10 things I think Week 1 edition

[jterrell]

1. DAL mgmt won the Zeke holdout. Zeke won too for sure but the funny money aside, Zeke's cap number just dropped and only 28M is fully GTD. Everything else is conditional. Zeke got GTD money but he only walked away with 7.5M this week. So year 1 total money was neutral.

2. Computer simulations hate DAL. LOL Seriously what is with these weird 6-8 win projections. Computers are impartial so what's the deal? I am going to guess it's largely based on the computers not factoring in how much better DAL was once they added Amari and how much better they are with Travis and Witten returning. Plus draft-wise minus a #1 DAL likely gets no rookie bump from computer yet Pollard and Wilson will help this team win games this year. Removing emotion is great for gaining accuracy and clarity but you still need some rational understanding of the game that computers lack.

3. Dallas contracts are almost tear-worthy. These long extensions are so well done because they maintained the cheaper front end and allowed restructures later after the new CBA is announced. It gives the team tons of flexibility without leveraging themselves heavily. Stephen Jones is a bean counter extraordinaire. He is ironically the perfect balance to Wildcatting life of the party Jerry Jones. Pretty easy to see a way forward now even when Dak gets huge money.

4. Dallas starts 3-0. I make zero bones about being a homer. This is my team of all teams. Across all sports and all times. I am not even pretending to be unbiased BUT I do try to maintain rationality. This team starts versus 3 of the arguably 5 worst teams in the NFL. WAS should field a very competitive defense but no offense. NYG and MIA simply look to be full on tanking/rebuilding. DAL IS a top 5 team in talent. Hard to imagine losing any of the first 3 even for a DAL team that generally starts very slowly.

5. Kellen Moore's influence will be subtle but it will also show up in Dak's stats. He will do a lot more to get Dak going. Linehan was dine with an Ace WR but not very good without it because he didn't do much of the little things to get the QB hot. Moore will do that because he's very aware of what a QB needs to succeed. Dak will provide fantasy owners some love this year.

6. Tony Pollard will contribute a lot. He is going to end up with the role they created for Tavon Austin. The speed element, game breaker. Not sure when we start to see him in the game same time as Zeke but it will eventually happen. Expect him to get those Kamara/McCaffrey type looks.

7. Amari with 1 year playing for all the money in the world? I like it. Cooper is special but he's also been inconsistent. This is the most motivation a player can ever have. Super Bowl contender with millions upon millions personally on the line. You will get all Amari Cooper has to give. And he broke an Eagle last year....

8. Donovan Wilson will play a lot this year. That kid just knows how to play. He should supplant Heath and finally give DAL the safety corps fans have been craving for about a decade now.

9. I'm all aboard the Super Bowl train here.No brakes. So what keeps me up at night? Injuries. DAL has no Amari replacement. They have no Dak replacement. They have no DLaw replacement. Those 3 guys need to stay on the field. Even at 75% they are better than their back ups.

10. Garrett is on the final year of his deal and I think he may want to leave. He could take a year off and then take a high profile college job back in the Northeast. Life would be more peaceful and he'd have more time with family. He could do more things his Dad did. I have a sneaking suspicion winning it all might be all the push he needs to close shop.

 

[Sydla]

2. Computer simulations hate DAL. LOL Seriously what is with these weird 6-8 win projections. Computers are impartial so what's the deal? I am going to guess it's largely based on the comptuers not factoring in how much better DAL was once they added Amari and how much better they are with Travis and Witten returning. Plus draft-wise minus a #1 DAL likely gets no rookie bump from computer yet Pollard and Wilson will help this team win games this year. Removing emotion is great for gaining accuracy and clarity but you still need some rational understanding of the game that computers lack.

On your #2, the computers take into account things like point differential, etc. And the reality is with our point differential last year, we outstripped what our win total historically would have been with only a +15 (regular season) point differential. It's also wrong to assume that with Cooper that would change. Prior to Cooper, the team had a +17 point differential. In games Cooper played, the point differential was actually -2. And if you want to adjust for the shut out to Indy, the point differential would still only be +21, so there wasn't that much a difference pre and post Cooper in point differential, for example.

And there should be no excuse for not starting 3-0.

 

[John813]

I definitely could see Wilson playing safety near full time by the bye week.

If the PS Wilson carries over to the regular season they found a starter from day 3.

 

[jazzcat22]

Great stuff and spot on.
I am pumped, let's get to Sunday 3:25 CST.

