2023-2024 NFL Season Thread • Page 1851

Like, I know no one's ever gonna give me a pat on the back for this but I actually am good at being critical of stats and data.

It's a pet peeve of mine because poor math / problem formulation OFTEN leads to presumptive conclusions being drawn, like real research studies by think tanks get debunked all the time cause the math is improperly formulated, but people will still point to the numbers and say "its STATISTICS its 100% objective" without realizing 1) how subjective the metrics are for creating advanced statistics in the first place & how they are weighed, 2) AND how easy it is to draw the wrong conclusion because you misinterpreted what the math is actually saying.

I mean this literally happened during COVID and right wing pundits ran away with it, constantly referencing terrible methodology that was debunked by better scientists and mathemeticians as "proof" that it's a non-dangerous hoax.

so of course that happens BIG TIME in much less academic realms like sports - such as people thinking completeion percentage measures accuracy, or, such as people treating EPA and DVOA like the Bible when they spit out a lot of misleading results

 

Yes, I do acknowledge that. And some stats are more predictive than others. And I literally said he can improve and be an outlier. But having multiple areas where no one has ever become good in is definitely concerning. And matters more than something like completion percentage because that is often an area QBs improve in whereas, like you've constantly said about Fields, processing time and taking sacks is something that QBs don't often improve in.

 

big Tomlin fan.

lifetime extension

 

Love Giant Eagle

 

man, you know it's over when a fucking cookie cake is calling for your job

 

I didn't only bring up completion percentage though. I brought up 3 different stats that aren't shared by practically any other starting QB in the NFL. AR is definitely an outlier himself if he manages to have a full career as a starter.

 

Caleb certainly is concerning under pressure and has bust potential, though.

 

TD:INT ratio and wins aren't really predictive stats either. There are tons of college QBs with outstanding TD:INT ratios that are terrible NFL QBs. Wins aren't predictive either. Mahomes had one season without a losing record in college.

 

but vice versa doesn't exist, so a BAD TD:INT ratio actually is an indicator that that QB probably wont' be successful at the next level, even if a good TD:INT ratio doesn't help you determine their scucess

but isn't HE an extreme outlier here? What other successful QB didn't win a lot of games in college?

 

The criticism of Caleb that I’ve seen on my feed the most is how he plays, spending ample time in the pocket or freelancing to make awesome plays (which, don’t get me wrong, are awesome), but I do think it’s fair to question how replicable that style of play will be at the NFL level, especially if he’s behind a questionable offensive line and has a questionable receiving corps to throw to (see: Bears)

A lot can change, still plenty of season and free agency, but I think some of the criticism of Caleb is warranted or at least backed up

 

Let alone, the likelihood of stats falls dramatically the more stats you are combining together, the same way that your chance of winning a bet drops dramatically if you parlay it with other bets. This is objectively true in all math.

I'm not saying AR is only an outlier in comp percentage or only an outlier in losing more games than he won - he's an outlier in all three, which makes it that much more unlikely he'll beat the odds and be better than that in the NFL

 

like if option A has a 10% chance of happening, and option B has a 10% chance of happening

the chance of both happening is dramatically reduced to 1%

so the chance of a QB with low TD INT ratio AND low win rate AND low completion percentage being successful is compounded to be lower and is more predictive, than if we were just discussing one of those not-that-predictive stats

 

I keep bringing up things with ZERO percent hit rates. A QB with a bad TD:INT ratio is probably not going to be a good NFL QB. A QB with a bad play under pressure has NEVER been a good NFL QB. That's the difference.

but also since you ask - Brees, Ryan, Cutler, Stafford, Tannehill, and Brady. Josh Allen was close also as far as TD:INT rate goes.

Lamar didn't win a lot. Josh Allen didn't. Goff went 1-11 his first year and only won 5 games the next. A good QB can come from anywhere and sometimes that means they came from a mediocre team.

 

okay so by that logic you would agree in my assessment of taking four stats with 0% hit rates and combining them to be worried about Caleb? that's ignoring the fact that the stats you listed aren't predictive of NFL success

 

I think you should drop the word "predictive" because predictive analytics is far more layered and complex that what you're doing right now.

 

I'm not saying I'm doing predictive analytics I'm saying I'm looking at certain factors that are indicative of NFL success/failure. At a certain range, these stats become very predictive of how a player transitions into the NFL.

 

but QB talent and development isn't linear, and neither is the requirements of the QB position in the NFL. Players get better or regress, different teams ask QBs to take different levels of risk and execute varying difficulties of throws, supporting cast factors in, playcalling factors in, etc etc etc

trying to predict things from stationary stats just isn't how it works, and that's why its a lot more complex than you think it is. Every one of those stats is influenced by a lot of other variables that could change each time they change teams or enter a new season.

 

My boy AR is out here catching strays

- I agree with Randall that the completion % was a major red flag on the prospect and that “it’s correctable” <> “it will be corrected” and his ceiling is very limited if it doesn’t improve
- I agree with Elf that it’s correctable and there’s plenty of current examples of that happening and/or teams working around it until that skill develops; also agree that his pocket presence/processing ability has looked very positive in early sample
- no one has mentioned that Richardson is a literal superhuman on paper, while most QBs are not; to that point, it’s a better bet than the average QB prospect
- I’m actively offended by QB Winz being weighed as a serious concern against any prospect

 

yes, correct.

 

You've raised highly legitimate concerns but nothing about it is "predictive" and that's whats reading as pretentious to me. What you're doing isn't "predictive" either so that's not a basis from which you can refuse to address other statistics that, maybe aren't as granular, but still say a lot

 

AR had one of the lowest pressure to sack ratios of any prospect and had an above average turnover play percentage. On top of being one of the most physically gifted QB prospects ever, that VERY MUCH outweighs the low accuracy

 

It's not pretentious - please stop using that word lol. It is literally predictive at this range.

 

nerds