|
Lessail posted:What type of not happened are we talkin about here The type where the muff happened after the Pats touched the ball first.
|
# ? Dec 4, 2014 18:33 |
|
|
# ? Apr 25, 2024 03:28 |
|
These guys are doing some interesting things in team stats/predictions. While they describe the general ideas behind their methodology (and it sounds, well, sound), it's still closed to the public.
|
# ? Dec 4, 2014 18:36 |
|
African AIDS cum posted:Yeah thats him, the most boring man to listen to in the entire world. Glad he hasn't been on recently. I guarantee Simmons will have Schatz on the BS Report again the week before the playoffs so that Simmons can fantasize about the Patriots winning another Super Bowl.
|
# ? Dec 4, 2014 18:38 |
|
King Hong Kong posted:I guarantee Simmons will have Schatz on the BS Report again the week before the playoffs so that Simmons can fantasize about the Patriots winning another Super Bowl. Simmons' theory now is the Patriots threw the game by staying very vanilla so not to tip off other teams they have upcoming. Dude is unhinged.
|
# ? Dec 4, 2014 18:40 |
|
African AIDS cum posted:Simmons' theory now is the Patriots threw the game by staying very vanilla so not to tip off other teams they have upcoming. Dude is unhinged. And the funniest and saddest thing is that there have been at least one other Patriots writer to use that same theory for the loss.
|
# ? Dec 4, 2014 19:09 |
|
TubeStank posted:And the funniest and saddest thing is that there have been at least one other Patriots writer to use that same theory for the loss. Link?
|
# ? Dec 4, 2014 19:10 |
|
African AIDS cum posted:Link? http://www.boston.com/sports/columnists/wilbur/2014/12/in_patriots_loss_to_the_packers_bill_belichick_lul.html
|
# ? Dec 4, 2014 19:11 |
|
Blotto Skorzany posted:http://www.boston.com/sports/columnists/wilbur/2014/12/in_patriots_loss_to_the_packers_bill_belichick_lul.html loving lol. Incredible.
|
# ? Dec 4, 2014 19:14 |
|
Bill does it again Beautiful
|
# ? Dec 4, 2014 19:17 |
|
Blotto Skorzany posted:http://www.boston.com/sports/columnists/wilbur/2014/12/in_patriots_loss_to_the_packers_bill_belichick_lul.html Please don't troll.
|
# ? Dec 4, 2014 19:24 |
|
Huh?
|
# ? Dec 4, 2014 19:28 |
|
Losing to the Packers while trying to get homefield advantage throughout the playoffs? Genius. -Boston sports journalism
|
# ? Dec 4, 2014 19:54 |
|
Blotto Skorzany posted:http://www.boston.com/sports/columnists/wilbur/2014/12/in_patriots_loss_to_the_packers_bill_belichick_lul.html This reads like an Onion article.
|
# ? Dec 4, 2014 19:58 |
|
That's
|
# ? Dec 4, 2014 20:02 |
|
Advanced stats for football are like the study of economics, there are such a vast array of unquantifiable factors that, even if test conditions were ideal and controlled, there would be absolutely no agreement between experts as to what does what, causes what, or is a result of what. It's all a bunch of hocus-pocus guesswork in the guise of math, and offers no better a predictive model than an eyeball test. Or at least that's what I tell myself because I don't understand any of it, but I barely passed both stats and macroeconomics, so what the gently caress do I know.
|
# ? Dec 4, 2014 20:14 |
|
Thaddius the Large posted:Advanced stats for football are like the study of economics, there are such a vast array of unquantifiable factors that, even if test conditions were ideal and controlled, there would be absolutely no agreement between experts as to what does what, causes what, or is a result of what. It's all a bunch of hocus-pocus guesswork in the guise of math, and offers no better a predictive model than an eyeball test. You're 100% right
|
# ? Dec 4, 2014 20:18 |
|
Thaddius the Large posted:Advanced stats for football are like the study of economics, there are such a vast array of unquantifiable factors that, even if test conditions were ideal and controlled, there would be absolutely no agreement between experts as to what does what, causes what, or is a result of what. It's all a bunch of hocus-pocus guesswork in the guise of math, and offers no better a predictive model than an eyeball test. I don't know if they're similar or not, because football is real and the economy isn't, but this sounds pretty spot on to me.
|
# ? Dec 4, 2014 20:23 |
|
Plafop posted:JJ Watt is underpaid. He signed such a long contract that toward the end of it he'll probably be severely underpaid, same thing happened to Andre Johnson when he signed a super long contract making him the highest paid WR in the league and within a couple years he wasn't even top 5
|
# ? Dec 4, 2014 20:30 |
|
That Boston article is absolutely amazing. I didn't know ehuds old co worker had a blog
|
# ? Dec 4, 2014 20:33 |
|
Rooster Brooster posted:These guys are doing some interesting things in team stats/predictions. While they describe the general ideas behind their methodology (and it sounds, well, sound), it's still closed to the public. Do they explain what they do any better than this? http://massey-peabody.com/methodology/ Because that doesn't even have the barest of explanations of what their metrics actually do.
|
# ? Dec 4, 2014 20:39 |
|
Sounds like some queero poo poo to me
|
# ? Dec 4, 2014 20:42 |
|
quote:The Patriots lost their third game of the season at Lambeau Field, where the Packers are now 6-0 this season, and are averaging 41 points per game. See I know he means it's the Patriots third loss of the season but the way he writes is really bad and doesn't make it clear that the Patriots haven't lost three times at Lambeau in one year quote:How to explain Belichick not going for it on fourth down at a couple opportune moments in the first half, even knowing you risked giving the ball back to a quarterback in Rodgers who was doing donuts around the Patriots’ secondary?
