2024 NFL Draft • Page 47

People do have nuance sometimes, but 90% of people will see 10 RAS and think “this guy’s more athletic than Bo Jackson” and, no. Sometimes not even close.

 

That's entirely their fault then lol. That does not mean every statistic ever presented ever needs you to argue against it. That's just annoying.

 

Football “analytics” in particular are notoriously irresponsible with numbers. That take of mine isn’t going to change. People literally have whole careers from using numbers in a very non data science way to convince people they’re data scientists and their analysis is infallible

 

The whole point of advanced analytics is to communicate stuff like this effectively and not have people drastically exaggerate (or minimize) what the measurements are saying

 

"analytics" = literal professional data scientists with degrees who work directly with the NFL. I'm not talking about "fantasy analysts" on twitter lol

Any system that isn't perfect is bad is a WILD take. And that's also not the "whole point of advanced analytics"

 

Working with the NFL’s stat collectors doesn’t make the science itself sound. Again, many sham studies out there, conducted by people with Ph.d’s about more important things than football, that take conceivably accurate data points but draw conclusions from bad methodology. If the formula’s wrong it doesn’t matter if the input points are right. And vice versa too, the formula might be getting false data if the stat itself is collected in a way that has margin for doubt

If people misappropriate what numbers can and can’t prove then the analytics are quite literally meaningless.

 

And people complained about me being sleepy too much, sheesh lol

 

Sure, but it means they are a LOT more knowledgable and experienced at this than you.

No....not at all.

 

Dude, RAS is very unscientific lol cause if it followed the actual bell curve of a standard distribution the 10.0 would be infinitely rare and there wouldn’t be 117 of them. “Experience” doesn’t mean much in a field where the bar is so low in terms of having all your ducks accounted for in your formulas. When people take the math at face value and never question, it you can write whatever formula you want and get whatever wonky results you want. All of a sudden Trevor Lawrence is bad and dragging his team down because his EPA is bad and not the other way around

Is it? Because we just had 50% of Americans justify not wearing masks during a pandemic with bad math, thinking you can just divide current cases by current deaths to figure out the death rate.

If it’s communicated poorly, does it really matter if the math people understand how to compensate for that, when the non math people are all just shrugging their shoulders and saying “well that number is clearly fully objective and formulated perfectly” cause they don’t want to think hard enough to challenge the context of the numbers?

It’s the data scientist’s job to not only have an accurate system that matches statistical distributions but to also communicate it in a way that people dont draw the wrong conclusions from it.

 

It’s a pretty informed opinion. Cause I actually know how data science works from an advanced statistics perspective. But have fun playing in the sandbox with numbers that are incredibly arbitrary

 

It's actually incredibly uninformed, arrogant, and annoying.

 

@EntryLevelTank

 

There's also a huge correlation between translating to the NFL and RAS but you would actually have to do research and not talk out of your ass to know that lmao

 

i would describe, actively avoiding math at every level of education you’ve been in, but then trying to make all your points with numbers to establish some intellectual high ground, exactly what you said

 

How do data nerds do dance offs? There needs to be something where like, whoever does a pivot table fastest wins and then you guys can just shut up lol

 

GIMME

 

Dave this didn't say what you think it says...

 

Yes exactly, I am not a math expert. Therefore I literally listen to people who know more than me and actually do this work lmao. That's how that works.

 

I would very much enjoy him “falling” from #2 into my lap at #3 please

 

Like you’re definitely better at organizing logic than me and you definitely put more effort into researching, but just wish you’d acknowledge that I know what I’m talking about when it comes to numeric relationships. I’ve been awesome at math my whole life, I’m not just bullshitting when I point out flaws in this new hype trend, of pretending stats can be objective about sports without linear regression analysis

 

I want that for you though.

 

Also Allbright is one of the guys that I assume knows nothing and is just talking lol

 

It’s 100 times out of 100 times better to use more simple measurements and draw your own conclusions from there, by cross examining those numbers with actual contextual in-game examples of whatever you’re looking for

than to try to criss cross a bunch of different numbers and try to come up with one composite tell-all statistic. Without artificial intelligence 99% of us (including me) are not actually advanced enough at math to do that.

 

I mean, I know you're bullshitting because I know you don't do research and don't put in 50+ hours a week in this field working directly with the NFL. You've literally tried to describe the methodology behind certain stats before and had to be corrected because you didn't know what you were talking about. You also don't "point out flaws", you reject every statistic every time its brought up and then turn it into an argument and literally nobody wants you to do that and it's incredibly annoying.