Call For Research: Ted and Analytics

[pbmax]

OK, Maxie reposted an item yesterday in red's panic thread about the 2017 Draft.

It was about the SPARQ evaluation system and how it helps team identify useful physical characteristics from combine numbers. This was in partial response to my post of an article from The Ringer about how SPARQ and its descendants are making team very conscious of explosive athletic traits among its potential draftees.

Its all very much in a vein of analytics applied to player personnel. In this case, the draft.

There are two Packer connections: first is John Dorsey saying his Chiefs are 15% Oakland Athletics (research) and 85% Atlanta Braves (old school scouting). this was in the Ringer article.

Second is that within the last year or two, Ted said in a press conference he was hiring or talking to a couple people about analytics as it applied to player personnel. He may have said they have hired people or that he was talking to consultants.

Short version, there are a lot of Packer draftees in last two drafts who look good in the SPARQ system and I think you can make a case the Packers have been getting faster. (Best example for this might be Martinez as your ILB rather than Reggie Ragland - see Maxie's post).

So I need someone to find that Thompson PC where he mentions analytics. I think its on the board here; I know he said it but I neither remember when or where. He specifically said it was on his to do list and that he was talking to two (or a couple of) people to get up to speed.

The reason we need it is that a writer has expressed interest in pursuing this for an article after I pestered them on Twitter.

[pbmax]

February 23rd, 2015


The Packers rarely use analytics on the personnel side, as general manager Ted Thompson admitted in a recent interview with the Milwaukee Journal-Sentinel. Thompson, an old-school GM, said, "I'm not saying I believe in [analytics], but I'm not going to shut my ears to it. ... We are beginning the process of dabbling in analytics."

http://www.espn.com/espn/feature/story/_/id/12331388/the-great-analytics-rankings#nfl-gnb

[pbmax]

Found it!

http://archive.jsonline.com/sports/packers/packers-exploring-benefits-of-analytics-in-player-personnel-scouting-b99331427z1-271478841.html

[Fritz]

It's also known as "analitics" in some quarters.

[Maxie the Taxi]

A lot of good stuff here on SPARQ for the NFL on this FAQ: https://3sigmaathlete.com/faq/

An example:

There are a vast number of players who have tested out well but failed to make an impact in the NFL. The issue here is hit rate. While Wes Welker had a successful career as a marginal athletic tester, there aren’t very many Wes Welkers around. It’s hard to identify that player, even for the best of scouts. I prefer to take my chances with the better athlete, and the data bears this out: good players also tend to be good athletes.

[texaspackerbacker]

So Analitics gets you Damarious Randall, and scouting gets you Ladarius Gunter? and Geronimo Allison?

[Maxie the Taxi]

So Analitics gets you Damarious Randall, and scouting gets you Ladarius Gunter? and Geronimo Allison?Yes. But there are 100 times more Randall's than either Gunter or Allison. If you're looking for a classic SPARQ failure, it's not Randall. It's Jeff Janis. Nobody's arguing a pure athlete without natural football instincts is a sure bet to make it in the NFL. We're saying it's the best way to cull through the thousands of players in college and find the players worth pursuing by scouts.:razz:

[texaspackerbacker]

Yes. But there are 100 times more Randall's than either Gunter or Allison. If you're looking for a classic SPARQ failure, it's not Randall. It's Jeff Janis. Nobody's arguing a pure athlete without natural football instincts is a sure bet to make it in the NFL. We're saying it's the best way to cull through the thousands of players in college and find the players worth pursuing by scouts.:razz:

Agreed, however, I'd argue he's a product of scouting too. And unlike Randall, who I tend to be skeptical is ever gonna amount to more than a journeyman, I still think one of these years, something will click for Janis, and he will be a star in this league - hopefully with the Packers.

[Maxie the Taxi]

It will never "click" for Janis. Why? He's got an ugly girlfriend. hahahahahahahahaha


https://www.youtube.com/watch?v=pWgyy_rlmag

[Patler]

Ron Wolf is not a fan. This is from an April, 2016 article:


Former Packers general manager Ron Wolf sees little value in analytics as a rule. In retirement, Wolf spends his winters in Florida and often frequents the St. Louis Cardinals' spring training facility in Jupiter.