 

[jazzcat22]

On your #2, the computers take into account things like point differential, etc. And the reality is with our point differential last year, we outstripped what our win total historically would have been with only a +15 (regular season) point differential. It's also wrong to assume that with Cooper that would change. Prior to Cooper, the team had a +17 point differential. In games Cooper played, the point differential was actually -2. And if you want to adjust for the shut out to Indy, the point differential would still only be +21, so there wasn't that much a difference pre and post Cooper in point differential, for example.

And there should be no excuse for not starting 3-0.

But it still does not make the computers right. With these computer predictions, it is also garbage in / garbage out. I can write a program that shows a different outcome.
And it does not take into account human emotions, how pumped a team gets or does not get for a game. The drive and will for players to overcome and the desire to win.

They can create all the models they want. It is good or bad for the fans to like it or not. But I look at them as I do power rankings or player rankings. It is nice if you are at the top, not so nice if you are at the bottom. But it really means the same as looking at the toilet flush.

Agree, should start 3-0 regardless.

 

[morasp]

I'm hoping you're right on Wilson and Pollard. Getting contributors on day three is hard to do.

 

[cowboy_ron]

1. DAL mgmt won the Zeke holdout. Zeke won too for sure but the funny money aside, Zeke's cap number just dropped and only 28M is fully GTD. Everything else is conditional. Zeke got GTD money but he only walked away with 7.5M this week. So year 1 total money was neutral.

2. Computer simulations hate DAL. LOL Seriously what is with these weird 6-8 win projections. Computers are impartial so what's the deal? I am going to guess it's largely based on the comptuers not factoring in how much better DAL was once they added Amari and how much better they are with Travis and Witten returning. Plus draft-wise minus a #1 DAL likely gets no rookie bump from computer yet Pollard and Wilson will help this team win games this year. Removing emotion is great for gaining accuracy and clarity but you still need some rational understanding of the game that computers lack.

3. Dallas contracts are almost tear-worthy. These long extensions are so well done because they maintained the cheaper front end and allowed restructures later after the new CBA is announced. It gives the team tons of flexibility without leveraging themselves heavily. Stephen Jones is a bean counter extraordinaire. He is ironically the perfect balance to Wildcatting life of the party Jerry Jones. Pretty easy to see a way forward now even when Dak gets huge money.

4. Dallas starts 3-0. I make zero bones about being a homer. This is my team of all teams. Across all sports and all times. I am not even pretending to be unbiased BUT I do try to maintain rationality. This team starts versus 3 of the arguably 5 worst teams in the NFL. WAS should field a very competitive defense but no offense. NYG and MIA simply look to be full on tanking/rebuilding. DAL IS a top 5 team in talent. Hard to imagine losing any of the first 3 even for a DAL team that generally starts very slowly.

5. Kellen Moore's influence will be subtle but it will also show up in Dak's stats. He will do a lot more to get Dak going. Linehan was dine with an Ace WR but not very good without it because he didn;t do much of the little things to get the QB hot. Moore will do that becuse he's very aware of what a QB needs to succeed. Dak will provide fantasy owners some love this year.

6. Tony Pollard will contribute a lot. He is going to end up with the role they created for Tavon Austin. The speed element, game breaker. Not sure when we start to see him in the game same time as Zeke but it will eventually happen. Expect him to get those Kamara/McCaffrey type looks.

7. Amari with 1 year playing for all the mioney in the world? I like it. Cooper is special but he's also been inconsistent. This is the most motivation a player can ever have. Super Bowl contender with millions upon milluions personally on the line. You will get all Amari Cooper has to give. And he broke an Eagle last year....

8. Donovan Wilson will play a lot this year. That kid just knows how to play. He should supplant Heath and finally give DAL the safety corps fans have been craving for about a decade now.

9. I'm all aboard the Super Boiwl train here.No brakes. So what keeps me up at night? Injuries. DAL has no Amari replacement. They have no Dak replacement. They have no DLaw replacement. Those 3 guys need to atay on the field. Even at 75% they are better than their back ups.

10. Garrett is on the final year of his deal and I think he may want to leave. He could take a year off and then take a high profile college job back in the Northeast. Life would be more peaceful and he'd have more time with family. He could do more things his Dad did. I have a sneaking suspicion winning it all might be all the push he needs to close shop.

Nice post

 

[Sydla]

But it still does not make the computers right. With these computer predictions, it is also garbage in / garbage out. I can write a program that shows a different outcome.
And it does not take into account human emotions, how pumped a team gets or does not get for a game. The drive and will for players to overcome and the desire to win.

