitter @carlo_surace

At this time every year rookie hype is at its peak. Proven veterans seem to be kicked to the curb in pursuit of the new hotness. The question is, are these rookies “overpriced”? It is important to note that price/value, mean different things to different people. This article will refer to “market value” and will use ADP as a proxy for this. ADP is significantly correlated to what a player is perceived to be worth in a trade. Thus, no matter how you acquire these rookies, this article will examine the typical “capital gains” one may expect on their investment.

Some may rebut this analysis claiming, ‘market value is useless if the players never produce and help you win’. This is true, but often rookies are acquired in a rebuilding stage to insulate or grow total asset value and not for the purpose of immediately contributing. This is especially the case with WRs, which will be the focus of this analysis. To this end, it is important not to overpay for a rookie WR with a high chance of value decay which will limit the moves you can make later. Perhaps, rookies are a better “buy” following the rookie year once the shine has worn off.

ADP Movement by Pedigree

Below is a table displaying startup (Non-Superflex) ADP by NFL Draft Pedigree for players in their rookie year vs their second year. Nothing too noteworthy here; as expected draft pedigree is correlated to ADP. There is no discernible difference year over year barring round 5, however this is statistically insignificant. Therefore, there is no clear bet that one can make in terms of pedigree to insulate value. As for the percentage that maintained/improved their ADP, the number hovers around a 50/50 bet regardless of pedigree. The exception is round 4 exhibiting a 62% improvement, however this number is slightly overstated as WRs that fall outside the top 250 after year one were removed from the sample.

NFL Pedigree Rookie Year ADP 2nd year ADP Maintained or improved ADP
Round 1 72.399 70.4 44%
Round 2 125.607 119.25 46%
Round 3 177.766 163.664 55%
Round 4 203.455 191.523 62%
Round 5 220.015 184.199 50%

What if we just buy players that are cheap in terms of ADP relative to draft pedigree?

If this strategy were effective, we would expect that players with later ADPs yield an ADP improvement over their first year more often than players with an earlier ADP, at the same draft pedigree. This is somewhat intuitive, as expectations are lower for the players drafted later and thus have more room to exceed expectation and improve their perceived value. In this analysis, “Cheapness” will be measured by taking a player’s ADP and comparing it to the historical average ADP for players with the same pedigree.

If the hypothesis is true, one would expect that cheap players improve at a higher rate more often than expensive players. Below is a bar chart, each bar represents 10% of rookies in our sample. The furthest left Bar contains the 10% most “overvalued” rookies relative to pedigree. The second bar contains the 2nd most “overvalued” 10% and so on and so forth. The Y axis represents the percentage of rookies in each decile that exhibited an ADP improvement over that first year. We would expect a high rate of improvement to the right and a low rate of improvement to the left. We can clearly see there is no relation.

Decile chart

What does that mean?

Surprisingly, the conclusion is that the market is efficient to a certain extent. This means that the expectations that are built into a players ADP are correct ON AVERAGE. We can see this in the chart below, the player’s rookie ADP is a reflection of what the fantasy world expects from them and there is a correlation to actual points per game production.

PPG VS AOD

Obviously, we can see that there is a giant variation across the line of best fit, which means ADP only explains so much, but in general early ADP players score more than later ADP players.

 Conclusions – Where do we go from here?

Unfortunately, there is no naïve way of speculating on a rookie purely based off draft pedigree and ADP alone. Ideally, I was hoping to find that historically cheap first round WRs produce an “ADP return on investment”, or something of that nature, but that is not the case. There must be some wisdom in the masses; when the market puts a discount or premium on a player there is something to that. This is of course generally speaking, there is an equal chance of under performing or over performing across the early and late ADPs. We need to incorporate additional variables in order to make informed speculation. More to come on this.

Posted on Leave a comment

Analytics of Dynasty Bundle Sale!

With rookie draft season in full effect, I’m offering a unique bundle sale of Analytics of Dynasty content: You get The Analytics of Dynasty 2020 Edition and a month of Dynasty Patron content for $30.

The Dynasty Patrons get all the audio content, dynasty tiers, unique ADP data, and my rookie board.  Plus The Analytics of Dynasty 2020 Edition  will give you all the strategy you need for rookie and startup drafts.

To take advantage, go to patreon.com/analyticsofdynasty and become a Dynasty Patron for $10.  I will then send you a promo code to get The Analytics of Dynasty 2020 Edition for $20.