There he spends many hours commiserating with former Cardinals manager Tony LaRussa and general manager Walt Jocketty. One can only imagine the discussions the three have about scouting talent, making draft decisions and running a franchise.

None subscribes to the "Moneyball" philosophy.

Wolf said the closest he came to such a thing was an experiment Oakland used incorporating physical testing numbers.

"When I was with the Raiders in the '80s, we tried it with height-weight-and-speed guys, guys that tested very well at each position," Wolf said. "We tried to draft those guys, the high-testers. And not one of them made it. Not one. Seriously.

"I don't know enough about analytics in football. I don't know how you can incorporate that. As you well know, you can take a statistic and use it any way you want. You can twist that anytime you want to do that."

http://archive.jsonline.com/sports/packers/baseball-style-analytics-becoming-a-part-of-nfl-decision-making-b99711074z1-376816711.html


I wonder how the Wolf puppy feels about it?

[texaspackerbacker]

It will never "click" for Janis. Why? He's got an ugly girlfriend. hahahahahahahahaha


You know this how?

[Maxie the Taxi]

You know this how?I was joking.

Did you watch the youtube? Or the movie? It's great.

[Patler]

It will never "click" for Janis. Why? He's got an ugly girlfriend. hahahahahahahahaha


Does his wife know?
He better hope not. Mrs. Janis:

https://www.instagram.com/p/BL9jtVkjdlv/

(There is supposed tb be a picture of her here with the buck she shot.)

https://www.instagram.com/p/BL9jtVkjdlv/

[Maxie the Taxi]

Ron Wolf is not a fan. This is from an April, 2016 article:



I wonder how the Wolf puppy feels about it?Ron Wolf's quote is typical in-the-box, old school thinking. I imagine his conversations with Tony La Russa in St. Louis were very much like the conversation around the table in the Youtube clip I posted. lol

[pbmax]

There is no level to which old school types won't sink to torpedo the notion that a new idea might work. I have no doubt the height-weight-speed guys were busts at a higher rate for Wolf. It sounds entirely like they looked at it in the abstract, like the Raiders always overdraft speedy wideouts (Darius Hayward Bey).

However, Wolf had height and weight restrictions for several positions. He also advised Schneider against taking Russell Wilson based on his height. So Ronnie was tied to the same kind of thinking that he claimed proved to be unworkable.

LaRussa loved to bang on about analytics, but the hometown team that Ted mentions being well into analytics was the Houston Astros. The Astros a short time ago were fined and had to fire a high level scout who was a key player in their move to data based analysis. He stole information to help build the Astros analytics database.

The team he stole information from, the team with the most far reaching database in baseball?

The St. Louis Cardinals.

LaRussa protested too much. He and Jocketty wanted the credit not to be shared and so they denied their front office used the stuff. When Jocketty resisted its continued use, he was let go. LaRussa followed soon after.

[Maxie the Taxi]

^^^Don't they call it "confirmation bias" or something?

[pbmax]

Yes. But there are 100 times more Randall's than either Gunter or Allison. If you're looking for a classic SPARQ failure, it's not Randall. It's Jeff Janis. Nobody's arguing a pure athlete without natural football instincts is a sure bet to make it in the NFL. We're saying it's the best way to cull through the thousands of players in college and find the players worth pursuing by scouts.:razz:


Agreed, however, I'd argue he's a product of scouting too. And unlike Randall, who I tend to be skeptical is ever gonna amount to more than a journeyman, I still think one of these years, something will click for Janis, and he will be a star in this league - hopefully with the Packers.

Data based analytics are not a replacement for scouting. No one hires an analytics person then turns the scouting offices into a rec area. They help anchor scouting with measurements and grades that can cross college conferences and opponents, and even time.

They let you remember more lessons from the past, not fewer. They also help to bust myths. There probably is a huge advantage to being a tall QB. But there are levels of performance that make the limit worth ignoring. Analytics can help you figure out where that is with less subjectivity.

But mostly they help you at the margins, later in the draft, to find the overlooked talent that is still available because the database doesn't forget who got what grades. And the database also incorporates the grades the scouts gave the players.

