whaley's analytics dept

[Lurker]

I suspect "I don't know what the analytics department does" would've been a faster and less accusatory way to find out

Exactamundo...

[Repulsif]

if I remember well, the pick was called a gamble by the announcers. And although game metrics were good, he wasn't rated very high by draft metrics.

[birdog1960]

I suspect "I don't know what the analytics department does" would've been a faster and less accusatory way to find out

then let's define it. i believe my link does a pretty good job. the A's of Billy Bean and the harvard economist are the prototype. they are what next gen sports organizations aspire to. do you disagree?

 

Wow, this is revisionist history when Whaley was not the guy who drafted EJ and the analytics dept was created in October of last year. SMH

perhaps they should have had one much earlier

 

I think that there are a little more objective and easily obtainable statistics available for baseball - a game that is also much more individual match ups than football is.

 

I am not saying that analytics can't help in some way, but I do think that it can't stand alone without talent evaluation.

 

The condescending characterization you've made of "a roomful of dudes with average to to below average IQs evaluating talent on instinct" isn't called for, either. Without the talent evaluators, what is the analytics guy(s) going to base his analysis on?

it's an opinion like nearly everything else posted here. i don't see nix, whaley or marrone as elite football intellects much less above average overall intellects. Edited December 30, 2014 by birdog1960

[Hazed and Amuzed]

They probably go buy a bunch of draft preview magazines the week before the draft and see if anyone jumps off the page. Bunch of clowns.

 

Yeah they whiffed on guys like Kiko, Woods, Preston Brown, the Hughes trade, letting Byrd walk and Seantrell. What a bunch morons.

[NoSaint]

then let's define it. i believe my link does a pretty good job. the A's of Billy Bean and the harvard economist are the prototype. they are what next gen sports organizations aspire to. do you disagree?

 

perhaps they should have had one much earlier

 

it's an opinion like nearly everything else posted here. i don't see nix, whaley or marrone as elite football intellects much less above average overall intellects.

 

baseball is a whole different beast- so if you are expecting that, then yea you are probably off base. An at bat or a fielding opportunity are very individual plays to grade and a lot more reps for sample size. pick just about any play you've watched ever and show me how you'd breakdown the values ala the As do with an at bat (obviously it's a message board so I'm understanding you aren't going to be insanely detailed). And you are expecting them to have done it before the department existed.

 

In football Id venture the analytics hit much better for briad philosophies of game management or positional value than they do for individual predictions of success

Edited December 30, 2014 by NoSaint

[FLbills]

 

it's an opinion like nearly everything else posted here. i don't see nix, whaley or marrone as elite football intellects much less above average overall intellects.

 

Overall, I think Marrone, has done a good job. No, they didn't make it to the playoffs this year like we had all hoped. The tide is turning at OBD though and looking at the bigger picture, I like the foundation that Marrone has built. To be continued in 2015.... Go Bills

[birdog1960]

baseball is a whole different beast- so if you are expecting that, then yea you are probably off base. An at bat or a fielding opportunity are very individual plays to grade and a lot more reps for sample size. pick just about any play you've watched ever and show me how you'd breakdown the values ala the As do with an at bat (obviously it's a message board so I'm understanding you aren't going to be insanely detailed). And you are expecting them to have done it before the department existed.

 

In football Id venture the analytics hit much better for briad philosophies of game management or positional value than they do for individual predictions of success

i'm wondering what you think the bills envisioned when they started an "analytics dept". i'm betting the idea was parroting money ball whether it directly translates to football or not. i also think the bills have a history of going against conventional wisdom in drafting. that's fine if you're smarter than everyone else or have a system (e.g. analytics) but clearly the bills aren't and didn't with manuel.

 

my other point is that some non football minds (statisticians, behavioral scientists, social scientists etc) might add a great deal to drafting success. perhaps the bills are using such professionals extensively but if they are they aren't widely publicizing it.

