Late to the party Bedard on 98.5

[JMC00]

Balt Pitt

Baltimore's only quality win was against the Browns week 1.

Pitt lost at home to Tampa and was out played by the jests.

Sorry if I'm not buying in on Pitt and Baltimore as "good" teams

 

[Ring 6]

No it isn't. DVOA is a totally different animal.

Not really, it is just a formula based on those cumulative stats, breaking them down per play, without correcting the flaws that exist in cumulative stats.

 

[Oswlek]

Not really, it is just a formula based on those cumulative stats, breaking them down per play, without correcting the flaws that exist in cumulative stats.

No, it isn't. Seriously, Andy, and I mean this the most genial of intentions, you need to educate yourself about the mechanics of DVOA.

As for Balt and Pitt, as I said before, even if I concede those, you have to turn around and say that neither Miami and Buffalo are good teams. And since when did "good" become such a positive adjective? Virtually every "good" team has lost to someone who sucks, that's what good means. If they beat all comers, they wouldn't just be good.

I think it's time for me to bow out of this conversation.

 

[Ring 6]

Baltimore's only quality win was against the Browns week 1.

Pitt lost at home to Tampa and was out played by the jests.

Sorry if I'm not buying in on Pitt and Baltimore as "good" teams

Baltimore has won against Tampa, Atlanta, Tennessee, Carolina, Cleveland and Pittsburgh.
The AFCN is not strong, but it gets to play the AFCS and NFCS.

The AFCN is 12-1-1 against Tenn, Jax, Tampa, Atlanta and Carolina and 11-13 against everyone else.

 

[Ring 6]

No, it isn't. Seriously, Andy, and I mean this the most genial of intentions, you need to educate yourself about the mechanics of DVOA.

I totally understand it, and as I have said, it is flawed.

As for Balt and Pitt, as I said before, even if I concede those, you have to turn around and say that neither Miami and Buffalo are good teams. And since when did "good" become such a positive adjective? Virtually every "good" team has lost to someone who sucks, that's what good means. If they beat all comers, they wouldn't just be good.

you are the one who is introducing a slant due to opponent into the discussion.

I think it's time for me to bow out of this conversation.

your call.

 

[oldrover]

"I came here for an argument..."

"No you didn't."

 

[Haley]

The mere fact that we disagree on which teams are amazingsauce, good, bad, terribad, etc shows how subjective and elusive it is to pin down a stat that effectively reflects strength of schedule differences.

Also, as Belichick is fond of pointing out, team strength changes throughout the season. Some teams start out strong and finish soft. Others do the opposite. Sometimes its due to away vs home. Sometimes, short weeks. Other times, injuries. I should throw in a shout out to coaching. I just don't think numbers can accurately reflect these week to week changes.

 

[Kontradiction]

He said PFF has the Colts as a better defense. I don't think he ever said the Colts' defense was better.

Yet another example of why PFF is absolutely useless. They're in "tits on a bull" territory.

 

[Oswlek]

Yet another example of why PFF is absolutely useless. They're in "tits on a bull" territory.

I don't begrudge anyone saying that over the course of the entire season Indy has been better. I do take issue when people act as if that somehow is more predictive than how NE has played since the Cincy game, and I take particular issue with those that seemingly want to weigh all yards equally.

For the record, I have said the same about DVOA within the past week. They, at least, have a consistent model and do lower the weighting when games are out of hand. I can't fault their effort. I do think that, historically, Bill's defenses lax more in garbage time than most, though. It could just be me, but I suspect DVOA has penalized NE's defenses for this compared to the rest of the league.

 

[Ring 6]

Andy, you're very statements demonstrate a lack of understanding. You can keep up with the "yeah huh!" type responses or you can calmly, rationally accept there is something for you to learn.

I have no idea what this even means.

Once again I understand dvoa and still that doesn't change that football is not a game best evaluated by statistics.
Feel free to keep being condescending and imply that the only way anyone would disagree is ignorance but you are wrong about that.