I doubt Randall was simply an analytics find. I bet it took scouts who thought his skills translated to CB to make the case that he was worth a first rounder regardless of his measurables. Ted and Ron took flyers on great athletes long before data became a thing. Remember Ted said he likes his guys to remind him that the idea is not just to have measurables, its to take good football players. The data just helps you sort the tradeoff you are making.

[vince]

Exactly.

Another important thing about analytics is that it's an evolutionary process. The world is complex and simple formulas incorporating a few data points will never accurately reflect reality.

As actual results are measured against what the analytic models used have predicted, the models must be updated and additional data points may be added to continuously improve their predictive value.

In other words, if your initial models don't work Ron, you need to update them and likely enhance their sophistication through the incorporation of additional data and performance metrics to make them increasingly reliable.

It's not analytics that don't work, it's the people who don't know how to use them to enhance (not replace) their existing experience and judgment for better decision-making.

[texaspackerbacker]

I was joking.

Did you watch the youtube? Or the movie? It's great.

I thought you might be hahahaha. I just watched the little clip you posted.

[Maxie the Taxi]

I can almost guarantee you that the Randall and Rollins picks were the result of scout "reasoning" rather than analytics.

Randall wasn't even ranked by pSPARQ analytics as a CB, but as a FS. As a free safety he ranked #11 among prospects, just four spots better than Jermaine Whitehead who is currently on our Practice Squad.

Rollins was ranked #67 among all prospects tested at CB.

If you rank Randall's pSPARQ score among the CB's tested, it turns out he would have been #23 among CB's, certainly not 1st round draft material.

[texaspackerbacker]

Does his wife know?
He better hope not. Mrs. Janis:

https://www.instagram.com/p/BL9jtVkjdlv/

(There is supposed tb be a picture of her here with the buck she shot.)

https://www.instagram.com/p/BL9jtVkjdlv/

Nice! (the girl, but the buck too) ........ amazing how you can find just about anything on the internet these days.

[texaspackerbacker]

I can almost guarantee you that the Randall and Rollins picks were the result of scout "reasoning" rather than analytics.

Randall wasn't even ranked by pSPARQ analytics as a CB, but as a FS. As a free safety he ranked #11 among prospects, just four spots better than Jermaine Whitehead who is currently on our Practice Squad.

Rollins was ranked #67 among all prospects tested at CB.

If you rank Randall's pSPARQ score among the CB's tested, it turns out he would have been #23 among CB's, certainly not 1st round draft material.

Seriously? I'd sure like to hear the Ted "reasoning" for those picks.

If I was a GM, I'd go with college on field performance ahead of both scout "reasoning" and analytics - for everything except maybe RBs that is.

[Maxie the Taxi]

Moreover, if TT had relied on pSPARQ analytics in 2015 he would have most likely drafted Ronald Darby with his first round choice, a choice Tex might have liked a whole lot more than Randall.

[Maxie the Taxi]

Seriously? I'd sure like to hear the Ted "reasoning" for those picks.

If I was a GM, I'd go with college on field performance ahead of both scout "reasoning" and analytics - for everything except maybe RBs that is.Tex, scout reasoning IS based on college on field performance, but there is a HUGE amount of subjectivity involved, comparing performance from program to program, level to level and conference to conference for instance.

[vince]

Seriously? I'd sure like to hear the Ted "reasoning" for those picks.

If I was a GM, I'd go with college on field performance ahead of both scout "reasoning" and analytics - for everything except maybe RBs that is.
All aspects and dynamics of college on-field performance as well as scout "reasoning" can be and are incorporated into performance models. The magic of automation and the vast availability of data enable the development of algorithms that measure complex relationships and weigh hundreds of variables together.

Character ratings, quality of college competition, scheme contribution considerations, and much, much more are incorporated.

Players wear body telemetry instruments that track all kinds of physical activity and stress measurements, creating mounds of data that are being incorporated and corollated with strength training data, nutrition data, medical histories, as well as a variety of additional performance metrics to offer insight into how practice schedules, nutrition plans, strength training goals, and ultimately on-field performance may be able to be optimized.