They can create all the models they want. It is good or bad for the fans to like it or not. But I look at them as I do power rankings or player rankings. It is nice if you are at the top, not so nice if you are at the bottom. But it really means the same as looking at the toilet flush.

Agree, should start 3-0 regardless.

Sure, it doesn't mean the computers are or would be right. But if the excuse is that the computers aren't taking into account Cooper coming on midway through the year, that's a false premise because many of the stats were largely unchanged when Cooper came on board, such as point differential as I showed you.

The reality is that with some of these stats and our ridiculously good record in one score games last year, it should surprise no one who is paying attention that we'd be a prime candidate for regression. Doesn't mean it will definitely happen but this team can't expect to win 10 games in 2019 without greatly improving our point differential and expecting we'll just kick *** in one score games again.

 

[Stash]

Sure, it doesn't mean the computers are or would be right. But if the excuse is that the computers aren't taking into account Cooper coming on midway through the year, that's a false premise because many of the stats were largely unchanged when Cooper came on board, such as point differential as I showed you.

The reality is that with some of these stats and our ridiculously good record in one score games last year, it should surprise no one who is paying attention that we'd be a prime candidate for regression. Doesn't mean it will definitely happen but this team can't expect to win 10 games in 2019 without greatly improving our point differential and expecting we'll just kick *** in one score games again.

That's my issue with that take. Why does anyone assume they will be "one score games" at the same ratio?

 

[big dog cowboy]

10 for 10 JT!

Great post.

 

[Sydla]

That's my issue with that take. Why does anyone assume they will be "one score games" at the same ratio?

I am not sure the argument is necessarily that there will be the same number of one score games. But the point is they were way above the mean in winning one score games last year which means there was probably some luck involved in them winning 10 games.

For all we know, they may be so good they only have 3 one score games this year. But if they do find themselves in a bunch of one score games like last year, assuming they will win those games at an 80% clip probably isn't wise.

 

[jazzcat22]

I am not sure the argument is necessarily that there will be the same number of one score games. But the point is they were way above the mean in winning one score games last year which means there was probably some luck involved in them winning 10 games.

For all we know, they may be so good they only have 3 one score games this year. But if they do find themselves in a bunch of one score games like last year, assuming they will win those games at an 80% clip probably isn't wise.

Many games are decided by one score or less. So if a team is lucky to win a close game, then why do fans whine to lose a game by less than one scored because they then are unlucky.
So this computer program only takes into account for Dallas' 17 point differential but not other teams?

 

[Sydla]

Many games are decided by one score or less. So if a team is lucky to win a close game, then why do fans whine to lose a game by less than one scored because they then are unlucky.
So this computer program only takes into account for Dallas' 17 point differential but not other teams?

No, the computer doesn't just take into account Dallas' point differential.

I think what's at play here is you are purposely trying to be obtuse on these computer models because you simply don't like that they aren't fawning over Dallas.

All I am trying to do is explain why the models say what they say. It doesn't mean the models will be right, they get things wrong. But arguing that the models are based on, I guess, faulty data isn't an accurate statement.

 

[jazzcat22]

No, the computer doesn't just take into account Dallas' point differential.

I think what's at play here is you are purposely trying to be obtuse on these computer models because you simply don't like that they aren't fawning over Dallas.

All I am trying to do is explain why the models say what they say. It doesn't mean the models will be right, they get things wrong. But arguing that the models are based on, I guess, faulty data isn't an accurate statement.

Not saying faulty data, And I know what you are getting at.
I can care less what the computer says. They can say we will win the SB and I will still say so what.

I just wonder what criteria they use as it makes no sense. You have a team that everything is on the upward trail, and then to come up with they will miss the playoffs, but other teams in the same situation will make the playoffs as they did nothing any different to improve than Dallas had. Just doesn't fly IMO.
You have your opinion, but this does not change mine.

 

[Stash]

I am not sure the argument is necessarily that there will be the same number of one score games. But the point is they were way above the mean in winning one score games last year which means there was probably some luck involved in them winning 10 games.

'Bad luck' in fact, given that they were without their Pro Bowl center, were breaking in a rookie guard, and were among the worst of the worst in red zone scoring. None of which will be the case for 2019. That's my point anyway.

For all we know, they may be so good they only have 3 one score games this year. But if they do find themselves in a bunch of one score games like last year, assuming they will win those games at an 80% clip probably isn't wise.

No, if they're no better than where they were last year, their 'luck' is sure to run out. But again, I think that's the flaw in this logic and where these forecasts are failing.