 

finally, statistically speaking, it would appear wise to hire at least one high ranking administrator named "bill": as in polian, walsh or belichick

Edited December 30, 2014 by birdog1960

[bills_fan_in_raleigh]

Whaley described it yesterday in the pressers as he sometimes asks them to run some numbers?? WTF. Also Marrone claims he's the architect behind this yet he doesn't know that analytics would give u a high percentage of making a fourth and 1 ??? Its all proprietary they say yeah its proprietary cause its a bunch of bull **** if ya ask me

[OldTimer1960]

Whaley described it yesterday in the pressers as he sometimes asks them to run some numbers?? WTF. Also Marrone claims he's the architect behind this yet he doesn't know that analytics would give u a high percentage of making a fourth and 1 ??? Its all proprietary they say yeah its proprietary cause its a bunch of bull **** if ya ask me

 

A high percentage of making a 4th and 1? What is the league-wide average? What do you think the Bills' success rate would have been with a poor run-blocking OL and 2 backup RBs with no real threat of a big pass play to keep the defense honest?

 

You have to consider the hand that you are playing with when making decisions like this. The Bills team this year had an outstanding defense, very good special teams and an anemic offense. I completely understand why they wouldn't want to go for it on 4th and 1 in many situations - they were playing to their team strengths.

[Lurker]

i'm wondering what you think the bills envisioned when they started an "analytics dept".

As this paper from SAS points out, the use of analytics in football is significantly different from baseball.

 

http://www.sas.com/content/dam/SAS/en_us/doc/whitepaper2/iia-analytics-in-sports-106993.pdf

 

"In professional football, player and game performance analytics tend to be less sophisticated than in baseball or basketball. The complex interactions of a larger number of players on the field, and the difficulty in rating performance of players in each position on each play, make it more difficult to evaluate players or create optimal lineups."

 

Other intersting points:

 

 

"Another frontier data source for player and team performance is locational and biometric devices. These include GPS devices, radio frequency devices, accelerometers, and other types of biometric sensors... Some NFL (e.g., the Buffalo Bills) and NBA (e.g., San Antonio Spurs) use GPS devices in practices, which is the only time at which they are approved by their respective leagues. The locational devices are most frequently used to assess total activity (miles or kilometers run, steps taken, average speed) undertaken by players in a game or practice."

 

 

For several years, it’s likely that the analytics from such devices will be simple descriptive reports—what happened in numbers or charts, without any prediction or optimization. Over time, sophisticated teams will develop predictive and associational measures, such as these hypothetical examples:

 

• “This player tends to allow an average of three more points per game on defense when he runs more than 10,000 meters in a game.”

• “Players who heart rates exceed 160 beats/minute for over two consecutive minutes tend to run much slower in the last quarter of games.”

 

And this gem:

 

 

"The New England Patriots also had proprietary measures of player personality and attitude, and for several years selected players whom other teams had dismissed as having attitude problems. Such proprietary measures offer promise, but they are not without problems. At the Patriots, for example, the drafting of Aaron Hernandez, who is currently in jail on murder charges, may lead to less reliance on this proprietary data."

[OldTimer1960]

As this paper from SAS points out, the use of analytics in football is significantly different from baseball.

 

http://www.sas.com/c...orts-106993.pdf

 

"In professional football, player and game performance analytics tend to be less sophisticated than in baseball or basketball. The complex interactions of a larger number of players on the field, and the difficulty in rating performance of players in each position on each play, make it more difficult to evaluate players or create optimal lineups."

 

Other intersting points:

 

 

"Another frontier data source for player and team performance is locational and biometric devices. These include GPS devices, radio frequency devices, accelerometers, and other types of biometric sensors... Some NFL (e.g., the Buffalo Bills) and NBA (e.g., San Antonio Spurs) use GPS devices in practices, which is the only time at which they are approved by their respective leagues. The locational devices are most frequently used to assess total activity (miles or kilometers run, steps taken, average speed) undertaken by players in a game or practice."

