03-09-2020, 06:45 PM
They use the stat WAR (Wins Above Replacement) for the first four years of the players career so that is why they cut off at '16.
The math is way to complicated to go into, but basically they calculated a value for each pick (expected WAR) in the draft and then looked at the "actual WAR" of each specific player selected. I am not that familiar with WAR, but for some reason the WAR for QBs in naturally much higher for good QBs than any other position player. So they excluded QBs. That means we get no credit for drafting Dalton in the second round. Based on this formula the Bengals ranked 9th in the league in "WAR Over Expected".
Once they calculated "WAR Over Expected" they saw that this measurement was influenced too much by one great player found in later rounds. For example the Seahawks were ranked second over that ten year period, but if you remove just one player (5th round pick Richard Sherman) they would be middle of the pack. So they did a second calculation to determine "how often teams managed to find a player who ended up in a given percentile of his distribution of outcomes." I have no idea what that means but they say it would measure consistency and would diminish the impact of any one huge outlier. It also allowed them to include QBs in the formula. Based on this new formula for draft "consistency" the Bengals ranked 7th.
Then they decided that success in the early round was more important than success in later rounds. This makes since because it does not take a lot of production to exceed the projected WAR for a 7th round pick because the bar is set much lower than for 1st round picks. To account for this they did exactly what I would have done and built "a model that regresses the draft success as measured by WAR over expected against the consistent draft success in different rounds, and put the success of each team back into this regression."
After this final calculation the Bengals ranked 11th. That makes since when you look at their list of top 5 biggest hits "by percentile" in the AFC North. The Bengals have three of the five with 7th rounder Nedu Ndukwe, and 5th rounders George Iloka and Marvin Jones.
At #11 the Bengals drafted better than any team in the AFC North, and according to their calculations the Steelers drafted very poorly (19th). Browns were dead last in the league because they had more "draft capital" (high picks) than any other team but still did not find much talent.
They also included a few notes about the drafts since 2016. Auden Tate was in the top five biggest hits "by percentile", but in biggest misses in "WAR Over Expected" we had both number 2 (Price) and 3 (Ross). Bengals currently rank 30th in drafting over the last three years. Since the '15 draft was obviously one of the worst in the league that means we have to be near the bottom over the last 5 years despite having a decent draft in '16 (Jackson, Boyd, Vigil, Billings, Fejedelem).
We have to get back to drafting better if we are going to have any extended success.
The math is way to complicated to go into, but basically they calculated a value for each pick (expected WAR) in the draft and then looked at the "actual WAR" of each specific player selected. I am not that familiar with WAR, but for some reason the WAR for QBs in naturally much higher for good QBs than any other position player. So they excluded QBs. That means we get no credit for drafting Dalton in the second round. Based on this formula the Bengals ranked 9th in the league in "WAR Over Expected".
Once they calculated "WAR Over Expected" they saw that this measurement was influenced too much by one great player found in later rounds. For example the Seahawks were ranked second over that ten year period, but if you remove just one player (5th round pick Richard Sherman) they would be middle of the pack. So they did a second calculation to determine "how often teams managed to find a player who ended up in a given percentile of his distribution of outcomes." I have no idea what that means but they say it would measure consistency and would diminish the impact of any one huge outlier. It also allowed them to include QBs in the formula. Based on this new formula for draft "consistency" the Bengals ranked 7th.
Then they decided that success in the early round was more important than success in later rounds. This makes since because it does not take a lot of production to exceed the projected WAR for a 7th round pick because the bar is set much lower than for 1st round picks. To account for this they did exactly what I would have done and built "a model that regresses the draft success as measured by WAR over expected against the consistent draft success in different rounds, and put the success of each team back into this regression."
After this final calculation the Bengals ranked 11th. That makes since when you look at their list of top 5 biggest hits "by percentile" in the AFC North. The Bengals have three of the five with 7th rounder Nedu Ndukwe, and 5th rounders George Iloka and Marvin Jones.At #11 the Bengals drafted better than any team in the AFC North, and according to their calculations the Steelers drafted very poorly (19th). Browns were dead last in the league because they had more "draft capital" (high picks) than any other team but still did not find much talent.
They also included a few notes about the drafts since 2016. Auden Tate was in the top five biggest hits "by percentile", but in biggest misses in "WAR Over Expected" we had both number 2 (Price) and 3 (Ross). Bengals currently rank 30th in drafting over the last three years. Since the '15 draft was obviously one of the worst in the league that means we have to be near the bottom over the last 5 years despite having a decent draft in '16 (Jackson, Boyd, Vigil, Billings, Fejedelem).
We have to get back to drafting better if we are going to have any extended success.
