|
Nodoze posted:Raiders schedule honestly does not look bad. IDK why you would have any faith in the Chargers to not shoot themselves in the dick like they did last year.
|
#
?
Dec 8, 2022 21:09
|
|
|
|
| # ? Jun 7, 2024 19:55 |
|
|
It's funny the chargers could end up 6-11, 11-6, 6-6-5 or anything in between and I would not be shocked.
|
|
#
?
Dec 8, 2022 21:09
|
|
|
I think if the Jets can get to 10-7, they're in for sure. Any less and I'd probably say the Chargers have the easiest remaining but they're also a game back and not really prone to playing well for extended stretches. The Pats have decent chance of getting in on 9-8 tiebreakers if it goes there.
|
|
#
?
Dec 8, 2022 21:11
|
|
|
The funny thing is the Ravens are the team that'd be in danger of falling out with Lamar's injury but their schedule is: @Steelers, @Browns, Falcons, Steelers, @Bengals. Shouldn't have any issues hitting 10 wins even with me at QB.
|
|
#
?
Dec 8, 2022 21:12
|
|
|
Diva Cupcake posted:I think if the Jets can get to 10-7, they're in for sure. If the Pats lose to the Cards there is a real good chance they just lose out, the players seem really close to just checking out. mcmagic posted:IDK why you would have any faith in the Chargers to not shoot themselves in the dick like they did last year. I definitely don't. They cannot get out of their own way
|
|
#
?
Dec 8, 2022 21:13
|
|
|
In 2008 the Niners lost all their quarterbacks (Alex, Trent Dilfer, and Shaun Hill) to injuries and had to grab Chris Weinke off the street to start the last game of the season. It went about as well as you'd expect.
|
|
#
?
Dec 8, 2022 21:14
|
|
|
The Raiders are probably playing the best of those 4 teams right now lol
|
|
#
?
Dec 8, 2022 21:16
|
|
|
https://twitter.com/tejfbanalytics/status/1600590086550261764
|
|
#
?
Dec 8, 2022 21:17
|
|
|
The Raiders would already be in the playoffs in the could hold onto a 17 point lead
|
|
#
?
Dec 8, 2022 21:20
|
|
|
mcmagic posted:The Raiders are probably playing the best of those 4 teams right now lol Which is funny because they've recently lost to Jeff Saturday and needed OT against a hapless Broncos team. And yet, you aren't wrong.
|
|
#
?
Dec 8, 2022 21:20
|
|
|
Kalli posted:The funny thing is the Ravens are the team that'd be in danger of falling out with Lamar's injury You'd think that, but we lost to the Jaguars and just barely beat the Mayfield Panthers and WIlson Broncos. Four of these games are divisional. We will almost certainly lose to Cincinnati on the road at the end of the year. We will probably be playing Huntley at Pittsburgh Sunday and while I like Huntley as a backup he isn't as good as Lamar and won't be able to beat Pittsburgh on the road with the same effort we had against Denver. I really could see 10 wins being like a 50/50 chance for the Ravens the way they are playing. We have struggled in divisional matchups in recent years and could conceivably lose all four of those.
|
|
#
?
Dec 8, 2022 21:23
|
|
|
oh man speaking of "recently playing better". The Steelers have won their last two games (against the Saturday Colts and Falcons), and before that lost a 1 score game* to the Bengals. So they're playing "better." They've got the Ravens twice, @Panthers, Raiders and Browns left. If there's no Lamar against them... they could get to 9 wins. They don't have tiebreakers against either the Pats or Jets, but gently caress... that could happen. * They were down 14 and scored a TD with under a minute to go, so we're lying with the truth here.
|
|
#
?
Dec 8, 2022 21:23
|
|
|
panthers haven't beat the Steelers in a quarter century.
|
|
#
?
Dec 8, 2022 21:45
|
|
|
I hate the Chargers Go Chargers
|
|
#
?
Dec 8, 2022 21:55
|
|
|
Not sure why MIA/LAC would have a negligible impact on the Jets playoff chances but I’m definitely rooting for the Dolphins. https://twitter.com/PFF_Moo/status/1600956828564586496
|
|
#
?
Dec 8, 2022 21:58
|
|
|
Diva Cupcake posted:Not sure why MIA/LAC would have a negligible impact on the Jets playoff chances but I’m definitely rooting for the Dolphins. We should root for the Dolphins but it would be incredible to have a chance to knock them out in week 18.
|
|
#
?
