|
Doltos posted:Football statistics are heretical. They become quasi-paganistic with the NBA and are dogmatic doctrines with baseball. "Yeah, I know my man Uggla is sub-mendoza line on the season, but check out his on base percentage with runners in scoring position" *Uggla strikes out* "Huh, next time, Danny boy."
|
#
?
Apr 21, 2014 18:48
|
|
|
|
| # ? May 28, 2024 01:21 |
|
|
Manziel's statline vs top 10 defenses is so  He's second only to Luck overall IMO but went 1-5 because RIP No DefenseEDIT: Luck went 0-3 too so there's that
|
|
#
?
Apr 21, 2014 18:51
|
|
|
Elotana posted:Manziel's statline vs top 10 defenses is so That stat line is like 90% his two match-ups vs Alabama. He looked like an all-time great in both of those games.
|
|
#
?
Apr 21, 2014 18:53
|
|
|
I would guess top 10 defenses also includes Auburn this year (he owned), LSU both years (nnnope), and Florida last year (that would match the 1-5 record). But poo poo, 2012 Florida was his first game ever, so having that in the average is even more impressive.
|
|
#
?
Apr 21, 2014 19:01
|
|
|
I really mean more like, how well his stats seem to look in the aggregate.
|
|
#
?
Apr 21, 2014 19:02
|
|
|
Also this chart has me praying that the Clowney 1st - Savage 2nd/3rd rumors for the Texans are a smokescreen because goddamn that is some Bad Quarterbacking
|
|
#
?
Apr 21, 2014 19:04
|
|
|
Oh most definitely, but like Daltos keep saying... these stats really are no indicator of our well he's going to do in the Pros. I think the tape is as far as you have to look.
|
|
#
?
Apr 21, 2014 19:08
|
|
|
It's going to be really funny when we spend all this energy creating the TFF Equation for NFL QB Success and we just end up recreating QBR
|
|
#
?
Apr 21, 2014 19:20
|
|
|
bewbies posted:Uh, alright. Regression only shows correlation so I guess we're ok. It would be hell trying to build a good regression for QB play (at least an OLS model) because there is so much endogeneity involved, and the observations would not be I.I.D. I mean, you can already guess what the sign of most of the estimators would be (completion %, TD/Int, Games Started are all positively correlated with NFL success) so the only use for the regression would be to see their magnitude, and those numbers would not be accurate, and hypothesis testing on them would suck due to the endogeneity.
|
|
#
?
Apr 21, 2014 19:21
|
|
|
Blitz7x posted:It's going to be really funny when we spend all this energy creating the TFF Equation for NFL QB Success and we just end up recreating QBR At least then we'd understand QBR.
|
|
#
?
Apr 21, 2014 19:26
|
|
|
Blitz7x posted:It's going to be really funny when we spend all this energy creating the TFF Equation for NFL QB Success and we just end up recreating QBR What would really happen is you'd get a list of good quarterbacks, and a big fat loving WEEDEN right in the middle. Plus you'd miss a bunch of good players. It'd be interesting to see what sort of universals statistically fit most guys we'd consider franchise QB's, but I don't think there's any possible way to use that data because it fluctuates so much and means little to nothing. If a guy in a lesser conference is getting weaker completion percentage than a guy in a top conference (unless he's getting balls dropped like crazy), or two guys in the same conference have wildly different statlines, I think stats tell quite a bit, but outside of that, it's a crapshoot.
|
|
#
?
Apr 21, 2014 19:28
|
|
|
@evansilva ESPN's Adam Schefter said he expects Jadeveon Clowney, Greg Robinson, Khalil Mack, and Sammy Watkins to be the first 4 picks in the draft. Gordon and Watkins would be awesome to watch as a duo.
|
|
#
?
Apr 21, 2014 19:28
|
|
|
Blitz7x posted:It's going to be really funny when we spend all this energy creating the TFF Equation for NFL QB Success and we just end up recreating QBR Goon projects are the best, Johnny Five-Aces.
|
|
#
?
Apr 21, 2014 19:28
|
|
|
Ah man, fuuuuuuuuuuuuuuuuuuuuck, I was gonna make an  post, but then endogeneity happened.
|
|
#
?