 

[patfanken]

Well since this thread has degenerated into a stats focused thread, I guess I'll ask for more. All the stats I've seen thus far have been accumulated over the full 9 games that each team has played. Well I think we call can agree that under the new CBA. what teams are at the start of the season, have little in common with what teams become after a month or two. Don't forget that after week one, the Jets were 1-0 and we were 0-1.

So my question to all those with subscriptions to PFF and FO's, how would these 2 team compare over the last 4 games? Because the teams we see on the field Sunday will reflect those stats more than stats that consider the full 9 games.

 

[Haley]

As I'm writing all these posts out, my dog is puking all over my white rug. If its any consolation, it seems he didn't much care for my posts either.

 

[Haley]

Well since this thread has degenerated into a stats focused thread, I guess I'll ask for more. All the stats I've seen thus far have been accumulated over the full 9 games that each team has played. Well I think we call can agree that under the new CBA. what teams are at the start of the season, have little in common with what teams become after a month or two. Don't forget that after week one, the Jets were 1-0 and we were 0-1.

So my question to all those with subscriptions to PFF and FO's, how would these 2 team compare over the last 4 games? Because the teams we see on the field Sunday will reflect those stats more than stats that consider the full 9 games.

I don't think you need any kind of a subscription to gather this data but then maybe this wasn't what you wanted? Here is what I found. Feel free to double check me, I've been a little distracted while writing this to say the least.

http://www.footballoutsiders.com/dvoa-ratings/2014/week-6-dvoa-ratings
http://www.footballoutsiders.com/dvoa-ratings/2014/week-7-dvoa-ratings
http://www.footballoutsiders.com/dvoa-ratings/2014/week-8-dvoa-ratings
http://www.footballoutsiders.com/dvoa-ratings/2014/week-9-dvoa-ratings

Based on this, something definitely seems off with how they adjust for strength of schedule. I know some places use last year's ranking as the basis for the adjustment. That could be the problem here.

 

[Oswlek]

I don't think you need any kind of a subscription to gather this data but then maybe this wasn't what you wanted? Here is what I found. Feel free to double check me, I've been a little distracted while writing this to say the least.

http://www.footballoutsiders.com/dvoa-ratings/2014/week-6-dvoa-ratings
http://www.footballoutsiders.com/dvoa-ratings/2014/week-7-dvoa-ratings
http://www.footballoutsiders.com/dvoa-ratings/2014/week-8-dvoa-ratings
http://www.footballoutsiders.com/dvoa-ratings/2014/week-9-dvoa-ratings

Nice gather, but that isn't exactly what Ken is looking for. What you have is the evolution of their ratings over the last month; useful but problematic because the first four weeks are still built into the data.

As far as I know, the only way to truly isolate weeks or a series of weeks is to have a subscription to the site. I'd be interested as well, though I suspect DVOA is underrating NE even if we drop the first month off. I say this because DVOA wasn't that impressive with their win over Chicago, it was about 17%, which is an ordinary figure in victory and even attained in a loss at times. I came away far more impressed and thought there were a number of signs that indicated they would have success against Denver.

 

[Haley]

Nice gather, but that isn't exactly what Ken is looking for. What you have is the evolution of their ratings over the last month; useful but problematic because the first four weeks are still built into the data.

As far as I know, the only way to truly isolate weeks or a series of weeks is to have a subscription to the site. I'd be interested as well, though I suspect DVOA is underrating NE even if we drop the first month off. I say this because DVOA wasn't that impressive with their win over Chicago, it was about 17%, which is an ordinary figure in victory and even attained in a loss at times. I came away far more impressed and thought there were a number of signs that indicated they would have success against Denver.

Hah, thanks. I was thinking that was far too easy.

 

[goheels22002]

Indy
Denver, Philly, Jax, TN, Balt, Hou, Cincy, Pitt, NYG

NE
Miami, Minny, Oak, KC, Cincy, Buff, NYJ, Chi, Denver

They both played Denver and Cincy, so those can be ignored. Otherwise, NE played three dumpster fires (Minn, Oak, Chi), one bad team (NYJ) and three good teams (KC,Buff, Mia). Indy played two dumpster fires (Jax, NYG), a bad team (TN), an average team (Hou) and three good teams (Philly, Balt, Pitt).

