Do Seahawks employ any form of analytics?

Mick063 · Jan 6, 2016

Just reading up on the "Moneyball" direction of the Cleveland Browns and their hire of the baseball advisor from the Mets.

I know that "Field Gulls" often cites some type of chart of measurables for players (metrics as position dependent) and I forget what they call it. But they claim Pete uses these statistics religiously when bringing in new talent.

Is this a form of "moneyball"? I'm not a baseball fan so I am not educated on the subject.

What, exactly, are the Browns trying to do? Seattle has been very successful with undrafted rookies. Are the Browns attempting to emulate that?

hawksincebirth · Jan 6, 2016

Sparq

hawksincebirth · Jan 6, 2016

From Google..." The SPARQ Rating is a scoring system designed to measure sport-specific athleticism. The results from various tests in each of the areas of speed, power, agility, reaction and quickness are combined and weighted using a sport specific formula."
I've seem several times that js and pc use this. I believe they we're one of the first NFL teams to use this.

TwistedHusky · Jan 6, 2016

They did.

Actually way back in the day, too.

Whitsitt believe it or not.

A friend of mine was building decision support platforms (back before it was called Big Data and nobody knew what Analytics was...) and the Seahawks were one of the first teams to buy in back then.

I remember distinctly they tried to sell the platform to the Mariners as a way to help them find draft picks that would stand a better shot at making the playoffs (the problem being that the Mariners did not really want to go to the playoffs but instead only wanted to go to the playoffs a minimum amount just to keep the fanbase interested, but that is another story). But the Seahawks caughtt wind of it, called shortly after and started running simulations using it.

I remember that we ran a # of QBs through as blind college profiles to see if we could identify key traits that would separate the Akili Smiths (high draft picks that did not turn out) from the Tom Brady types. One of the most significant elements was the # of times that a QB came back from a deficit in the 4th qtr against a top 25 team. I remember that guys like Tom Brady, Kurt Warner, Jim Kelly and Joe Montana would repeatedly end up in the top scores when we ran draft tests, and guys like Akili Smith and other draft busts would fall out.

I would be really shocked if they didn't have more substantive platforms they were running now. But I do remember it worked, hell we started negotiations for guys that wanted to use the platform for fantasy golf and fantasy Nascar (which I was not aware even existed at the time...) shortly thereafter. But if you have college kids using it to pick fantasy picks, then you certainly have an analytics engine that is weeding through scouting reports to see how measurables can be evaluated in context.

Hell the entire concept of SPARQ almost demands it. If you are going to score athletes based on measurables, then you have to apply some system to identify which attributes combined with those measurables are most likely to produce benefits. Even then, it still becomes a low floor/high ceiling approach (which is why it works so well, because most draft strategy starts with trying to get the highest floor possible - they run from risk instead of toward the greatest opportunity).

I believe that also explains why it feels that our early round picks do not pan out as often as they should. I think we don't follow the rule of trying to get picks to become immediate contributors, we try to get picks that have a much higher chance of becoming good to great players - but with a higher risk of failure. To accomplish that effectively in a way that is measurable & repeatable (as in multiple drafts), you almost have to be running some analytics engine.

Popeyejones · Jan 6, 2016

SPARQ scoring has been around since 2004. For the past seven or eight years it has mostly been a promotional tool used by Nike to market sports apparel across sports with branded athletes (how do you get a non-soccer fan to care about your soccer sponsor: assign the guy a SPARQ athletic score so he is comparable to the basketball or football players that person's a fan of).

With regards to analytics, yeah, it's an advanced analytic of sorts in that it's aggregating different measurements into a score, but it's a measurement of overall athletic ability rather than of actual performance in a competitive sport.

This isn't based on any inside knowledge, but knowing the sports analytics/moneyball community my guess is that they'd probably look more unfavorably on SPARQ as a meaningful measurement than most, as the subtext of most of what they're doing is to figure out the relationship between performance and wins to GET AWAY from things like passing the "eye test", and SPARQ is kind of like an "eye test" on steroids.

