SSOG
Moderator
Without getting too much into subscriber content, I just read Tefertiller's "drafting rookies" article and felt that the rookies needed a champion to defend them a little bit.
The quote that caught my attention was the following: "So many rookies are drafted with high expectations. The position with highest relative rookie ADP is at running back. Over these eight seasons of the study, twenty-five rookie ball carriers were drafted as fantasy starters. Their collective ADP was RB28, but they finished with a net ranking of RB37."
The problem I have with that quote is that it gives no context whatsoever. How does that compare to 2nd year players? 3rd year players? 4th year players? There are two big reasons to believe that the phenomenon you found had nothing to do with rookies being a "losing proposition" and everything to do with how you framed your study. For starters, you only selected the most highly drafted rookies. The higher a player is drafted, the more likely he is to underperform his draft slot through no fault of his own. For instance, I could do a study to see how many #1 overall draft picks outperformed their draft position, and I'd guarantee you that it'd tell you that no #1 has ever in the history of football outperformed his draft pick. The number of #2 guys who outperform their draft pick is going to be astronomically small, and so on down the line. Any time you deliberately take a sample from the top of any population, that population is almost certainly going to wind up underperforming projections. Just look at Adrian Peterson and MJD last year. They were drafted 1st and 2nd. They finished 2nd and 3rd. Both of them technically "underperformed" their draft slot... but I bet most owners would take "underperformance" like that 8 days a week.
Second off, in this instance you shouldn't be using mean finishes, because the population is going to exhibit a strong skew. If a player is taken as the 11th WR, then the most he could outperform his draft position is by 10 slots. He could, however, underperform his draft position by 90+ slots. If you had two rookies taken 11th and 12th, and one of them finished as WR1 and another finished as WR100, using the mean would tell us that the rookies were taken 11.5th on average, and they finished 50.5th on average, so drafting rookie WRs was a losing proposition. That analysis does not accurately describe the reality of the situation.
To illustrate what I mean, I took the names of every 3rd, 4th, or 5th year RB drafted 36th or higher at his position last year. These guys are all in the prime of their career, and they're all known quantities. All told, there were 15 such RBs last year. They were drafted, on average, as RB17. They finished, on average, as RB24. Therefore, drafting 3rd-5th year RBs is a losing proposition, too.
According to the data from your study, 13 of the 25 rookies taken in the top 36 wound up outperforming their draft position (52% success). For comparison, just 4 of the 15 3rd-5th year RBs taken in the top 36 last year wound up outperforming their draft position (27% success). By that comparison, drafting RBs PERIOD is a losing proposition... but drafting rookie RBs is a WINNING proposition. Or- to paraphrase Doug Drinin's favorite paraphrasing of Winston Churchill- drafting rookies is the worst way to find value, except for all the other ways.
One last stat before I go. From 2004 to 2008 (5 years), 13 RBs were drafted outside of the top 25 who then wound up finishing inside the top 12. Six of them (almost half) were rookies. Once the top 25 options are off the board, rookies often represent the best possible chance of landing a surprise RB1.
Not trying to pick on you, because I think you do phenomenal work, I just had to post to say that I disagreed with the conclusions of this particular article and to explain why. In my opinion, drafting rookie RBs is a WINNING proposition. I do agree about rookie QBs and TEs, though- they're rubbish and are best avoided.
The quote that caught my attention was the following: "So many rookies are drafted with high expectations. The position with highest relative rookie ADP is at running back. Over these eight seasons of the study, twenty-five rookie ball carriers were drafted as fantasy starters. Their collective ADP was RB28, but they finished with a net ranking of RB37."
The problem I have with that quote is that it gives no context whatsoever. How does that compare to 2nd year players? 3rd year players? 4th year players? There are two big reasons to believe that the phenomenon you found had nothing to do with rookies being a "losing proposition" and everything to do with how you framed your study. For starters, you only selected the most highly drafted rookies. The higher a player is drafted, the more likely he is to underperform his draft slot through no fault of his own. For instance, I could do a study to see how many #1 overall draft picks outperformed their draft position, and I'd guarantee you that it'd tell you that no #1 has ever in the history of football outperformed his draft pick. The number of #2 guys who outperform their draft pick is going to be astronomically small, and so on down the line. Any time you deliberately take a sample from the top of any population, that population is almost certainly going to wind up underperforming projections. Just look at Adrian Peterson and MJD last year. They were drafted 1st and 2nd. They finished 2nd and 3rd. Both of them technically "underperformed" their draft slot... but I bet most owners would take "underperformance" like that 8 days a week.
Second off, in this instance you shouldn't be using mean finishes, because the population is going to exhibit a strong skew. If a player is taken as the 11th WR, then the most he could outperform his draft position is by 10 slots. He could, however, underperform his draft position by 90+ slots. If you had two rookies taken 11th and 12th, and one of them finished as WR1 and another finished as WR100, using the mean would tell us that the rookies were taken 11.5th on average, and they finished 50.5th on average, so drafting rookie WRs was a losing proposition. That analysis does not accurately describe the reality of the situation.
To illustrate what I mean, I took the names of every 3rd, 4th, or 5th year RB drafted 36th or higher at his position last year. These guys are all in the prime of their career, and they're all known quantities. All told, there were 15 such RBs last year. They were drafted, on average, as RB17. They finished, on average, as RB24. Therefore, drafting 3rd-5th year RBs is a losing proposition, too.
According to the data from your study, 13 of the 25 rookies taken in the top 36 wound up outperforming their draft position (52% success). For comparison, just 4 of the 15 3rd-5th year RBs taken in the top 36 last year wound up outperforming their draft position (27% success). By that comparison, drafting RBs PERIOD is a losing proposition... but drafting rookie RBs is a WINNING proposition. Or- to paraphrase Doug Drinin's favorite paraphrasing of Winston Churchill- drafting rookies is the worst way to find value, except for all the other ways.
One last stat before I go. From 2004 to 2008 (5 years), 13 RBs were drafted outside of the top 25 who then wound up finishing inside the top 12. Six of them (almost half) were rookies. Once the top 25 options are off the board, rookies often represent the best possible chance of landing a surprise RB1.
Not trying to pick on you, because I think you do phenomenal work, I just had to post to say that I disagreed with the conclusions of this particular article and to explain why. In my opinion, drafting rookie RBs is a WINNING proposition. I do agree about rookie QBs and TEs, though- they're rubbish and are best avoided.