Apr 21, 2014 19:29
|
|
|
Ozu posted:@evansilva
|
|
#
?
Apr 21, 2014 19:30
|
|
|
You can do whatever stats you want, all I care about is getting people to stop calling the youtube videos they watch "tape"
|
|
#
?
Apr 21, 2014 19:30
|
|
|
Groucho Marxist posted:You can do whatever stats you want, all I care about is getting people to stop calling the youtube videos they watch "tape" I know right, real game tape is so much better for judging player efficacy.
|
|
#
?
Apr 21, 2014 19:32
|
|
|
Groucho Marxist posted:You can do whatever stats you want, all I care about is getting people to stop calling the youtube videos they watch "tape" Depends on which video you watch. Highlight videos are tape, but they aren't "game tape". There are many, many videos out there that show the best angle on every single snap of a game for a prospect. Those are great but not ideal. Alas I cannot wait for the day where All-22 is the norm for viewing football.
|
|
#
?
Apr 21, 2014 19:36
|
|
|
Ozu posted:@evansilva That means Bridgewater to the Raiders. 
|
|
#
?
Apr 21, 2014 19:47
|
|
|
Is it conceivable the Browns go Oline? I've been seeing it as an option lately but it seems like a stupid idea.
|
|
#
?
Apr 21, 2014 19:51
|
|
|
Atticus Finch posted:Is it conceivable the Browns go Oline? I've been seeing it as an option lately but it seems like a stupid idea. No not really unless they want Robinson as their road grader and eventual replacement for Thomas. That'll only happen if they roll the dice on Hoyer next year while waiting for Winston/Mariota.
|
|
#
?
Apr 21, 2014 19:55
|
|
|
If they think they can get their quarterback later, there's no reason the Browns can't go with BPA at 4. Can't dismiss the trades that can happen at spots 2 and 4 as well.
|
|
#
?
Apr 21, 2014 19:57
|
|
|
Nawid posted:That means Bridgewater to the Raiders. Didn't you hear? Bridgewater isn't a first round guy anymore  I have to say, I'm getting less and less hyped about the draft with each passing day. Just so burnt out on it. Stupid Goodell
|
|
#
?
Apr 21, 2014 20:02
|
|
|
For some reason all my game tape has lovely nu-metal on it
|
|
#
?
Apr 21, 2014 20:08
|
|
|
My VCR doesn't even have wifi so you guys are way beyond me
|
|
#
?
Apr 21, 2014 20:12
|
|
|
Intruder posted:Didn't you hear? Bridgewater isn't a first round guy anymore At least Walter is now saying that there is no reason to not pick Clowney since Bridgewater is going to the second. Which seems to imply that he actually is a first round talent and he's just buying into bullshit but hey, logic is hard.
|
|
#
?
Apr 21, 2014 20:13
|
|
|
Rasczak posted:For some reason all my game tape has lovely nu-metal on it That's a real knock, IMO. When there's some hardcore rap going in the background, you know you're lookin at a real difference maker.
|
|
#
?
Apr 21, 2014 20:13
|
|
|
So, I'm doing this thing. I'm populating stats from 30 randomly chosen QBs drafted in the first 4 rounds since 1998.Pron on VHS posted:It would be hell trying to build a good regression for QB play (at least an OLS model) because there is so much endogeneity involved, and the observations would not be I.I.D. I would have thought this as well, but so far at least based on an eyeball of the data there doesn't seem to be a lot of overlap. I'm a little more concerned about multicollinearity but I've kept my parameters limited and none really has much do with any others insofar as football is concerned so the output might be alright.
|
|
#
?
Apr 21, 2014 20:15
|
|
|
AAA DOLFAN posted:My VCR doesn't even have wifi so you guys are way beyond me What's a VCR? Would it be an upgrade over my reel to reel?
|
|
#
?
Apr 21, 2014 20:16
|
|
|
I watch all of my game film on this. One game is 10 drat pounds of radial slides.
|
|
#
?