 

 

For several years, it’s likely that the analytics from such devices will be simple descriptive reports—what happened in numbers or charts, without any prediction or optimization. Over time, sophisticated teams will develop predictive and associational measures, such as these hypothetical examples:

 

• “This player tends to allow an average of three more points per game on defense when he runs more than 10,000 meters in a game.”

• “Players who heart rates exceed 160 beats/minute for over two consecutive minutes tend to run much slower in the last quarter of games.”

 

And this gem:

 

 

"The New England Patriots also had proprietary measures of player personality and attitude, and for several years selected players whom other teams had dismissed as having attitude problems. Such proprietary measures offer promise, but they are not without problems. At the Patriots, for example, the drafting of Aaron Hernandez, who is currently in jail on murder charges, may lead to less reliance on this proprietary data."

 

Aww man, now you've taken the fun out of saying "the Bills buffoons just don't get it!"

 

This make perfect sense to me (and probably anyone who thinks about it for a bit). Another poster and I both pointed out that football was significantly different than baseball and were dismissed because we obviously don't get this new-fangled analytics that genius Mike Shoop babbles on endlessly about.

 

Thank you also for including the bit about the Bills using GPS tracking - OMG how could they be on the leading edge of something???

 

Further, thanks for the point about the Pats and their proprietary measures of player personality. While maybe not as sophisticated as the Patriots analysis, I am certain that teams have been extensively investigating players off-field behaviors and backgrounds all the way back into HS and before - nothing new there.

 

Now, if someone can devise a test that can find the one Tom Brady out of the 1,000 similar athletes playing QB in college, that would be worth a bundle!

Edited December 30, 2014 by OldTimer1960

[Canadian Bills Fan]

1st rule about Analytics department is that you dont talk about analytics department

 

 

CBF

[birdog1960]

A high percentage of making a 4th and 1? What is the league-wide average? What do you think the Bills' success rate would have been with a poor run-blocking OL and 2 backup RBs with no real threat of a big pass play to keep the defense honest?

 

You have to consider the hand that you are playing with when making decisions like this. The Bills team this year had an outstanding defense, very good special teams and an anemic offense. I completely understand why they wouldn't want to go for it on 4th and 1 in many situations - they were playing to their team strengths.

the 4th down dilemma is one of the obvious places to apply statistics. it so happens that a berkley economist did just that and his conclusions don't fit with conventional nfl practice: http://static.espn.g...eg/1453717.html. an algorithm could likely be developed that accounts for differences in specific teams personnel. it's seems an obvious place to get an edge over the majority of the league with minimal investment or effort. interestingly, some of the most revered names in coaching agree with his basic premise including another economics major from weslyan who happens to coach in the boston area. perhaps, the bills should be looking for an economist for their analytics dept.

 

showing a few failures of analytics is hardly, well, an analytical approach to assessing it's value. as in any field, there are good, bad and indifferent actors involved. the real potential of the approach can only be judged by looking at the best in the field.

Edited December 31, 2014 by birdog1960

[OldTimer1960]

the 4th down dilemma is one of the obvious places to apply statistics. it so happens that a berkley economist did just that and his conclusions don't fit with conventional nfl practice: http://static.espn.g...eg/1453717.html. an algorithm could likely be developed that accounts for differences in specific teams personnel. it's seems an obvious place to get an edge over the majority of the league with minimal investment or effort. interestingly, some of the most revered names in coaching agree with his basic premise including another economics major from weslyan who happens to coach in the boston area. perhaps, the bills should be looking for an economist for their analytics dept.

 

showing a few failures of analytics is hardly, well, an analytical approach to assessing it's value. as in any field, there are good, bad and indifferent actors involved. the real potential of the approach can only be judged by looking at the best in the field.

 

Woah there cowboy. Nobody gave a single example to impune analytics. What are you talking about?

 

If you read my posts in this thread, you will see that I agree that analytics can complement solid football knowledge and scouting. I am only arguing that analytics, in the absence of solid football knowledge and scouting is blind - just as it is in any other application.

