Bills Teammates Rave about Tyrod Taylor

[LB3]

The thing about the mean and median is that they're two different statistical tools and in order to analyze data, understanding the difference is critical. I'm really not sure how else to explain it to you.

 

Here are some links you may find helpful and hopefully will do a better job of explaining mean and median than I have:

 

https://www.udacity.com/course/intro-to-statistics--st101

http://writing.colostate.edu/guides/guide.cfm?guideid=67

http://www.pixyland.org/peterpan/

:lol:

[Hapless Bills Fan]

Guys, I understand what the median and the mean are. I really do. I just think the term is invoked to here as a stand-in for actual analysis. It's the equivalent hand-waving ( http://en.wiktionary.org/wiki/handwaving) , in my opinion, and the use of it here doesn't make sense to me.

 

Anyway, please tell me this: why would anyone dispense with performances in games that matter more than others (divisional games) against a quality defense by either cutting them out altogether or claiming that they're skewed results even though they're 1/6 of a small sample size?

 

Ha, ha! Astute, Dave. Actually, somehow my computer dumped a long post with an analysis I did of Orton's actual statistics, at least 2 that I think matter in individual games (completion % and YPA).

Somehow it came through as a single sentence defining them and none of the actual analysis.

 

I don't feel like retyping it all. The bottom line is there is absolutely no justification whatsoever in completion percentage for tossing out either Jets game - the median and the mean differ by less than a percentage point and are not significantly impacted by deleting the Jets games.

 

One of the two Jets games has a clear outlier in YPA - a number 2x the mean. The mean and the median differ, and when that number is tossed out, the mean shifts while the median doesn't change. But there's nothing in the other Jets game to indicate an outlier. It's above the median, but 3 numbers are higher and one very close. And when we look "under the hood" at that game, we see Orton only made 17 pass attempts, one of which broke a short pass for 66 yds when the Jets lost contain on Watkins. So that single 66 yd pass play on a low number of attempts skewed the YPA number despite an entirely pedestrian completion percentage of 58% that game.

 

Rating in both Jets games is unusually high, but most people don't seem to put a lot of stock in rating as a predictive metric for winning or as a good measure of QB performance so meh.

 

Bottom line, I agree with you that in this case, the term "embrace the median" is invoked as a stand-in for actual analysis and amounts to hand-waving. In my opinion, and I am a scientist by profession and use statistics extensively, neither the use of it here nor excluding the two Jets games from analysis of Orton make any sense to me.

Edited June 5, 2015 by Hopeful

[dave mcbride]

 

Ha, ha! Astute, Dave. Actually, somehow my computer dumped a long post with an analysis I did of Orton's actual statistics, at least 2 that I think matter in individual games (completion % and YPA).

Somehow it came through as a single sentence defining them and none of the actual analysis.

 

I don't feel like retyping it all. The bottom line is there is absolutely no justification whatsoever in completion percentage for tossing out either Jets game - the median and the mean differ by less than a percentage point and are not significantly impacted by deleting the Jets games.

 

One of the two Jets games has a clear outlier in YPA - a number 2x the mean. The mean and the median differ, and when that number is tossed out, the mean shifts while the median doesn't change. But there's nothing in the other Jets game to indicate an outlier. It's above the median, but 3 numbers are higher and one very close. And when we look "under the hood" at that game, we see Orton only made 17 pass attempts, one of which broke a short pass for 66 yds when the Jets lost contain on Watkins. So that single 66 yd pass play on a low number of attempts skewed the YPA number despite an entirely pedestrian completion percentage of 58% that game.

 

Rating in both Jets games is unusually high, but most people don't seem to put a lot of stock in rating as a predictive metric for winning or as a good measure of QB performance so meh.

 

Bottom line, I agree with you that in this case, the term "embrace the median" is invoked as a stand-in for actual analysis and amounts to hand-waving. In my opinion, and I am a scientist by profession and use statistics extensively, neither the use of it here nor excluding the two Jets games from analysis of Orton make any sense to me.

Thanks!

