2019 Texas Football Season - "#8 Go Fuck Yourself Bitch"

[texifornia 57224]

 

[Fud 9560]

5 hours ago, closetojumping said:

The premise of a 4 loss season has nothing to do with anything logically being discussed. The original tweet referenced Texas finishing the season unranked. Not the regular season, which would have been a bizarre and irrelevant point, but yet that is the premise you, sarvanaash and a multitude of others seem bizarrely attached to in this exercise.

There will not be an 8-4 finish to Texas this season, irrespective of whether or not they're ranked at the time of season completion. They'll be one of: 0-12, 1-11, 2-10, 3-9, 4-8, 5-7, 5-8, 6-7, 7-6, 8-5, 9-4, 9-5, 10-3, 10-4, 11-2, 11-3, 11-4, 12-2, 12-3, 13-1, 13-2, 14-1, 15-0. Notice there isn't a fucking 8-4 in there, rube-ric. Barring a natural disaster, act of terror, or aliens shooting laser beams from outer space as they invade our planet, it is fucking impossible for Texas to go 8-4 this season. And, at the point in which they are 8-4, should they hit that mark, they'll have another game to play, and after that game, they may or may not finish the fucking season unranked. 

My original issue is that sarvanassh has a pathological need to poormouth Texas results, and it carries over into patently false, disprovable outcomes, such as a Texas team hitting 8-4, and then not going to a bowl, and finishing the season, with that record, unranked. You didn't pause long enough to consider this reality, as you and dipshits like you can't wait for me and my ilk to be wrong so that you can throw rocks at me from the back of the crowd and point out to no one in particular what a dumbass I am. That, as usual for you, didn't work out so well.

As to the premise of whether or not Texas could lose 4 games this year? I just pointed out that that is possible, which I never fucking claimed otherwise. It's "possible" that Texas could lose 12 games this season. Chaos, while a ladder, is also a weird thing that I do not discount. Anything could happen, and I have fucking accounted for most of the known anythings in this very post. 

 

 

He’s tasked with creating a model that is predictive for the most teams possible. It just happens to be that 2019 Texas probably doesn’t fit the model very well. If he wanted to, he could probably run custom cases for random teams that appear to be outliers, but then he runs into the issue potentially being biased, since he wouldn’t give the same consideration to every team. 

If Texas starts beating the likes of Tulsa, Kansas, Kansas State, Baylor, and Texas Tech (with its backup QBs) by more than one score, then all of the sudden the model will quickly correct. Maybe Texas should start doing that?

I’m not as bothered by the S&P+ ranking than most. It captures most teams more or less correctly; 2018 Texas going forward towards 2019 just isn’t most teams. 

[Hank Chinaski 6324]

1 hour ago, Fud said:

He’s tasked with creating a model that is predictive for the most teams possible. It just happens to be that 2019 Texas probably doesn’t fit the model very well. If he wanted to, he could probably run custom cases for random teams that appear to be outliers, but then he runs into the issue potentially being biased, since he wouldn’t give the same consideration to every team. 

If Texas starts beating the likes of Tulsa, Kansas, Kansas State, Baylor, and Texas Tech (with its backup QBs) by more than one score, then all of the sudden the model will quickly correct. Maybe Texas should start doing that?

I’m not as bothered by the S&P+ ranking than most. It captures most teams more or less correctly; 2018 Texas going forward towards 2019 just isn’t most teams. 

Yup. It's like BMI. 

[Not a Sock 2342]

1 hour ago, Fud said:

He’s tasked with creating a model that is predictive for the most teams possible. It just happens to be that 2019 Texas probably doesn’t fit the model very well. If he wanted to, he could probably run custom cases for random teams that appear to be outliers, but then he runs into the issue potentially being biased, since he wouldn’t give the same consideration to every team. 

If Texas starts beating the likes of Tulsa, Kansas, Kansas State, Baylor, and Texas Tech (with its backup QBs) by more than one score, then all of the sudden the model will quickly correct. Maybe Texas should start doing that?

I’m not as bothered by the S&P+ ranking than most. It captures most teams more or less correctly; 2018 Texas going forward towards 2019 just isn’t most teams. 