Apr 21, 2014 20:19
|
|
|
Doltos posted:This is absolutely awesome. A super sperg lord on Reddit compiled the stats of most draft eligible QBs of the last 3 years: Completion % vs top ten defenses is pretty telling. If the player is at or above 60% they seem to be good pros. Of course there's only a few games for each player except Manziel. Completion % vs top 35 defenses is also interesting for at or above 60% although you get a lot more of those guys who seem like average backup QBs in the NFL like Ponder. Interestingly the jump from top 35 to top 10 cuts out guys like Tebow and Ponder while keeping Luck, Wilson, Bradford. If you want to read a whole lot into that the guys who have greater than 60% vs top 35 but not top 10 will at most be mediocre QBs like Manuel, Geno, Tebow etc. Either way the main things that stand out in that comparison: 1. Manziel is loving ridiculously accurate 2. Ditto Luck and Wilson his final year 3. Bridgewater is solid. His completion % actually doesnt vary at all from top 35 to top 10 4. Carr looks like the 3rd QB behind Manziel and Bridgewater 5. I love the Andy Dalton completion % Defenses outside of top 35: 63.6% Top 35: 57.5% Top 10: 43% (Admittedly only one game, but still hilarious) We really need one more year's worth of players to compare if only because there aren't enough decent pros on the list to read into.
|
|
#
?
Apr 21, 2014 20:32
|
|
|
bewbies posted:So, I'm doing this thing. I'm populating stats from 30 randomly chosen QBs drafted in the first 4 rounds since 1998. How are you quantifying "success in the NFL"? Aaron Rodgers was allowed to sit 3 years and work with an excellent QB development coach. Is that a fair comparison with a hypothetical plauer who has a career-ending injury in pre-season of his rookie year. Being drafted by the Browns also seems to be a curse on its own.
|
|
#
?
Apr 21, 2014 20:37
|
|
|
bewbies posted:So, I'm doing this thing. I'm populating stats from 30 randomly chosen QBs drafted in the first 4 rounds since 1998. There isn't really a problem if the explanatory variables like games started and TD/INT ratio have correlation, it's just that there are so many undiscovered variables lurking in the error term (injuries, offensive scheme) that would muck things up. edit: to be honest I am kind of being a pussy, this would be a great and fun thing to do and in undergrad I used NFL data for all econometrics projects, its just that the results won't have very good properties and be very useful. If we spend months and months and get as close as we can then there's probably a chance the resulting model could get TFF hired at footballoutsiders or something also I don't know if MJBuddy still reads these forums in the offseason but he would be a good source to ask regarding how to do an accurate regression. I don't know much beyond OLS and regression discontinuity models. Pron on VHS fucked around with this message at 21:17 on Apr 21, 2014 |
|
#
?
Apr 21, 2014 21:13
|
|
|
Atticus Finch posted:Is it conceivable the Browns go Oline? I've been seeing it as an option lately but it seems like a stupid idea. It is conceivable if they want to draft Robinson as a RT which is a bad idea but it's the Browns so who knows.
|
|
#
?
Apr 21, 2014 21:14
|
|
|
Robinson would be a loving road grader at RT, while also being the LT of the future.
|
|
#
?
Apr 21, 2014 21:27
|
|
|
Chilichimp posted:Robinson would be a loving road grader at RT, while also being the LT of the future. Yeah but if you want a road grader you can't do much better than Lewan. Plus you can get him 10 picks later.
|
|
#
?
Apr 21, 2014 21:41
|
|
|
I don't see how Robinson lands at the intersection of best player available and team need, but I'd be OK with it
|
|
#
?
Apr 21, 2014 21:56
|
|
|
Doltos posted:Yeah but if you want a road grader you can't do much better than Lewan. Plus you can get him 10 picks later. But then you're doing the objectively morally repugnant action of hiring Taylor Lewan
|
|
#
?
Apr 21, 2014 21:56
|
|
|
Alouicious posted:But then you're doing the objectively morally repugnant action of hiring Taylor Lewan Being banished to Cleveland seems like suitable punishment
|
|
#
?
Apr 21, 2014 22:12
|
|
|
|
| # ? May 28, 2024 01:21 |
|
|
Intruder posted:Didn't you hear? Bridgewater isn't a first round guy anymore Same, I just want it over with already. If the first four teams all pass on Bridgewater they'll each deserve the five years of losing they'll get for it.
|
|
#
?
Apr 21, 2014 22:33
|
|