[KDIGGZ]

i think TT is going to be a great practice and preseason QB so i can understand the early hype. when the plays break down he can improvise and he will be more exciting than the other options. but when teams start game planning for him and do what they have to in order to keep him in the pocket and force him to make the throws i think we will see he is not an accurate passer. we have too many offensive weapons. we just have to get the ball in their hands and let them make the plays. we don't need our QB to be running around trying to do too much

[dave mcbride]

The thing about the mean and median is that they're two different statistical tools and in order to analyze data, understanding the difference is critical. I'm really not sure how else to explain it to you.

 

Here are some links you may find helpful and hopefully will do a better job of explaining mean and median than I have:

 

https://www.udacity.com/course/intro-to-statistics--st101

http://writing.colostate.edu/guides/guide.cfm?guideid=67

http://www.pixyland.org/peterpan/

Um ... I do understand the difference, and I understand why it matters. I've understood it for decades. I just think that the concept of the median is being used here to deflect attention from a weak-to-bordering-on-non-analysis of Orton. No one but Hopeful is actually dealing with the facts at hand ...

Edited June 5, 2015 by dave mcbride

[Jauronimo]

Um ... I do understand the difference, and I understand why it matters. I've understood it for decades. I just think that the concept of the median is being used here to deflect attention from a weak-to-bordering-on-non-analysis of Orton. No one but Hopeful is actually dealing with the facts at hand ...

I don't know what else to tell you. Did you explore the links I provided? There is an important distinction between mean and median (we'll get to mode later) and I really can't explain it any better than I have.

 

http://en.wikipedia.org/wiki/Median

http://en.wikipedia.org/wiki/Mean

 

You may not agree, but its a basic fact. If they were the same, why would there be two unique wikipedia entries? Can we at least agree on that?

[Hapless Bills Fan]

i think TT is going to be a great practice and preseason QB so i can understand the early hype. when the plays break down he can improvise and he will be more exciting than the other options. but when teams start game planning for him and do what they have to in order to keep him in the pocket and force him to make the throws i think we will see he is not an accurate passer. we have too many offensive weapons. we just have to get the ball in their hands and let them make the plays. we don't need our QB to be running around trying to do too much

 

We'll see. I do hope Roman and Rex are not intending to get all cutsy swapping QB in and out with him wildcat-style.

 

His track record in college was that his completion percentage and passing yardage went up each year.

His one game, EJ's 14 games, and Matt Cassel lifetime are all within spitting distance of each other for completion percentage. (58.6, 58.6 ,59).

Right now I would have to say that's the "?" for both EJ and TT, can they stand in the pocket and make the throws accurately?

[John from Riverside]

 

We'll see. I do hope Roman and Rex are not intending to get all cutsy swapping QB in and out with him wildcat-style.

 

His track record in college was that his completion percentage and passing yardage went up each year.

His one game, EJ's 14 games, and Matt Cassel lifetime are all within spitting distance of each other for completion percentage. (58.6, 58.6 ,59).

Right now I would have to say that's the "?" for both EJ and TT, can they stand in the pocket and make the throws accurately?

What I want to know is if Taylor trusts what he throws and actually releases the ball.......this is what seems to be the only (but major) hangup for EJ so far.

 

It is said that TT throws a really nice long ball......but does how does he do on short to intermediate stuff? Does his lack of height limit him?

[Hapless Bills Fan]

What I want to know is if Taylor trusts what he throws and actually releases the ball.......this is what seems to be the only (but major) hangup for EJ so far.

 

It is said that TT throws a really nice long ball......but does how does he do on short to intermediate stuff? Does his lack of height limit him?

 

Throws I've seen from TT are rainbows. Pretty to watch but not good for the Ralph in December.

[Gugny]

What I want to know is if Taylor trusts what he throws and actually releases the ball.......this is what seems to be the only (but major) hangup for EJ so far.

 

It is said that TT throws a really nice long ball......but does how does he do on short to intermediate stuff? Does his lack of height limit him?

 

I don't know why you need to come here and bash EJ every chance you get. We get it. You don't like him. But enough is enough.