Dec 8, 2022 21:59
|
|
|
I have absolutely no clue what "shannon entropy" is but it's from this new PFF guy
|
|
#
?
Dec 8, 2022 22:02
|
|
|
Turns out that chart was strictly for playoff seeding and not playoff chances. That’s this one. https://twitter.com/PFF_Moo/status/1600961229698174978
|
|
#
?
Dec 8, 2022 22:12
|
|
|
Kalli posted:Which is funny because they've recently lost to Jeff Saturday and needed OT against a hapless Broncos team. Football is just so fun this year
|
|
#
?
Dec 8, 2022 22:15
|
|
|
Would just like to say I'm glad this place exists because good lord Twitter is a cesspool regarding the imaginary Tua vs Herbert rivalry
|
|
#
?
Dec 8, 2022 22:48
|
|
|
FizFashizzle posted:panthers haven't beat the Steelers in a quarter century. First time we've had an oline since like 2006 and our secondary might make something happen but I still just hope we lose out.
|
|
#
?
Dec 8, 2022 22:51
|
|
|
YOLOsubmarine posted:Football is just so fun this year Just imagine if we had an Urban Meyer this year to top it all off.
|
|
#
?
Dec 8, 2022 23:10
|
|
|
SKULL.GIF posted:I have absolutely no clue what "shannon entropy" is but it's from this new PFF guy To go back to defensive coaches and needing talent more than scheme: look at Seattle and Dallas on this chart and remember Dan Quinns back in the news to be a head coach again.
|
|
#
?
Dec 8, 2022 23:15
|
|
|
SKULL.GIF posted:I have absolutely no clue what "shannon entropy" is but it's from this new PFF guy you can think of shannon entropy as a mathematical measure of how random a signal is. the higher the entropy, the more random the signal. this is important in some engineering, math, and cryptography domains. it is, in theory, a cool thing to apply here as some measure of how "multiple" and varied a defense might be. however, as with so much else of pff, it is rendered a completely meaningless phrase because he doesn't loving define what he's applying the measurement to. "coverage scheme shannon entropy" might be about taking the entropy of a signal that looks like ["cover-3", "cover-2", "2-man", "cover-3", "zero"] i guess... but how are those defined? how is zone separated from man? are zone schemes separated within the shell counts by like whether there's a robber, whether it's match, etc etc? also however, is this actually like... a reasonable thing to measure? could a team that plays many different kinds of coverage actually just mean that they're giving up 3 yard plays all the time and they have one coverage they always use on 1st and 10, and a different one on 2nd and 6, and a different one on 3rd and 3? who knows, but there's a cran project that says it does it, let's make a graph None of these shitheads ever provide actual information on what their graphs mean, how they make their data, what might make the graph less useful, what flaws in their methodology might be, what their loving methodology IS, they just go  it's a cool graph man can't argue with the graph if you argue with the graph you're a caveman
Phobeste fucked around with this message at 23:59 on Dec 8, 2022 |
|
#
?
Dec 8, 2022 23:56
|
|
|
Is someone questioning the integrity of an organization run by Austin Collinsworth?
|
|
#
?
Dec 9, 2022 00:04
|
|
|
yeah but it kinda sucks because i like cool graphs, i want more cool graphs, pff paying a bunch of mopes to chart everything out and make it available (or at least more available than team internal data) is cool even if it's not perfect. i just wish they'd do normal poo poo like saying how the graphs get made
|
|
#
?
Dec 9, 2022 00:09
|
|
|
posted:"It's a very conservative pass game," the Cardinals defensive coordinator said during his press availability period. "Lots of screens. All kinds of screens. It's like a defensive guy is calling offense." Wow lmao
|
|
#
?
Dec 9, 2022 00:30
|
|
|
Nodoze posted:Wow lmao Hasn't the patriots offense been this since Moss left?
|
|
#
?
Dec 9, 2022 00:33
|
|
|
Phobeste posted:yeah but it kinda sucks because i like cool graphs, i want more cool graphs, pff paying a bunch of mopes to chart everything out and make it available (or at least more available than team internal data) is cool even if it's not perfect. i just wish they'd do normal poo poo like saying how the graphs get made I agree with everything you said. Existing advanced stats would be a much more interesting topic of conversation if we could literally do more with them than just take Aaron Schatz at his word that he's a football math god king.
|
|
#
?
Dec 9, 2022 00:44
|
|
|
fsif posted:I agree with everything you said. Existing advanced stats would be a much more interesting topic of conversation if we could literally do more with them than just take Aaron Schatz at his word that he's a football math god king. Yea, but if the PFF needs published that information then people who actually understand stats would be empowered to call them dipshits on Twitter and point out why their charts are stupid.
|
|
#
?