It’s more that bill Conolley is just a huge bitch and his system creates huge fucking outliers and then he doubles down and tries to lay blame on the teams for not fitting his models. 

[texifornia 57224]

4 minutes ago, Not a Sock said:

It’s more that bill Conolley is just a huge bitch and his system creates huge fucking outliers and then he doubles down and tries to lay blame on the teams for not fitting his models. 

Honestly it's not terrible but we really notice it because WE'RE the huge outlier.

(Or at least Tom Herman teams. Strong's teams were accurately judged to be shitty.)

[closetojumping 63088]

2 hours ago, Fud said:

He’s tasked with creating a model that is predictive for the most teams possible. It just happens to be that 2019 Texas probably doesn’t fit the model very well. If he wanted to, he could probably run custom cases for random teams that appear to be outliers, but then he runs into the issue potentially being biased, since he wouldn’t give the same consideration to every team. 

If Texas starts beating the likes of Tulsa, Kansas, Kansas State, Baylor, and Texas Tech (with its backup QBs) by more than one score, then all of the sudden the model will quickly correct. Maybe Texas should start doing that?

I’m not as bothered by the S&P+ ranking than most. It captures most teams more or less correctly; 2018 Texas going forward towards 2019 just isn’t most teams. 

I think you quoted the wrong post. I’ve got an understanding of Connelly’s stuff and no real beef or intetest with it. The post you quoted was the closing argument to the logic and reader errors being espoused a group of twits on this board. 

[Fud 9560]

8 minutes ago, closetojumping said:

I think you quoted the wrong post. I’ve got an understanding of Connelly’s stuff and no real beef or intetest with it. The post you quoted was the closing argument to the logic and reader errors being espoused a group of twits on this board. 

Was meant to be this one. I guess you were somewhere in the proximity when I went back to find it 

8 hours ago, sushihorn said:

The idiot who wrote this has been constantly wrong about Tom Herman and seems to take the proof of his incompetence personally.  Connelly refuses to explicitly acknowledge Herman's demonstrated ability to get his team up for the big game.  5-2 vs the top 25 in 2018 and 6=0 at UH.  The only exception was with a dysfunctional 2017 Texas rebuilding from 3 straight losing seasons and starting a true freshman QB.

The chance of losing a game the Horns shouldn't from the remaining 10 on the schedule makes this knob think a ranking of #31 or even #26 is deserved at 11-1 or even 10-2.  I doubt that even an 8-4 Texas is outside the top 25 at the end of the regular season.  Yet the same idiot has aggy ranked much higher even though Vegas has them as dogs in 5 games and 7.5 season win O/U compared to 9.5 for Texas.

Talk is cheap.  The people wagering actual money think Texas is a much better team than Connelly does.  They also have an actual stake in the outcome.

 

[HtownHorn 14916]

He uses a 5 year rolling average of results, yet there isn't a single player on the Texas roster that played in the 2014 season. Also the coaching staff has completely changed, as has the offensive philosophies. It's hilarious that team A is 17-10 in the last 2 years, and lost a lot of defensive starters, their leading rusher, and their leading receiver while Team B is 16-10 over the past 2 years, and lost a lot of defensive starters, their leading rusher, and their leading receiver.

One is rated in the 31, the other is 11.

[Pato del Muerto 52858]

Team A is losing to Oklahoma state and Maryland, team B is losing to Alabama and Clemson. 

[Patricio Swayze 17058]

Oh for fuck’s sake, Saturday can’t get here soon enough.

[sushihorn 3233]

16 minutes ago, Pato del Muerto said:

Team A is losing to Oklahoma state and Maryland West Virginia, team B is losing to Alabama and Clemson. Mississippi State and Auburn.

FIFY

Of course you have to also disregard wins completely and focus on "quality losses" or win-losses to make this argument anyway.

[Pato del Muerto 52858]

I’m assuming he does, or is there another explanation for the expected performance delta between the two teams?

[sushihorn 3233]

24 minutes ago, HtownHorn said:

He uses a 5 year rolling average of results, yet there isn't a single player on the Texas roster that played in the 2014 season. Also the coaching staff has completely changed, as has the offensive philosophies. It's hilarious that team A is 17-10 in the last 2 years, and lost a lot of defensive starters, their leading rusher, and their leading receiver while Team B is 16-10 over the past 2 years, and lost a lot of defensive starters, their leading rusher, and their leading receiver.