[Kirby Jackson]

i think TT is going to be a great practice and preseason QB so i can understand the early hype. when the plays break down he can improvise and he will be more exciting than the other options. but when teams start game planning for him and do what they have to in order to keep him in the pocket and force him to make the throws i think we will see he is not an accurate passer. we have too many offensive weapons. we just have to get the ball in their hands and let them make the plays. we don't need our QB to be running around trying to do too much

So you are a Cassel guy? You can always tell where someone stands based on their interpretation of what's needed. Accuracy, play in the pocket, good decision making are the same qualities that I am looking for. For me at the moment Cassel is the safest for those reasons too. If someone else develops and can do similar things AND make more plays I am all for it.

 

We'll see. I do hope Roman and Rex are not intending to get all cutsy swapping QB in and out with him wildcat-style.

 

His track record in college was that his completion percentage and passing yardage went up each year.

His one game, EJ's 14 games, and Matt Cassel lifetime are all within spitting distance of each other for completion percentage. (58.6, 58.6 ,59).

Right now I would have to say that's the "?" for both EJ and TT, can they stand in the pocket and make the throws accurately?

Nice post and it makes sense. That is the 8,000,000,000 question.

[dave mcbride]

I don't know what else to tell you. Did you explore the links I provided? There is an important distinction between mean and median (we'll get to mode later) and I really can't explain it any better than I have.

 

http://en.wikipedia.org/wiki/Median

http://en.wikipedia.org/wiki/Mean

 

You may not agree, but its a basic fact. If they were the same, why would there be two unique wikipedia entries? Can we at least agree on that?

I think we're talking past each other. I know what the terms mean and what they imply but I honestly don't care. My issue is with discounting the games to make a (cherry-picked) point about Orton's below average-ness despite the fact the opponent in said games was good competition. It's one of those "he doesn't have that much power once you subtract all the home runs hit on Tuesdays and Thursdays" sorts of arguments. I have no time for those. To reiterate, I really don't care about the terminological issue that you seem to be concerned about. I regard it as a MacGuffin in this whole debate. With regard to Orton, I honestly don't know where you stand. I do know where you stand on the importance of the differences between medians and means, but that's irrelevant to me. What point are you yourself trying to make about Orton's performance, if any?

Edited June 5, 2015 by dave mcbride

[KDIGGZ]

So you are a Cassel guy? You can always tell where someone stands based on their interpretation of what's needed. Accuracy, play in the pocket, good decision making are the same qualities that I am looking for. For me at the moment Cassel is the safest for those reasons too. If someone else develops and can do similar things AND make more plays I am all for it.

Nice post and it makes sense. That is the 8,000,000,000 question.

I just feel like we have so many weapons on this team, why go with a QB that tries to do too much? 3-steps and throw and get it in the hands of the real playmakers. Be safe with the ball, be accurate, make good decisions. From what little I have seen from the guys so far, I think this will be Cassel. He gets the ball out on time and makes the right decisions. If we can get 3-4 yards on 1st and 2nd down and don't ask him to do too much by puting ourselves in 3rd and long situations then I think he can be successful.

[billykaykay]

Like Tyrod Taylor, seems like a very good #2 option. So does Cassel. Simms seems like a good camp arm.

 

The NFL is a business and it would be shocking if the Bills cut a former 1st round pick. EJ has a great attitude by all accounts, is a team guy, has shown the ability to learn and adapt while also having the size, arm strength and mobility ideal for the position. All of that coupled with the fact they would have to pay him a lot of money to hit the street just seems like a drastic move that wouldn't fit.

 

I'm not sure what EJ did or didn't do for some of the people that are so strongly against him but he is the front runner whether people like it or not and he isn't going anywhere in all likelihood. I hope he succeeds and gives the Bills a QB of the future.

Is Simms Practice Squad eligible ?

