NGT: PFF

Most others use them for stats accumulation that can be verified. Subjectively, their opinions matter as much as many of ours

1. no one had heard of analytics in the NFL a few years ago, it's a recent thing.

2. PFF may not have been the first but it was one of the first.

3. Unlike other analytics companies, part of PFF was targeted directly to NFL teams. Being newer as I understood most NFL teams bought the PFF package which gave them more legitimacy. Many NFL teams has since starting their own in-house departments.

4. Cris Collinsworth personally invested in PFF and then used his network and media connections to raise its profile in the media.

I feel like they got so popular because their stats were very niche. "Player X is the most accurate when given over 5 seconds in the pocket and when Player Y did this and Player Z did that." It provides the opportunity to showcase a player's stats that are often overlooked even though they're oddly specific and usually don't hold much water.

In comment 14923426 Giantsfan79 said:

Quote:

1. no one had heard of analytics in the NFL a few years ago, it's a recent thing. 2. PFF may not have been the first but it was one of the first. 3. Unlike other analytics companies, part of PFF was targeted directly to NFL teams. Being newer as I understood most NFL teams bought the PFF package which gave them more legitimacy. Many NFL teams has since starting their own in-house departments. 4. Cris Collinsworth personally invested in PFF and then used his network and media connections to raise its profile in the media.

Interesting. How did they first establish their connections with the NFL?

conflates analytics with a lot of what they do. They don't offer much in analytics. They do their own rating system. That isn't analytics - that is subjective grading.

They perform two major services. Their quantitative compiling of snap and formation counts. This is what NFL teams use.

Their other service is their subjective ratings. Which is fodder for fans and media outlets, but holds little authority in meaning anything.

with the preface I'm no expert of their business model so by all means check yourself, but PFF hired a bunch of retired coaches to do their film breakdown. I know for certain Kevin Gilbride was one of their first big gets after he left the Giants. But that was more to build credibility with NFL teams not the general public.

My impression is PFF really took off with the media after Collinsworth became an investor.

.

the basic services provided with film breakdown to NFL teams, and their hook really imo is fantasy football enhancement.

RedZone, PFF, etc go towards FF and luring in new "fans" to eat up that shit.

Its film cuttup work. People really oversell PFfs importance to NFL teams. They don't understand its essentially intern work

helps to aid in that confusion when they market that they provide data to 80% of the NFL. Fans often confuse what they provide as being their ratings (of which no teams subscribe to).

In reality, it is their quantitative stats on snap counts and formations based off the film

It's why I call then snakeoil salesmen.

a ton of traffic volume to their site.

Given the sometimes sizable difference between their rankings and the eye test, along with some "catchy" names for some of their stats, they are a very smart business strategy.

...can write two separate articles on the same player, one "proving" that he's great and one "proving" that he sucks.

The just cherry pick data to back up their biases.

when they discredit them.

Joking aside, they're basically like any other metric of evaluation, there are biases baked in that you have to keep in mind. The grades are best used for things like offensive and defensive line play where the game is more straightforward and free of context.

There's a middle ground between PFF is useless and PFF is gospel. Personally, I've always had the opinion that their take mattered more than most because they are actually watching every snap in the NFL. Most posters here are only following the ball on a televised feed.

issue is that in that middle ground between useless and gospel, they are passing subjective analyses off as being grounded in data and facts.

They are doing more harm than good in the overall effort to bring "analytics" to game because what they do isn't implementing analytics, even though they want fans to believe they are to build credability.

PFF indexes and tags tons of data in a bunch of different dimensions. It’s a lot of work and is the starting point for analysis. Snap counts, formation counts, xyz scenarios. It’s the beginning point to understand if a scenario informs an outcome. Soon enough software will replace that labor and the analysis.

PFF has a bunch of interesting stats, that are good fodder for debate at both the amateur and professional level. The industry needs to be pushed to dispel some of its own silly assumptions, so pushing that conversation is a good thing for football. Strong outside voices are always good to kick a stagnant industry in the ass. That’s their primary long term contribution.

Trying to grade a player is both silly and boring. It assumes not only the player executed the assignment, but even more fundamentally it assumes the play and play design were correct for the situation (plus lots of other variables). Unless you know the likelihood and value of the outcome, grading players comes down to 1) did they appear to win the one-on-one match up or 2) compile meaningful stats in the aggregate. Those are fine dimensions to look at, but not a complete picture.

ok I guess. Just another piece of content to look at.

The one service I really like is Football Outsiders. I think their DVOA as an attempt at "objective power rankings" to be pretty indicative of teams' offensive/defensive/overall strength.

They also have a collection of stats that the typical sports site doesn't track. It's useful to look at their stats for certain teams in specific situations (i.e. points per drive, punts per drive, 3 and outs per drive, etc).

Used in conjunction with other sources and the eye test, helps gives you a good picture on teams.

out and their site was free I was a frequent visitor.

every week for whatever reason I wanted to know who played, how many snaps, run or pass play, etc.

I'd eagerly await every Tuesday morning for the Giants stats to become available.

then they made it a paid site, something like $19.99 for the season. I happily paid it for a couple years.

then for a short period of time once they became popular they revoked the individual subscriptions altogether. It was only an option for the media and/or teams and it was very expensive.

So F-them. I can get the snap counts now on pro-football-reference now or from a beat reporter who puts them in their stories.

They now I believe offer a consumer subscription again at affordable prices but like I said F-them.

I always found their ratings interesting - though polarizing, like others have said, an opinion, one data point out of many, but it was never why I frequented the site. I did always look at them though.

And it's definitely not snake oil. their stat data has been confirmed as 99+% accurate and their "PFF rating" has never been billed as analytics as far as I know, but they do believe it is statistically generated and without bias. Others disagree (with good reason some times).

You just bought some of it.

with their snap count data. They sell filmwork to the teams in the league, it is essentially the filmwork interns did in the past. It's down and distance stuff. formation tendencies things ike that.

Their rating nonsense has no credibility.

In comment 14924552 Sneakers O'toole said:

Quote:

You just bought some of it.

snake oil doesn't mean what you think it means.

They are a con. Bait and switch.

about the marketing practices?

What did they claim about their site that was exaggerated or even untrue?

What did they bait and switch?

Nothing. You can say their ratings are useless, but they don't claim them to be anything they're not and if that's what interests people about the site, great. They know what they're getting.

But again, the snap count data is 99+% verified accurate and that's why I paid for a $19.99 annual subscription for a couple years and the service met or exceeded my expectations. If they didn't remove it from being commercially available I'd still possibly be subscribing to them.

So, how exactly is it snake oil?

To give their horseshit legitimacy