As for your question, back in Feb ESPN ranked all the major sports teams and their reliance on sports analytics. The Seahawks for them ranked as about average for their reliance on analytics in the NFL. FWIW that write-up never mentions their reliance on SPARQ, which at least weakly suggests that it's not a major selling point in the analytics community. (I can see its potential usefulness at the margins (e.g. once you get two S.D.s out and towards the back of the draft), but overall I think its best use might be its most frequent use: a marketing tool to sell shoes and such).

Write-up here: http://espn.go.com/espn/feature/story/_ ... s-rankings

Laloosh · Jan 6, 2016

The shit you know is spooky at times, popeye.

Year of The Hawk · Jan 6, 2016

TwistedHusky":2etgm40r said:
They did.

Actually way back in the day, too.

Whitsitt believe it or not.

A friend of mine was building decision support platforms (back before it was called Big Data and nobody knew what Analytics was...) and the Seahawks were one of the first teams to buy in back then.

I remember distinctly they tried to sell the platform to the Mariners as a way to help them find draft picks that would stand a better shot at making the playoffs (the problem being that the Mariners did not really want to go to the playoffs but instead only wanted to go to the playoffs a minimum amount just to keep the fanbase interested, but that is another story). But the Seahawks caughtt wind of it, called shortly after and started running simulations using it.

I remember that we ran a # of QBs through as blind college profiles to see if we could identify key traits that would separate the Akili Smiths (high draft picks that did not turn out) from the Tom Brady types. One of the most significant elements was the # of times that a QB came back from a deficit in the 4th qtr against a top 25 team. I remember that guys like Tom Brady, Kurt Warner, Jim Kelly and Joe Montana would repeatedly end up in the top scores when we ran draft tests, and guys like Akili Smith and other draft busts would fall out.

I would be really shocked if they didn't have more substantive platforms they were running now. But I do remember it worked, hell we started negotiations for guys that wanted to use the platform for fantasy golf and fantasy Nascar (which I was not aware even existed at the time...) shortly thereafter. But if you have college kids using it to pick fantasy picks, then you certainly have an analytics engine that is weeding through scouting reports to see how measurables can be evaluated in context.

Hell the entire concept of SPARQ almost demands it. If you are going to score athletes based on measurables, then you have to apply some system to identify which attributes combined with those measurables are most likely to produce benefits. Even then, it still becomes a low floor/high ceiling approach (which is why it works so well, because most draft strategy starts with trying to get the highest floor possible - they run from risk instead of toward the greatest opportunity).

I believe that also explains why it feels that our early round picks do not pan out as often as they should. I think we don't follow the rule of trying to get picks to become immediate contributors, we try to get picks that have a much higher chance of becoming good to great players - but with a higher risk of failure. To accomplish that effectively in a way that is measurable & repeatable (as in multiple drafts), you almost have to be running some analytics engine.

Thanks for that post. I agree with your angle on draft picks for us. Can be frustrating but also rewarding when it works.

Sealake80 · Jan 6, 2016

TwistedHusky,

Epic post. Thanks! :th2thumbs:

hawksincebirth · Jan 6, 2016

http://www.nfl.com/news/story/0ap20...e-seahawks-using-formula-to-find-future-stars

Per Field Gulls, Seattle's 2013 draft class boasted the NFL's highest collective SPARQ score. The early indications from 2014 calculations suggest a repeat for this year's class.

Among the Seahawks' late-round SPARQ stars are Super Bowl MVP Malcolm Smith, tight end Luke Willson, wide receiver Jermaine Kearse and special-teams dynamo Jeremy Lane.

There is more SPARQ-sanctioned help on the way.

The coaching staff has spent the offseason lavishing praise on "Making the Leap" candidate Christine Michael and third-year linebacker Korey Toomer, the two players with the highest SPARQ scores on the roster.