 

[Fletch]

10 for 10 JT!

Great post.

Brown-noser!

 

[xwalker]

On your #2, the computers take into account things like point differential, etc. And the reality is with our point differential last year, we outstripped what our win total historically would have been with only a +15 (regular season) point differential. It's also wrong to assume that with Cooper that would change. Prior to Cooper, the team had a +17 point differential. In games Cooper played, the point differential was actually -2. And if you want to adjust for the shut out to Indy, the point differential would still only be +21, so there wasn't that much a difference pre and post Cooper in point differential, for example.

And there should be no excuse for not starting 3-0.

Computer simulations of NFL seasons are nonsense.

Teams change significantly from season to season.

The Cowboys change and their opponents change.

The modern NFL had become very match-up sensitive.

The Cowboys defense was at a huge match-up disadvantage to the Rams offense.

The Cowboys held the Saints to 10 points but couldn't stop the Rams. That's a match-up issue.

Yes, the Rams rolled out some new tweaks for the playoffs with some specifically aimed at the Cowboys;
however, it still fit perfectly into their basic run blocking scheme and ability to use the pass to setup the run.

Opponents change. The actual teams the Cowboys play change but also all opponents themselves change.
If an opponent gets better or worse at QB that is huge change (see Colts).

There is no way to really incorporate the changes for a new season into simulations.

Maybe they can re-run Colts simulations without Luck, but there is no way to really quantify all changes to an NFL team over the off-season.

I'm not saying this from a "Those new newfangled computers are wrong" type of perspective. I design computer systems for a living and I have Math & Engineering degrees; therefore, I understand the statistics and modeling concepts involved.

If it could be done with significant accuracy, it wouldn't be shared with the public. The people running the simulations would be racking up millions by betting on games.

 

[jazzcat22]

Computer simulations of NFL seasons are nonsense.

Teams change significantly from season to season.

The Cowboys change and their opponents change.

The modern NFL had become very match-up sensitive.

The Cowboys defense was at a huge match-up disadvantage to the Rams offense.

The Cowboys held the Saints to 10 points but couldn't stop the Rams. That's a match-up issue.

Yes, the Rams rolled out some new tweaks for the playoffs with some specifically aimed at the Cowboys;
however, it still fit perfectly into their basic run blocking scheme and ability to use the pass to setup the run.

Opponents change. The actual teams the Cowboys play change but also all opponents themselves change.
If an opponent gets better or worse at QB that is huge change (see Colts).

There is no way to really incorporate the changes for a new season into simulations.

Maybe they can re-run Colts simulations without Luck, but there is no way to really quantify all changes to an NFL team over the off-season.

I'm not saying this from a "Those new newfangled computers are wrong" type of perspective. I design computer systems for a living and I have Math & Engineering degrees; therefore, I understand the statistics and modeling concepts involved.

If it could be done with significant accuracy, it wouldn't be shared with the public. The people running the simulations would be racking up millions by betting on games.

I have Engineering degrees as well. And have written computer programs, not just for work, but for myself as well.
I also taught other Designers, Engineers, Drafters on different computer systems and CAD systems. One thing I stressed to them. Do not take for granted what the computer tells you or shows you. Use your common sense as you must have an idea of what the results are before it gives you an answer. If someone thing seems off, then check into why and prove if it is right or wrong.

Unfortunately I came across many new college graduates coming to work, they were adamant the computer is correct always. When if fact that is not the case. Even though many times it was user input error. Or just not knowing the best way. As I then had to show them why or why not. Not often, but it happened.

 

[Sydla]

Not saying faulty data, And I know what you are getting at.
I can care less what the computer says. They can say we will win the SB and I will still say so what.

I just wonder what criteria they use as it makes no sense. You have a team that everything is on the upward trail, and then to come up with they will miss the playoffs, but other teams in the same situation will make the playoffs as they did nothing any different to improve than Dallas had. Just doesn't fly IMO.
You have your opinion, but this does not change mine.

The criteria they use is historical data. It's simply a computer program. They can't account for things like "well we lost Gregory but signed Quinn" that fans try to apply

They simply take data, crunch it and spit out results. That's all it is. Complaining about it and how it doesn't account for this or that just shows that people aren't understanding what these models are doing.

Also note, nowhere did I say anything about their accuracy to predict. It's not exactly great. But I chuckle when people ***** about a computer model and how it didn't account for this change or the hiring of this coach or whatever. That's not what these models are built for.

 

[Outlaw Heroes]

Interesting thoughts on Garrett, OP. Super speculative. But interesting.