AFC South is going to be tough this season

[OILERMAN]

33 minutes ago, ctm said:

Colts appear to be in essentially the same situation as the Titans.  Defense will have to carry the team early on.

 

I'm not sure the Titans offense will be a problem early on, why?

[ctm]

47 minutes ago, OILERMAN said:

 

I'm not sure the Titans offense will be a problem early on, why?

 

A couple of potential issues that we don't know about at this point:

The OL both running and pass pro.

Tanny needing some time to get on the same page with all new receivers.

[StephenIsLegend]

1 hour ago, ctm said:

 

A couple of potential issues that we don't know about at this point:

The OL both running and pass pro.

Tanny needing some time to get on the same page with all new receivers.

Except by most accounts he looks good in camp. I can agree with the same page stuff, but I don’t think the offense struggles as much as others do early on. 

[ctm]

2 minutes ago, StephenIsLegend said:

Except by most accounts he looks good in camp. I can agree with the same page stuff, but I don’t think the offense struggles as much as others do early on. 

 

I hope you are right.  But I don't put a lot of faith in training camp 7 on 7 or even 11 on 11 without pads on.  Need to see more.

[Amy]

2 hours ago, ctm said:

 

A couple of potential issues that we don't know about at this point:

The OL both running and pass pro.

Tanny needing some time to get on the same page with all new receivers.

I do not think Tannehill getting on same page as his receivers will be an issue.  They have been practicing everyday together unlike last year when no one practiced.  I do hope Vrabel makes his decision on OL soon so they have time to gel together 

[Jamalisms]

 

 

 

[2ToneTerror]

3 hours ago, OILERMAN said:

 

I'm not sure the Titans offense will be a problem early on, why?

I think we need to see Tannehil get a few series in the next preseason game with the #1’s and we might be able to get a taste.

If Burks and Phillips get as open as they did in the first preseason game and Tannehill hits them, I’m not worried at all.

[Justafan]

34 minutes ago, 2ToneTerror said:

I think we need to see Tannehil get a few series in the next preseason game with the #1’s and we might be able to get a taste.

If Burks and Phillips get as open as they did in the first preseason game and Tannehill hits them, I’m not worried at all.

Agreed.  I think it was a mistake to not play starters at all last year in the preseason, especially when they had so little time to gel in camp as well.  You could tell how rusty they were early on.  Hopefully, they've learned.  I'm not asking for a half or even a quarter, just two series or so to let them knock some of the rust off.  

[Einstein]

17 hours ago, JT2000 said:

A "regression analysis " isn't appropriate here and I'm guessing that "Einstein " is counting on most of the board not knowing what a regression analysis is or what its limitations are.

 

For Einstein/logical:

https://blog.minitab.com/en/adventures-in-statistics-2/regression-analysis-tutorial-and-examples

 

 

If you need assistance understanding what a regression analysis is, what it does, or how it helps in a particular situation, i’m happy to help. Your original claim was “In a single game you have too much of a potential variable that drastically effects results”. In order to test that theory, you must conduct a study. But which? What model of analysis allows us to identify if a variable has an impact on a game, and likewise, whether a variable fluctuates too greatly to be of use to have an impact? The answer to that question is in your link: 

 

 

A regression analysis is not only appropriate for this situation, but it is the most appropriate, as there are very few other options to find the answer. One may arrive at the same conclusion via one or two other analysis models, but a regression analysis would be the simplest and cleanest method to determine if individual game variability impacts results or if it is a factor that can be ignored. Using your hypothesis as an example, the number of possessions (grouped) would be our independent variable, and the outcome of the game would be our dependent variable. This same experiment (and vice versa) could be regressioned across an entire season, leaving you with clean results on whether the fluctuating nature of the possession variable leads to any skewing in results both at the micro (1 game) level versus the macro (season) level.

 

I must confess that I am not the first to think of using a regression analysis in this manner, as it was the thesis of Stephen Bouzoanis that spawned the idea. It was titled: “Predicting the Outcome of NFL Games Using Logistic Regression. An excerpt from said thesis:

 

The following mathematical equation represents the way in which the probability of a victory is determined:
 Prob[𝑊𝑖𝑛] = 1 1+𝑒−𝑈
 where:
6

𝑈 = 𝛽0 + 𝛽1𝑥1 + 𝛽2𝑥2 + 𝛽3𝑥3 + ⋯
and 𝑥𝑖 are the various independent variables (team performance metrics) chosen for the model.
The model coefficients, 𝛽𝑖, i.e., the weights by which the independent variables affect
the outcome, are determined using Maximum Likelihood Estimation (MLE). For each variable
(performance metric), the corresponding coefficient/weight can be interpret as follows: All
else equal, each unit increase (decrease) in the variable 𝑥𝑖 will multiply (divide) the odds of
winning, i.e., Prob[𝑊𝑖𝑛] , by a factor of 𝑒𝛽𝑖 . Therefore, a positive (negative) coefficient, indicates a Prob[𝐿𝑜𝑠𝑠]
positive (negative) relationship between that variable (performance metric) and the odds of winning. For further explanation on logistic regression, see Fitzmaurice (2001).