One is rated in the 31, the other is 11.

It's a status quo model.  It relies on the same ceteris paribus (all else being equal) assumption that produces so many flawed projections in static economic models.  Radical changes in personnel quality will be reflected slowly.  Radical changes in coaching quality will not be measured at all except after the fact.  In both cases, the 5 year look back period ensures that his projections are literally years behind the curve when a program does hit an inflection point.

Even so, it's less that Connelly's system is flawed than that he is unwilling to admit or even accept it.  If he had just stopped 2 years ago when he wrote that his system doesn't predict outcomes for Tom Herman coached teams none of this would be an issue.  The snide commentary, the insistence that his model is correct and the anger at Texas for refusing to conform to it are exceedingly unprofessional. 

[Tex Long 4509]

The test of any model is quality of its predictions. Are his of any use in wagering? 

[Machinator 32099]

2 minutes ago, Tex Long said:

The test of any model is quality of its predictions. Are his of any use in wagering? 

I think his model has performed in the 50%-55% range against the spread in previous years..

[Tex Long 4509]

So, not much better than a coin flip. As expected.

[SydneyCarton 48296]

If he achieved 5% over a coin flip for Vegas then that is pretty impressive. 

[Fud 9560]

48 minutes ago, SydneyCarton said:

If he achieved 5% over a coin flip for Vegas then that is pretty impressive. 

I think it fluctuates between 51-54%. That’s not bad as a starting point since you can then apply other factors like injuries, etc that aren’t considered in the model 

[Magus Ossis 2883]

Teams that win a lot, score a lot, etc can reasonably be expected to continue winning and scoring, all things being equal, as @sushihorn noted that BC predicts. His “model” is just data mining. It isn’t evil or wrong. As stated, though, he defends it as if it embodies some truth with which we are out of synch. We are merely a team with significant contribution from factors he ignores in his formula. His formula gets a few things right, even about us. It down rates us because Hager couldn’t get to the passer and because he’s been replaced. It downrates us because Strong was the highest losing percent coach we’ve ever had. It fails to penalize us for Floyd’s condition. It is a simplistic model that gets enough right to embolden its author to bleat publicly in defiance of its failings. Against the spread, it is roughly as insightful as just betting home dogs. BC has a platform which he uses, among other things, to take shots at us. We have a platform pit from which we Surls take a few in return.

[Droopy 2634]

18 hours ago, SydneyCarton said:

 

I’ve gotta say man, posts like this honestly make me wonder why you even follow UT and college football. I mean, what’s the fucking point for you? Living in a perpetually dark cave with no hope, optimism, or chance for success. It seems like a miserable, wretched existence. 

Because he’s an aggy. He has a tell, besides his shitty takes. Sloppy. 

[SydneyCarton 48296]

6 minutes ago, Droopy said:

Because he’s an aggy. He has a tell, besides his shitty takes. Sloppy. 

If he’s an Aggie this dedicated to this persona, then that’s an even more miserable existence. Because it takes a lot of time and work to follow UT the way satya does...he’s extremely informed. Occam’s Razor says he’s just miserable. 

[Droopy 2634]

13 minutes ago, SydneyCarton said:

If he’s an Aggie this dedicated to this persona, then that’s an even more miserable existence. Because it takes a lot of time and work to follow UT the way satya does...he’s extremely informed. Occam’s Razor says he’s just miserable. 

Aggy. At the end of every YouTube video posted, you can see suggestions/previous videos watched by the poster. His always have aggy highlights. JFF, beating Bama  that one time, etc... Never Texas. Seems like a tell when mixed with his takes and obsession. Occam’s razor alright. 

Edited August 29, 2019 by Droopy

[hookem2010 2644]

An 8-4 season isn't 'miserable'. It just represents a step back in line with the guys we're losing and the uncertainty at RB and LB. No one was 'miserable' when we were 10-3 in 2007 and Colt struggled coming off the heels of a promising 2006 season.

Loss to aggy, loss to ou, and blowout loss at home to ksu... yea, I'd say that 2007 was pretty fucking miserable.