Dec 9, 2022 00:51
|
|
|
https://twitter.com/LATimesklein/status/1600997433018044417
|
|
#
?
Dec 9, 2022 00:58
|
|
|
Phobeste posted:you can think of shannon entropy as a mathematical measure of how random a signal is. the higher the entropy, the more random the signal. this is important in some engineering, math, and cryptography domains. it is, in theory, a cool thing to apply here as some measure of how "multiple" and varied a defense might be. basic reading here for anyone who cares: https://en.wikipedia.org/wiki/Entropy_(information_theory) let's say a defense played cover-2 50% of the time, cover-3 25% of the time, and cover-1 25% of the time. each defense type of a separate class, so let's count them as bits 0/1. -1* (0.5*log2(0.5)+2*0.25*log2(0.25) ) = 1.5 in his entropy rating if using log base 2 for one-hot encoding of defense types, for that example. say it was only 30% cover-2, and adding in 20% cover-0. -1 * (0.3*log2(0.3)+2*0.25*log2(0.25)+0.2*log2(0.2)) = 1.985 basically the same ballpark as the numbers he's posting with my 'simple' examples. so somewhere at their site, they have data saying what formation the defense was running for every play, and then they just encoded those as above. it's lovely that he didn't at least link the math, it's not complicated math
|
|
#
?
Dec 9, 2022 00:58
|
|
|
SKULL.GIF posted:I have absolutely no clue what "shannon entropy" is but it's from this new PFF guy Shannon Entropy was a well-known sports analyst who had built a successful career as an expert commentator on American football. She was known for her sharp insights and engaging personality, and had become a household name among sports fans across the country. But what many people didn't know was that Shannon Entropy was actually a pseudonym for Shannon Sharpe, a former NFL player who had retired from the game a few years earlier. Shannon had adopted the new name as a way to distance herself from her playing career and to focus on her new career in broadcasting. Despite her best efforts to keep her identity secret, Shannon's true identity eventually came to light, and she was met with a mix of surprise and admiration from her fans. Many were impressed by her intelligence and quick wit, and by her ability to excel in two very different fields. Shannon continued to thrive as Shannon Entropy, and became even more popular as fans learned more about her background and her journey from football player to sports analyst. She remained a beloved and respected figure in the world of American football, and her name will always be associated with her unique combination of intelligence and athletic prowess.
|
|
#
?
Dec 9, 2022 01:01
|
|
|
I mean probably the most valuable takeaway from that plot is that it's just a fuckin' big blob of points, clearly showing that shannon entropy has no correlation whatever to their expected points allowed/play metric. literally does not matter if your coverage scheme is 'predictable' or 'mixed up' a lot. jim's and joes not x's and o's?
|
|
#
?
Dec 9, 2022 01:07
|
|
|
SKULL.GIF posted:I have absolutely no clue what "shannon entropy" is but it's from this new PFF guy Vikings sneaking into the "good defense" frame should be a clue this graph is bunk.
|
|
#
?
Dec 9, 2022 01:11
|
|
|
https://twitter.com/ndjomo76/status/1600946987913781249?t=odZ6DQ4fPSVermgwFvUAFg&s=19
|
|
#
?
Dec 9, 2022 01:12
|
|
|
pmchem posted:I mean probably the most valuable takeaway from that plot is that it's just a fuckin' big blob of points, clearly showing that shannon entropy has no correlation whatever to their expected points allowed/play metric. literally does not matter if your coverage scheme is 'predictable' or 'mixed up' a lot. jim's and joes not x's and o's? I think it’s probably more subtle than simply “have good players, put them in wherever scheme” but PFF can’t provide any deeper insight than “it could be good OR bad to do one thing OR the other.”
|
|
#
?
Dec 9, 2022 01:12
|
|
|
mcmagic posted:Hasn't the patriots offense been this since Moss left? No
|
|
#
?
Dec 9, 2022 01:27
|
|
|
Dammit Roman is out of the running (heh) for the Stanford job
|
|
#
?
Dec 9, 2022 01:30
|
|
|
|
| # ? Jun 7, 2024 19:55 |
|
|
sharknado slashfic posted:Dammit Roman is out of the running (heh) for the Stanford job for Jason Garrett of all people I may have my issues with Greg Roman but I guarantee he's running a more interesting college offense than Garrett
|
|
#
?
Dec 9, 2022 02:05
|
|