The Colts shutting out a pretty solid Cincinnati offense was a major deal, as was the blowout by Pittsburgh. They gave up zero points and then 51 points including 6 TD passes in back-to-back games. The Colts defense is the most schizophrenic in the NFL. It's almost as if they don't have a clue how to scheme against different opponents.

The home loss to the Eagles (30-27) may be the most instructive for the Patriots. It was an early loss to a team with with a hurry up capability and a defense that forced a turnover and tightened the noose late in the game when needed. The Eagles scored the last ten points of that game and came back to hand the Colts a home loss that exposed the limitations of both the offense and defense of the Colts.

 

[patfanken]

Hah, thanks. I was thinking that was far too easy.

But I appreciate the effort, Haley. Thanks

 

[Oswlek]

After that long discussion, I figured it might be useful to actually explain several ways that DVOA reads things differently than cumulative stats do. It is absolutely true that DVOA looks at every play and the yards accumulated (or stopped) all matter, but the following also have a great impact on how teams grade out:

1) Opponent adjustment - the most obvious one. 72 yards rushing against the 2006 Vikings would be considered substantially more impressive than a 126 yard day against the 2006 Colts.

2) Garbage Time - another obvious one. Passing for 300 yards in a game you trailed 27-3 midway through the second quarter isn't going to be weighted as heavily as one that was competitive throughout. Aaron has actually tried to remove garbage time results entirely, but was never able to find a useable line of demarcation. Every attempt ended up with a worse result. It should be noted that even in these situations, they are opponent adjusted (and situationally adjusted as well, as discussed below).

3) Situational adjustment - A 7 yard run on 3rd and five is more useful than a 17 yard run on 3rd and twenty five. DVOA is very good at eliminating (or diminishing) these extraneous yards that fill a stat sheet but don't do much to help you win.

4) Explosive plays - Oddly enough, DVOA has found that the predictive value tends to lessen the longer the play. Said another way, while it would consider Welker's 99 yard catch in the 2011 opener to be more impressive than a 25 yarder where the safety came over to make the tackle, the difference isn't nearly as large as it is in Wes' receiving stats.

There are plenty more, these are just a few that I can list off the top of my head. It is also important to note that these factors can be blended as well. Maybe a team is average at stopping the run, but are terrific in short yardage... how you do against them on 3rd and 2 involves situational and opponent adjustments. Perhaps explosive (if the RB breaks through a stacked line and rumbles for a 73 yard TD) and garbage ones as well (depending on the score).

They really do a great job that looks significantly deeper than just a reorganization of cumulative yards. Not that it doesn't have problems, of course. Their ratings don't consider the 1998 Vikings an all-time great offense because they had an unnatural amount of big plays. A few years back the Jets had a bizarrely high offensive rating, that was ultimately attributed to the fact that their poor drives were generally three and out and their successful drives tended to be 12 play voyages. It created an odd mismatch between the number of successful plays to the actual scoring on the field.

Unlike PFF, though, footballoutsiders don't hide these issues or pretend that we must side with their ratings when a discrepancy arises. In fact, they regularly point out these oddities and use them to facilitate discussion.

Alright, back to your regularly scheduled programming....

 

[crew1954]

but if you separate out the meaningless stats ( stats acculumulated in a loss are meaningless to me, as are piling on when you already have won) then stats would be a good gauge.

Otherwise winning is the #1 stat for a QB IMO.

I disagree with the bold parts, at least on certain occasions. If a team's D is just matadoring the ops O to the end zone and a QB is matching the opposition TD for TD only to lose by a point 41-40 when he misses a 2 point conversion as the clock runs out, that is meaningful, to me. If the wheels fall off for a half on the entire team and the QB then puts on a clinic to catch up from down 28-0 only to lose 41-40, that is meaningful ( sort of like Brady vs. SFO last year).

As for winning, we are back to the argument that Bob Griese and Eli, to name just 2, were better than any of their less winning peers.

 

[Rob0729]

This thread has gotten off topic a bit. Can we get back to the most important thing to come out of today's Bedard segment?!? Felger is a douche.