Sent from my SM-G950U using Tapatalk

[KYHorn 17264]

15 minutes ago, Droopy said:

Aggy. At the end of every YouTube video posted, you can see suggestions/previous videos watched by the poster. His always have aggy highlights. JFF, beating Bama  that one time, etc... Never Texas. Seems like a tell when mixed with his takes and obsession. Occam’s razor alright. 

Unless it's a screen capture, I'm pretty sure those are video suggestions based on your watching history.

[Droopy 2634]

14 minutes ago, KYHorn said:

Unless it's a screen capture, I'm pretty sure those are video suggestions based on your watching history.

Nope. I watch a shit load of YouTube and don’t watch aggy highlights. It’s something I noticed a while ago and never said anything but dude loves to doom and gloom. Other posters of videos do the same thing. Just my opinion, but curious to see if it changes going forward. I assume you could go look at old videos he’s posted and it would still show the same videos I’m talking about. Doesn’t really matter I guess but since I’ve noticed it I take his posts with a lot of salt. 

Edited August 29, 2019 by Droopy

[SydneyCarton 48296]

13 minutes ago, KYHorn said:

Unless it's a screen capture, I'm pretty sure those are video suggestions based on your watching history.

 

DUN DUN DUN!

[Lerka Lerka 1094]

So Droopy just outed himself as aggy. Noted.  

[texifornia 57224]

10 minutes ago, Droopy said:

Nope. I watch a shit load of YouTube and don’t watch aggy highlights. It’s something I noticed a while ago and never said anything but dude loves to doom and gloom. Other posters of videos do the same thing. Just my opinion, but curious to see if it changes going forward. I assume you could go look at old videos he’s posted and it would still show the same videos I’m talking about. Doesn’t really matter I guess but since I’ve noticed it I take his posts with a lot of salt. 

I get aggy video suggestions all the time, the algorithm just captures that you like college football in Texas, it's not great at getting much more granular than that.

The same type of algorithm that's good at presenting you similar music choices is naturally going to suck at figuring out why this kind of "related" content doesn't work the same way.

[KYHorn 17264]

8 minutes ago, Droopy said:

Nope. I watch a shit load of YouTube and don’t watch aggy highlights. It’s something I noticed a while ago and never said anything but dude loves to doom and gloom. Other posters of videos do the same thing. Just my opinion, but curious to see if it changes going forward. I assume you could go look at old videos he’s posted and it would still show the same videos I’m talking about. Doesn’t really matter I guess but since I’ve noticed it I take his posts with a lot of salt. 

Just watched a few videos he embedded, each one twice: once in my regular browser (and it had me logged into youtube) and once in incognito mode (didn't know who I was). In regular browser, it gave me personalized videos based on my history, which was mostly Texas football clips and some CR stuff, plus an ad for my employer. In incognito mode, it suggested videos from various teams, with a few being Texas football. Only Aggy video in any of them was a Texas fan talking with an Aggy fan. Not pointing any fingers, nor defending Satya, but I am saying that I think it's pretty unlikely that he's Aggy, and more likely he's just not a blast a parties.

[HOOK'EMHOOAH 1610]

18 hours ago, Tex Long said:

Well-nigh obligatory:

F.

B.

C.

 

... and his dog...

... and his mother ...

... and his mother's dog ...

... and his dog's mother.

 

 

 

You forgot the horse he rode in on.

[Huckleberry 33523]

1 hour ago, Magus Ossis said:

Teams that win a lot, score a lot, etc can reasonably be expected to continue winning and scoring, all things being equal, as @sushihorn noted that BC predicts. His “model” is just data mining. It isn’t evil or wrong. As stated, though, he defends it as if it embodies some truth with which we are out of synch. We are merely a team with significant contribution from factors he ignores in his formula. His formula gets a few things right, even about us. It down rates us because Hager couldn’t get to the passer and because he’s been replaced. It downrates us because Strong was the highest losing percent coach we’ve ever had. It fails to penalize us for Floyd’s condition. It is a simplistic model that gets enough right to embolden its author to bleat publicly in defiance of its failings. Against the spread, it is roughly as insightful as just betting home dogs. BC has a platform which he uses, among other things, to take shots at us. We have a platform pit from which we Surls take a few in return.

