The Dean Posted January 1, 2015 Posted January 1, 2015 we know these things: bill belichik read the berkley study on 4th down. he goes on 4th down more often than the bills. the bills paid a huge price for a receiver that appears to have several similar less costly peers. taking into account injury risks for receivers and general unpredictability it's hard to imagine a statistically based algorithm that favors the watkins trade. many other teams valued manuel significantly less than the bills. reportedly, many use analytics. i think the odds that the bills did the end run and the others followed the statistical analysis is much more likely than the converse. given more time, i can produce more examples. lets see your counter examples that point toward a successful analytical methodology from the bills. Well, you may be right on some of these accounts. But we really don't KNOW what other teams might have done had the Bills not picked EJ because it didn't happen. Don't rely on speculation. But the real fact is, there is no one approach to analytics. That is simply the case.
birdog1960 Posted January 1, 2015 Author Posted January 1, 2015 (edited) Well, you may be right on some of these accounts. But we really don't KNOW what other teams might have done had the Bills not picked EJ because it didn't happen. Don't rely on speculation. But the real fact is, there is no one approach to analytics. That is simply the case. agreed. but lets look at some premises for the watkins trade that might be incorporated into the algorithm: !.nfl players careers avg less than 3 years. 1a. the bills have a poor record of resigning star players after rookie contracts.2. a significant number of 1st round draft picks do not become stars (someone linked to a very subjective analysis of this but it was way below 50% - so lets throw 50 out there as a liberal estimate), injuries are extremely common, including career ending or changing injuries, it has recently been dramatically displayed that some nfl top players have outlier social and psychological profiles that threaten and sometimes end their careers. given all that, statistically it would appear to make sense to draft as many risky but top talented players (with the probable exception of the qb position given its importance and the rarity of exceptional players in this position) as possible to hedge ones bets. the bills did the opposite. the only reasonable conclusion was that this was a "gut" decision based on little more than personal opinion. it was one of the biggest decisions for the bills in recent years. it implies less than stong commitment to analytics in guiding decisions. Edited January 1, 2015 by birdog1960
The Dean Posted January 1, 2015 Posted January 1, 2015 agreed. but lets look at some premises for the watkins trade that might be incorporated into the algorithm: !.nfl players careers avg less than 3 years. 1a. the bills have a poor record of resigning star players after rookie contracts. 2. a significant number of 1st round draft picks do not become stars (someone linked to a very subjective analysis of this but it was way below 50% - so lets throw 50 out there as a liberal estimate), injuries are extremely common, including career ending or changing injuries, it has recently been dramatically displayed that some nfl top players have outlier social and psychological profiles that threaten and sometimes end their careers. given all that, statistically it would appear to make sense to draft as many risky but top talented players (with the probable exception of the qb position given its importance and the rarity of exceptional players in this position) as possible to hedge ones bets. the bills did the opposite. the only reasonable conclusion was that this was a "gut" decision based on little more than personal opinion. it was one of the biggest decisions for the bills in recent years. it implies less than stong commitment to analytics in guiding decisions. I don't know using averages in this case (especially how long one remains in the NFL) means a whole lot when it comes to a 1st round draft pick. Sure, some leave the NFL due to injury, but most leave because they just aren't good enough. That was never a concern with Sammy. But trying to look at football analytics from a baseball perspective will get you in a lot of trouble---again, IMO. There are too many moving parts, and variables in football. Much of what happens has to do with what another player, or players, does. Baseball is a bit more of an individual sport, when it comes to analytics.
birdog1960 Posted January 1, 2015 Author Posted January 1, 2015 (edited) I don't know using averages in this case (especially how long one remains in the NFL) means a whole lot when it comes to a 1st round draft pick. Sure, some leave the NFL due to injury, but most leave because they just aren't good enough. That was never a concern with Sammy. But trying to look at football analytics from a baseball perspective will get you in a lot of trouble---again, IMO. There are too many moving parts, and variables in football. Much of what happens has to do with what another player, or players, does. Baseball is a bit more of an individual sport, when it comes to analytics. and i would argue it's mostly the injuries that dictate the differences in career length between mlb and nfl. the differences at the top a talent are razor thin in almost every sport. but they're measurable and meaningful. and the true elites in baseball generally have very long careers while it's fairly common for an elite nil player to face early retirement/disability. given the choice again, i'll bet bo jackson avoids the nfl like the plague and plays mlb for many years. Edited January 1, 2015 by birdog1960
TSOL Posted January 1, 2015 Posted January 1, 2015 How about that end around they called for sammy last week? Duuh, think they used analytics to figure that if they run that play, once a year for 10 years, it will lead to a successful outcome 3-4 times?
Recommended Posts