|
# ? Dec 4, 2014 20:48 |
|
Pats fans are pathetic.
|
# ? Dec 4, 2014 21:02 |
|
I think that article is a joke, guys
|
# ? Dec 4, 2014 21:48 |
|
The_Hat posted:I think that article is a joke, guys Based on what?
|
# ? Dec 4, 2014 21:57 |
|
The_Hat posted:I think that article is a joke, guys it's very very real
|
# ? Dec 4, 2014 22:04 |
|
It's a matter of probability. What is more likely, a Patriot's writer being self aware enough and a good enough writer to compose a complete piece of satire without ever breaking character or they are simply delusional enough to think any loss is some sort of 5th dimensional chess game by Billy 5 Aces.
|
# ? Dec 4, 2014 22:09 |
|
Football is like four balls on the edge of a cliff
|
# ? Dec 4, 2014 22:11 |
|
tractor fanatic posted:Do they explain what they do any better than this? http://massey-peabody.com/methodology/ Because that doesn't even have the barest of explanations of what their metrics actually do. I don't think there is much more detail on the mechanics, but their stated goal is to predict spread lines for the purpose of winning at betting.
|
# ? Dec 4, 2014 22:17 |
|
African AIDS cum posted:Simmons' theory now is the Patriots threw the game by staying very vanilla so not to tip off other teams they have upcoming. Dude is unhinged. At least on his podcast he brought this up as a joke theory and said he didn't believe it,
|
# ? Dec 4, 2014 22:38 |
|
My new goal is to start a blog doing some legitimate football analysis (there's got to be some interesting thing to chart that nobody tracks yet, which doubles as an excuse to buy game rewind) and keep it going long enough that Aaron Schatz follows me on twitter, then goatse him
|
# ? Dec 4, 2014 22:41 |
|
Blotto Skorzany posted:(there's got to be some interesting thing to chart that nobody tracks yet, which doubles as an excuse to buy game rewind) You know how everyone talks about "wearing out the defense" with time of possession and high play counts? You could test to see if that's true, if defenses that give up lots of plays or lots of TOP start giving up more and more average yards per play later in the game.
|
# ? Dec 4, 2014 23:03 |
|
Make up your own stats language and be a step ahead of the R kids.
|
# ? Dec 5, 2014 00:30 |
|
Came for the thread title, stayed for the ridiculous Pats article. Never change, Boston.
|
# ? Dec 5, 2014 00:50 |
|
Hand Row posted:Make up your own stats language and be a step ahead of the R kids. That's literally what DVOA is
|
# ? Dec 5, 2014 00:58 |
|
Hand Row posted:Make up your own stats language and be a step ahead of the R kids. The ~data science~ crowd has already sort of done this with Julia (python/scipy is also popular with them), but imo Haskell or OCaml would probably work very well. Brushing up on one of them would also probably get me lucrative job offers provided I can hit the mute button on my conscience. full disclosure: I used R to do analysis and visualizations for a couple of projects in school, it was great compared to Excel. Just having a reasonable syntax and a repl and built in arbitrary precision arithmetic goes really far. I switched to Common Lisp for my actual ECSE department prob/stats course (basically to check my answers on homework assignments, some of the integrals get real hairy to do analytically), it's a better language but doesn't match R's built in visualization libraries.
|
# ? Dec 5, 2014 01:16 |
|
Brannock posted:You know how everyone talks about "wearing out the defense" with time of possession and high play counts? You could test to see if that's true, if defenses that give up lots of plays or lots of TOP start giving up more and more average yards per play later in the game. I'm definitely curious about this. I bet that at anecdotal levels of high school and even college games, where lack of conditioning and talent disparity is much greater, that it happens. But in the pros, especially when dlinemen are being rotated? Who knows
|
# ? Dec 5, 2014 02:06 |
|
mastershakeman posted:I'm definitely curious about this. I bet that at anecdotal levels of high school and even college games, where lack of conditioning and talent disparity is much greater, that it happens. But in the pros, especially when dlinemen are being rotated? Who knows You're going to wear out the 300-pound linemen, probably not so much the corners and other "skill" players. I think the real benefit of the perma-no-huddle and the like is running plays quickly enough that your opponent doesn't have much time to make adjustments and in some cases substitutions.
|
# ? Dec 5, 2014 02:15 |
|
OneThousandMonkeys posted:You're going to wear out the 300-pound linemen, probably not so much the corners and other "skill" players. I think the real benefit of the perma-no-huddle and the like is running plays quickly enough that your opponent doesn't have much time to make adjustments and in some cases substitutions. Those corners are running way up and down the field, and it's way easier on you to run routes as a receiver than having to cover them as a corner.
|
# ? Dec 5, 2014 02:47 |
|
|
# ? Apr 25, 2024 03:28 |
|
DVOA was pretty revolutionary when it came out. It's massively more predictive than anything that was in use at the time, which was basically strait yardage stats, wins, and narrative. I'm sure there are folks doing better work these days but I haven't heard of them. PFF is interesting but it doesn't do a Non-rhetorical question: what "advanced stat" out there is currently the best at predicting future wins?
|
# ? Dec 5, 2014 07:56 |