This is basically the only thing that bothers me about Connelly. His sensitivity and defiant defense of his system as though it is The Truth and others just don't get it. As a fellow producer of advanced ratings systems I think I'm qualified to note that they are inherently imperfect and those who produce them should be more aware of their imperfections than anyone. 

The fact that there is a subset of teams who share specific traits in style of play and whom your system struggles to get a handle on doesn't mean they play wrong like Connelly implies. It means your system doesn't handle them well. If that subset is small enough it's probably not worth it to try to change your system to capture them if it's intended to handle the entire team universe. Just say that instead of whining and getting defensive when they're pointed out. 

Then there's the fact that as proven this offseason, the entire time he's combatively defending his system against all questioning he's simultaneously researching changes and tweaks he can make to his system to improve it because he knows it's not perfect. After the 2018 season he changed his ratings to keep preseason data in all season long, he changed them to adjust ratings based on the conference a team is in. This last change is based on a completely arbitrary human grouping of teams that makes zero sense from an analytical perspective. He is using conference grouping as a proxy for team strength and may not even realize it. Essentially it is a post-processing corrective step that simply shifts good teams up and bad teams down. He made major changes, and they are definitely major, to his system immediately after spending 2018 defending his system as though it was perfect.

And for all of his arbitrary changes to the system, consider the fact that if Alabama and South Florida had completely swapped coaching staffs on February 15th, 2019 that his preseason ratings would not change. At all. Any reasonable person would understand and publicly recognize inherent weaknesses like that in their creation.

Edited August 29, 2019 by Huckleberry

[Droopy 2634]

10 minutes ago, KYHorn said:

Just watched a few videos he embedded, each one twice: once in my regular browser (and it had me logged into youtube) and once in incognito mode (didn't know who I was). In regular browser, it gave me personalized videos based on my history, which was mostly Texas football clips and some CR stuff, plus an ad for my employer. In incognito mode, it suggested videos from various teams, with a few being Texas football. Only Aggy video in any of them was a Texas fan talking with an Aggy fan. Not pointing any fingers, nor defending Satya, but I am saying that I think it's pretty unlikely that he's Aggy, and more likely he's just not a blast a parties.

Fair enough. Still have my doubts. 

[Hiphopopotamos 20894]

 

[SydneyCarton 48296]

29 minutes ago, KYHorn said:

 Not pointing any fingers, nor defending Satya, but I am saying that I think it's pretty unlikely that he's Aggy, and more likely he's just not a blast a parties.

Not sure any of us should ever, ever, defend Satya

29 minutes ago, KYHorn said:

Not pointing any fingers, nor defending Satya, but I am saying that I think it's pretty unlikely that he's Aggy, and more likely he's just not a blast a parties Life.

FIFY

 

[Rickylovesweed 14829]

19 hours ago, SydneyCarton said:

This was going to be my third point, but you beat me to it...nothing bad ever happens to any other team except UT in his mind. It's unbelievable. OU is just going to turn in 5 more NFL draftable OL this very season and they'll be dominant from the start. 

So much this. I can see why people think he’s an aggy. 

It’s not even that he’s pessimistic about everything Texas related it’s the fact he also gives schools like OU and A&M the benefit of the doubt on everything. 

I don’t think he’s an aggy because nobody can be that dedicated to following Texas. He’s just a miserable SOB that nobody likes. 

 

[Post Oak 20327]

23 minutes ago, Hiphopopotamos said:

 

Damn.  Hug your babies

[Zavala 1384]

 

how many days

[immamac 43555]

 

 

[Walden Ponderer 12639]

1 hour ago, texifornia said:

I get aggy video suggestions all the time, the algorithm just captures that you like college football in Texas, it's not great at getting much more granular than that.

The same type of algorithm that's good at presenting you similar music choices is naturally going to suck at figuring out why this kind of "related" content doesn't work the same way.

CSB, I am having to tweak my Facebook language to reference "Texas" instead of "UT" because, now that I'm east of the Mississippi, I've been deluged with Rocky Top advertising.

Computers are great, except for when they aren't. 

[Katfid54 8635]

1 minute ago, Walden Ponderer said:

now that I'm east of the Mississippi

This is wholly and entirely your fault.