The fine folks at SB Nation's Field Gulls blog have detailed coach Pete Carroll's connection to Nike's SPARQ rating, a formula developed with the help of Seahawks strength and conditioning coach Chris Carlisle.

http://www.fieldgulls.com/seahawks-...e-of-athlete-the-seahawks-hunt-for/in/4125461

I disagree respectfully . The Seahawks use a modified version of sparq. Evident by the sparq freaks we continue to draft and sign via free agency

hawksincebirth · Jan 6, 2016

I'm sure they use other methods, but look at the players we draft and look at their sparq scores.

Jville · Jan 6, 2016

Mick063":2psoosn0 said:
Just reading up on the "Moneyball" direction of the Cleveland Browns and their hire of the baseball advisor from the Mets.

I know that "Field Gulls" often cites some type of chart of measurables for players (metrics as position dependent) and I forget what they call it. But they claim Pete uses these statistics religiously when bringing in new talent.

Is this a form of "moneyball"? I'm not a baseball fan so I am not educated on the subject.

What, exactly, are the Browns trying to do? Seattle has been very successful with undrafted rookies. Are the Browns attempting to emulate that?

The term "analytics" is a relatively recent buzz word. "Moneyball" surely popularized growth of the term "analytics". However, like so many buzz words, the quality, usefulness and purpose of what is published under the term "analytics" varies greatly.

The actual discipline that the terms draw their inspiration from is called "Operations Research". In western history, the modern origins of "Operations Research" predates World War II. However, the life and death struggle of World War II greatly accelerated interest in "Operations Research". Without it, the outcome of that war would have been different.

My earliest recollection of attempts to draw upon the discipline of "Operations Research" for NFL operations was pioneered by Tom Landry ..... a schooled industrial engineer and Dallas Cowboy head coach. Tom began with pencil and paper. Eventually, the Cowboy organization bought a computer and hired a programmer to automate Landrys' process and add to it. Concurrently, applications in that time era exploded with the growth of computer hardware and software tools.

Cornell Engineering currently has a concise and useful description of .... "What is Operations Research?" .... on their web site here >>> [urltargetblank]http://www.orie.cornell.edu/about/whatis.cfm[/urltargetblank]

Note: Cornell’s School of OR&IE has changed the “I” in their name, from “Industrial” to “Information,” to become the School of Operations Research and Information Engineering. (Schools thru out the world have made a similar name change.)

Popeyejones · Jan 7, 2016

Laloosh":3dlz86c1 said:
The shit you know is spooky at times, popeye.

:lol:

Thanks, but TBF you're giving me too much credit, man. I use posting to learn stuff also.

I just knew Nike either created or licensed SPARQ for marketing purposes and then googled from there b/c I was curious. The guess about SPARQ in the analytics community is seriously just a guess, and I remembered the ESPN analytics breakdown of all major sports teams b/c I thought it was pretty neat (and because Paraag Marathe of the 9ers is pretty involved in that community).

UK_Seahawk · Jan 7, 2016

Didn't they have some analytics set up initially with Nike but now the British guy in charge of it now works exclusively for the Seahawks? There was an interview with him on Seahawks draft blog.

Scottemojo · Jan 7, 2016

It was a weird draft this year, some Seattle bloggers were so in tune on the sparq stuff that they had called Glowinski, Sokoli, Frank Clark, and a couple of secondary players.

And when a UDFA signs, he usually has a good sparq rating. it is a metric that changes by position. Sokoli is an athlete at any position, but at center he would be a sparq god.

Analytics, on the other hand, is far more encompassing. Some might describe a breakdown of how a team passes or runs from specific formations or personnel groups analytics. Others might lump moneyball free agency thinking in that word.

IIRC, one of the reasons Tomsula fought with Paraig was over analytics.

I can tell you who didn't use analytics. Ruskell. Old expensive free agents, small and slow players who had done well for 4 years in college at big schools, he would have been the old guard Billy Beane had to get around in Moneyball.