 

Analytics without any application-specific knowledge is just a bunch of numbers without context. Yes, if you crunch enough numbers (with application-specific knowledge), you may identify something unique. But, I will argue, that you can crunch data for eternity and learn nothing if 1) the data is unfounded because it is not perfectly objective (football scouting nd coaching isn't ) and 2) you don't know the context of what you are analyzing is.

 

As for yorur grudging acceptance that maybe you ought to consider specific situations over general analytics trends, well yeah. If you have a great defense and a sh*tty offense, does it not make sense to rely on your defense over your offense?

 

I AM NOT saying that if you have a competent offense that can gain 2 yards no matter down and distance that it would be a bad idea to risk going for it on 4th and 1, but to make that broad-based assertion as a fact regardless of your team's strengths and weaknesses is silly - regardless of how many economists disagree.

 

 

Just my $.02 worth, but then I am sure the economists don't think that is relevant.

 

 

 

 

[BringBackFergy]

1st rule about Analytics department is that you dont talk about analytics department

 

 

CBF

LOL

[Dibs]

........the analytics dept was created in October of last year. SMH

 

Yeah, but Nix/Whaley should have hired some guys to predict what the future analytics department was going to predict about the QBs in 2013.

[thewildrabbit]

Just a shame that in all their "consensus" they didn't have one decent offensive mind among them.

[birdog1960]

Woah there cowboy. Nobody gave a single example to impune analytics. What are you talking about?

 

If you read my posts in this thread, you will see that I agree that analytics can complement solid football knowledge and scouting. I am only arguing that analytics, in the absence of solid football knowledge and scouting is blind - just as it is in any other application.

 

Analytics without any application-specific knowledge is just a bunch of numbers without context. Yes, if you crunch enough numbers (with application-specific knowledge), you may identify something unique. But, I will argue, that you can crunch data for eternity and learn nothing if 1) the data is unfounded because it is not perfectly objective (football scouting nd coaching isn't ) and 2) you don't know the context of what you are analyzing is.

 

As for yorur grudging acceptance that maybe you ought to consider specific situations over general analytics trends, well yeah. If you have a great defense and a sh*tty offense, does it not make sense to rely on your defense over your offense?

 

I AM NOT saying that if you have a competent offense that can gain 2 yards no matter down and distance that it would be a bad idea to risk going for it on 4th and 1, but to make that broad-based assertion as a fact regardless of your team's strengths and weaknesses is silly - regardless of how many economists disagree.

 

 

Just my $.02 worth, but then I am sure the economists don't think that is relevant.

gotcha, chief…if you read my posts you'd understand that i'm questioning the bills commitment to analytics. the decisions from the fo most often appear empirical including the manuel pick. i'm suggesting the analytics dept in regards to the bills might well represent more a brandon marketing move than a substantive philosophical change in management. Edited December 31, 2014 by birdog1960

[Manther]

the decision to draft ej described here sounds a lot like a bunch of guys shooting the bull: "I always say it's a Buffalo Bills pick," Bills general manager Doug Whaley said. "You guys can assign blame wherever you want, but when we're in draft meetings it is a consensus recommendation for our whole organization of where this guy's valued and where we should pick him in the draft."

 

so do you need an "analytics dept" to do this or can 3 or 4 guys give their opinion and then majority rules? somehow i see the bills doing it this way rather than an IT guru handing out carefully manicured and harvested probabilities of success in different aspects of performance metrics integral for a desired position players likelihood of mediocrity or domination.

 

anyone with direct insight into how i reasllyt happens beyond whaley's comments here?

The analytics department was Brandon. And Marrone was the one who said he can use it better than anyone. How is this bashing Whaley?

[birdog1960]

The analytics department was Brandon. And Marrone was the one who said he can use it better than anyone. How is this bashing Whaley?

if you believe analysis is superior to empirical decision making then whaley's consensus approach is undesirable and less likely to be successful.