[HOOK'EMHOOAH 1610]

1 hour ago, Huckleberry said:

This is basically the only thing that bothers me about Connelly. His sensitivity and defiant defense of his system as though it is The Truth and others just don't get it. As a fellow producer of advanced ratings systems I think I'm qualified to note that they are inherently imperfect and those who produce them should be more aware of their imperfections than anyone. 

The fact that there is a subset of teams who share specific traits in style of play and whom your system struggles to get a handle on doesn't mean they play wrong like Connelly implies. It means your system doesn't handle them well. If that subset is small enough it's probably not worth it to try to change your system to capture them if it's intended to handle the entire team universe. Just say that instead of whining and getting defensive when they're pointed out. 

Then there's the fact that as proven this offseason, the entire time he's combatively defending his system against all questioning he's simultaneously researching changes and tweaks he can make to his system to improve it because he knows it's not perfect. After the 2018 season he changed his ratings to keep preseason data in all season long, he changed them to adjust ratings based on the conference a team is in. This last change is based on a completely arbitrary human grouping of teams that makes zero sense from an analytical perspective. He is using conference grouping as a proxy for team strength and may not even realize it. Essentially it is a post-processing corrective step that simply shifts good teams up and bad teams down. He made major changes, and they are definitely major, to his system immediately after spending 2018 defending his system as though it was perfect.

And for all of his arbitrary changes to the system, consider the fact that if Alabama and South Florida had completely swapped coaching staffs on February 15th, 2019 that his preseason ratings would not change. At all. Any reasonable person would understand and publicly recognize inherent weaknesses like that in their creation.

Just throwing shit against the wall here... why not instead build a system to explain why the results became what they were? Inmho, that would be more valuable and a far more acceptable system than trying to weigh one thing more than another. In 2016, when Kansas beat Texas, there were metrics that led to that conclusion. They happened in-game and might not be predicted before hand. When App State beat Michigan, could that be predicted before the result? If the data set is taken as a week on basis, then it gets more to the point and less about giving any specific team any kind of brand recognition and building the outcome from it.to me, that is far more applicable than weighing teams on some arbitrary and bullshit equation.

[Shut up Lou Holtz 2090]

Fuck the offseason. 

Fuck ou. 

Fuck it, Texas kicks off in 2 days so let’s fucking go

[HOOK'EMHOOAH 1610]

34 minutes ago, Zavala said:

 

how many days

To motherfukcing many.

[Tex Long 4509]

29 minutes ago, HOOK'EMHOOAH said:

Just throwing shit against the wall here... why not instead build a system to explain why the results became what they were? Inmho, that would be more valuable and a far more acceptable system than trying to weigh one thing more than another. In 2016, when Kansas beat Texas, there were metrics that led to that conclusion. They happened in-game and might not be predicted before hand. When App State beat Michigan, could that be predicted before the result? If the data set is taken as a week on basis, then it gets more to the point and less about giving any specific team any kind of brand recognition and building the outcome from it.to me, that is far more applicable than weighing teams on some arbitrary and bullshit equation.

In the  case of FBC, Any value at all would be more value. It's not that entertaining, and unless it rises above 52% accuracy it has no value - or rather, negative value - for wagering, at the 110/100 vig window. Plus, as many have stated, his apparent personal biases ruin both entertainment and financial values. 

[ShowmeyourTDs 1156]

11 hours ago, Patricio Swayze said:

Oh for fuck’s sake, Saturday can’t get here soon enough.

This post deserves more credit.  I can’t wait until Texas Football starts.

[Big D 249]

Anyone have the link to this college football schedule web site that has the tv listings included?

 

Thanks

[Machinator 32099]

5 minutes ago, Big D said:

Anyone have the link to this college football schedule web site that has the tv listings included?

 

Thanks

http://www.lsufootball.net/tvschedule.htm

[Big D 249]

Just now, Machinator said:

http://www.lsufootball.net/tvschedule.htm

Perfect.  Thanks 

[Machinator 32099]

 

[Junior Miller 1502]

14 minutes ago, Machinator said:

 

Texas 11 wins. Oklahoma 9 wins. Hurts is a fraud who can't throw. Defense will still be shitty. 

I can almost begin to smell the hitting in the air. Almost.