 

In our situation, the performance metric (listed above) would be possessions and points per possession. If you need any further assistance, my university email is attached to my profile.

 

 

Edited August 14, 2022 by Einstein

[Einstein]

1 hour ago, 2ToneTerror said:

 

If Burks and Phillips get as open as they did in the first preseason game and Tannehill hits them, I’m not worried at all.

 

I think it would be wise to temper expectations on Burks second preseason game. It is important to remember that the majority of plays that Burks was open was a direct result of blown coverage by the opponent.

[OILERMAN]

It was a small sample size but when the Titans offense was healthy last year they played at a very high level. Unlike last year they are getting work together this camp. 

 

Henry seems healthy and explosive and the overall pass catching group seems improved over last year

 

We need to see more evidence of a good pass blocking OL

[9 Nines]

Da charges

Edited August 14, 2022 by 9 Nines

[AussieTitanFan08]

The Colts should rightfully be discussed as in the race to win the AFC South Division crown but its laughable how over the top the NFL media has been about them being a real threat to win the Super Bowl.

 

 

[dcblanton]

1 hour ago, AussieTitanFan08 said:

The Colts should rightfully be discussed as in the race to win the AFC South Division crown but its laughable how over the top the NFL media has been about them being a real threat to win the Super Bowl.

 

 

Damn.  Admittedly, I had forgotten all about the Texans being good enough to win the division for a few years there.  I knew Jax had done it and got to the AFCCG.  I knew that the Titans had won the last couple.  I assumed the Colts had the other ones and it was the Texans.  

 

I guess the Colts are like the AFC South's Chargers.  The Chargers are always picked to win the division.  I understand the frustration now.

[JT2000]

8 hours ago, Einstein said:

 

If you need assistance understanding what a regression analysis is, what it does, or how it helps in a particular situation, i’m happy to help. Your original claim was “In a single game you have too much of a potential variable that drastically effects results”. In order to test that theory, you must conduct a study. But which? What model of analysis allows us to identify if a variable has an impact on a game, and likewise, whether a variable fluctuates too greatly to be of use to have an impact? The answer to that question is in your link: 

 

 

A regression analysis is not only appropriate for this situation, but it is the most appropriate, as there are very few other options to find the answer. One may arrive at the same conclusion via one or two other analysis models, but a regression analysis would be the simplest and cleanest method to determine if individual game variability impacts results or if it is a factor that can be ignored. Using your hypothesis as an example, the number of possessions (grouped) would be our independent variable, and the outcome of the game would be our dependent variable. This same experiment (and vice versa) could be regressioned across an entire season, leaving you with clean results on whether the fluctuating nature of the possession variable leads to any skewing in results both at the micro (1 game) level versus the macro (season) level.

 

I must confess that I am not the first to think of using a regression analysis in this manner, as it was the thesis of Stephen Bouzoanis that spawned the idea. It was titled: “Predicting the Outcome of NFL Games Using Logistic Regression. An excerpt from said thesis:

 

The following mathematical equation represents the way in which the probability of a victory is determined:
 Prob[𝑊𝑖𝑛] = 1 1+𝑒−𝑈
 where:
6

𝑈 = 𝛽0 + 𝛽1𝑥1 + 𝛽2𝑥2 + 𝛽3𝑥3 + ⋯
and 𝑥𝑖 are the various independent variables (team performance metrics) chosen for the model.
The model coefficients, 𝛽𝑖, i.e., the weights by which the independent variables affect
the outcome, are determined using Maximum Likelihood Estimation (MLE). For each variable
(performance metric), the corresponding coefficient/weight can be interpret as follows: All
else equal, each unit increase (decrease) in the variable 𝑥𝑖 will multiply (divide) the odds of
winning, i.e., Prob[𝑊𝑖𝑛] , by a factor of 𝑒𝛽𝑖 . Therefore, a positive (negative) coefficient, indicates a Prob[𝐿𝑜𝑠𝑠]
positive (negative) relationship between that variable (performance metric) and the odds of winning. For further explanation on logistic regression, see Fitzmaurice (2001).

 

In our situation, the performance metric (listed above) would be possessions and points per possession. If you need any further assistance, my university email is attached to my profile.

 

 

I have your alt on ignore for a reason. You obviously don't use these tools in work - I do.

 

You just said a bunch of crap that didn't address why you can't set up a proper regression analysis. 

 

You go ahead and set it up and show us and it will be obvious why you have to force things to make the regression fit.