[Jauronimo]

I think we're talking past each other. I know what the terms mean and what they imply but I honestly don't care. My issue is with discounting the games to make a (cherry-picked) point about Orton's below average-ness despite the fact the opponent in said games was good competition. It's one of those "he doesn't have that much power once you subtract all the home runs hit on Tuesdays and Thursdays" sorts of arguments. I have no time for those. To reiterate, I really don't care about the terminological issue that you seem to be concerned about. I regard it as a MacGuffin in this whole debate. With regard to Orton, I honestly don't know where you stand. I do know where you stand on the importance of the differences between medians and means, but that's irrelevant to me. What point are you yourself trying to make about Orton's performance, if any?

I don't know how you can say that when MacGuffin himself, the godfather of modern statistics, declares unequivocally that median and mean are not the same. Understanding the difference between mean, median, mode, and range is paramount to deeper statistical analysis. I know that you're reluctant to accept my explanation of the differences, subtle as they are, but you should find the links I provided to be both credible and accessible for novices and experts alike. Its not my intention to be rude or difficult, but I cannot in good conscience just agree to disagree on this matter. I don't know what else to say.

[Kirby Jackson]

I just feel like we have so many weapons on this team, why go with a QB that tries to do too much? 3-steps and throw and get it in the hands of the real playmakers. Be safe with the ball, be accurate, make good decisions. From what little I have seen from the guys so far, I think this will be Cassel. He gets the ball out on time and makes the right decisions. If we can get 3-4 yards on 1st and 2nd down and don't ask him to do too much by puting ourselves in 3rd and long situations then I think he can be successful.

Nice thoughts and very similar to where I stand. I think that roster makeup has a lot to do with it as well. If the Bills didn't have a lot of playmakers than they would certainly need more plays to come from the QBs. I would think that both EJ & TT are capable of making more plays. This roster though needs the best game manager (mean that term in a good way). I don't care who it is as long as it is the guy that is the best at what you mentioned above.

[San-O]

He's thrown very few NFL passes, so it's hard to tell from live game stats. He might be great!!!

http://espn.go.com/nfl/player/stats/_/id/14163/tyrod-taylor

[dave mcbride]

I don't know how you can say that when MacGuffin himself, the godfather of modern statistics, declares unequivocally that median and mean are not the same. Understanding the difference between mean, median, mode, and range is paramount to deeper statistical analysis. I know that you're reluctant to accept my explanation of the differences, subtle as they are, but you should find the links I provided to be both credible and accessible for novices and experts alike. Its not my intention to be rude or difficult, but I cannot in good conscience just agree to disagree on this matter. I don't know what else to say.

I now think you're just trying to yank my chain. Is this some sort of elaborate inside joke about Holcomb's/Edward's/Orton's Arm and regression to the mean?

 

Anyway, I'm not disagreeing with you; I just don't care. There's a big difference between not caring and agreeing to disagree -- sorta like the difference between the median and the mean.

Edited June 5, 2015 by dave mcbride

[KDIGGZ]

Nice thoughts and very similar to where I stand. I think that roster makeup has a lot to do with it as well. If the Bills didn't have a lot of playmakers than they would certainly need more plays to come from the QBs. I would think that both EJ & TT are capable of making more plays. This roster though needs the best game manager (mean that term in a good way). I don't care who it is as long as it is the guy that is the best at what you mentioned above.

glad we are on the same page. hopefully the team feels the same way

[Hapless Bills Fan]

I just feel like we have so many weapons on this team, why go with a QB that tries to do too much? 3-steps and throw and get it in the hands of the real playmakers. Be safe with the ball, be accurate, make good decisions. From what little I have seen from the guys so far, I think this will be Cassel. He gets the ball out on time and makes the right decisions. If we can get 3-4 yards on 1st and 2nd down and don't ask him to do too much by puting ourselves in 3rd and long situations then I think he can be successful.

 

Well, the counterpoint to that is that we kind of have a 10 year track record on Cassel. He may think his best football is ahead of him, typically by the time a guy is in his 30s and has been in the league 10 years he is what he is ... I can't think of an exception besides Gannon.

 

So I think Cassel is the fall back "safety net" guy, but if either EJ or Tyrod seem to have similar skills right now, we'll ride with whoever looks best because the young guys just might have a higher ceiling.