NFL labor has a lot of rules that kind of fight the core principles of moneyball. Rookie contract scales, minimum cap floor, stuff like that. Not overpaying a linebacker because he is a linebacker might seem sound in baseball principles, but NFL positional blurring is way different than MLB. That linebacker might get sacks, or cover the middle on third down very well, and so a hard and fast positional approach isn't something I ever see being part of the NFL. But all corners need 32 inch arms? That kind of analytics is part of Seattle, and other teams, already.

UK_Seahawk · Jan 7, 2016

Is OP talking about analytics for drafting or measuring performance of his own players. I'm fairly sure we have schemes for both.

Popeyejones · Jan 7, 2016

Scottemojo":2g6jnibe said:
It was a weird draft this year, some Seattle bloggers were so in tune on the sparq stuff that they had called Glowinski, Sokoli, Frank Clark, and a couple of secondary players.

And when a UDFA signs, he usually has a good sparq rating. it is a metric that changes by position. Sokoli is an athlete at any position, but at center he would be a sparq god.

Analytics, on the other hand, is far more encompassing. Some might describe a breakdown of how a team passes or runs from specific formations or personnel groups analytics. Others might lump moneyball free agency thinking in that word.

IIRC, one of the reasons Tomsula fought with Paraig was over analytics.

I can tell you who didn't use analytics. Ruskell. Old expensive free agents, small and slow players who had done well for 4 years in college at big schools, he would have been the old guard Billy Beane had to get around in Moneyball.

NFL labor has a lot of rules that kind of fight the core principles of moneyball. Rookie contract scales, minimum cap floor, stuff like that. Not overpaying a linebacker because he is a linebacker might seem sound in baseball principles, but NFL positional blurring is way different than MLB. That linebacker might get sacks, or cover the middle on third down very well, and so a hard and fast positional approach isn't something I ever see being part of the NFL. But all corners need 32 inch arms? That kind of analytics is part of Seattle, and other teams, already.

Really good post, and just quibbling on two points:

1. Small & slow college players and Billy Beane: Hard to say on that one, as it really depends on which types of analytics are being leaned on. I think if we could port Bean into football today these might be the types of guys that he thinks are undervalued, as they don't pass the traditional scouting "eye test." I know he retired after a year but Chris Borland was a great example, as he fell due to being small and slow, and in his one season was one of the highest rated ILBs in the NFL.

I think guys like Borland, Russell Wilson, Antonio Brown, etc. may have very well been the types of players guys like Beane would have been targeting.

2. The song got played again this year, but the analytics "fight" rumor on the 9ers was supposedly between Harbaugh and Paraag as it related to in-game strategy. By the time Tomsula was up to bat this year Paraag was only involved in in-game analytics in name only, and didn't even attend those meetings.

seahawks08 · Jan 7, 2016

Analytics is really good to tell a story, there are multiple angles from what you infer from it. The data is that, but the futures is all a bit of luck involved. You can have all the data you want, but human intervention, experience and thinking with an open mind are critical for any space you are researching at. We had an architect who would use analytics and slice and dice only the parts that helped him tell the story rather than go with what was more right for the organization. It's great when you can use it the right way, but people have used analytics in ways that I can't imaging people do. Some use as fear mongering tactics, some use it as a high sales pitch, some to predict when people can fall sick or probability base on insurance costs. There are so many variations, if not done right, it is extremely and double edged sword. From a seahawks perspective, if they didn't have a very strong scouting personnel or group of people who look at everything, we would not have a strong team. Lots and lots of input is from Pete from the college days and the knowledge he had of players from high school on. I am sure he still talks to his connections to get a feel for each of the players they go after in the draft.

Do Seahawks employ any form of analytics?

Mick063

Well-known member

hawksincebirth

Active member

hawksincebirth

Active member

TwistedHusky

Well-known member

Popeyejones

Active member

Laloosh

New member

Year of The Hawk

Well-known member

Sealake80

New member

hawksincebirth

Active member

hawksincebirth

Active member

Jville

Well-known member

Popeyejones

Active member

UK_Seahawk

Well-known member

Scottemojo

Active member

UK_Seahawk

Well-known member

Popeyejones

Active member

seahawks08

Well-known member

Latest posts