Stats that matter IMO (1 Viewer)

[mister pc]

There is a number of stats that can be used to see who is the best. If you are using offensive time of possesion you also need to see the opponets TOP on defense. Compare them and come up with the offense TOP then do the other team.



I agree on all but the 1st 2. Points allowed and scored are skewed. Thats why I look at YPP and PPP because it gives me a better judgement on how they are scoring. PA and PS like you mentioned just tells me the whole picture which is fine. But imo to examine a team indepth and get a clear picture on who is good and who is not you must use stats that break down the team per yard,per point, per play and per possesion.

points allowed/points scored is arguably the most important of stats. When a team is allowing fewer points than they are scoring, thats how you get wins. Try to find a flaw in that argument... you cant.

Seriously though, points allowed is the first stat by which I measure how good a defense is.

 

[LSSpam]

I think the Saints are particularly illustrative in this instance.

The Saints are 5th/10th on offense, which has to be fairly comparable to where we were last year. At least, comparable enough that the defensive improvement should outweigh it. Yet we're 9-3 and don't have thee same margins of victories we had at this point last year.

The simplest explanation is turnovers, which are not accurately reflected in your statistics. The offense is committing more of them, which represents both an aborted drive and poor defensive field position, and the defense is not getting as much, which represents an extended drive and poor offensive field position.

I suppose to an extent your statistical catagories do, indirectly, measure some of this. A team with high OYPP for instance would, on average, likely have more manageable third downs and therefore a higher conversion rate. And by the same token, a team consistently getting good field position would likely benefit in terms of OPPP. So maybe I was wrong to call them "worthless". But this is an indirect way of measuring the more reductive statistics, i.e. simply looking at 3rd downs and turnovers (and redzone efficiency).

For instance the Saints are 5th in OYPP, but we're 2nd in 3rd down conversion percentage. I.e. some teams are just better at converting even long third down situations and therefore sustaining drives.

 

[LSSpam]

points allowed/points scored is arguably the most important of stats. if a team is allowing fewer points than they are scoring, thats how you get wins. Try to find a flaw in that argument... you cant.

Well it doesn't explain why a team is scoring or preventing scores. So it introduces a fluke factor that makes it hard to be predictive.

 

[Galbreath34]

#1 TO differential
#2 ANY/A differential
#3 anything else, but all stats lie to some degree.

Of course, all tell the truth to some degree. If you feed your family by gambling, more power to you. Not many even experts in Vegas do that by direct gambling. More do it by selling the phantasm of a system that gives an edge than by actually using one.

 

[LSSpam]

Extending this further

This is Footballoutsiders Drive stats
http://www.footballoutsiders.com/stats/drivestats

As you can see Indianapolis and New Orleans are 1 and 2 in YPD, 3 and 4 in PPD, 2 and 5 in Touchdowns PD. These are things your statistics roughly approximate.

Despite this anyone watching football can just feel Indianapolis and New Orleans' offenses just aren't as good, at least over the course of the full 12 games played, as they were the first 12 games of last year. Why?

Something your stats don't measure. Indianapolis and New Orleans are 27 and 30 in INTs PD and 18 and 9 in Fumbles PD. More turnovers.

 

[mister pc]

Well it doesn't explain why a team is scoring or preventing scores. So it introduces a fluke factor that makes it hard to be predictive.

Why do we need to know why a team is preventing scores? all that matters is that they are preventing them. If a defense is getting lots of turnovers, thats one way to reduce an opponents scoring opportunities. if they are holding offenses to field goals and punts, that can do the trick as well. It doesnt matter so much how your defense stops the other team from putting up points, so long as the scoreboard shows the end result. Imo the best defense in the league is the one with the fewest points allowed.

 

[The Mongoose]

Plus i have fed my family for years using those stats so either I am lucky or i know how to use them. i will go with the 2nd part of my statement.

With all due respect, you don't have a formula that uses stats to beat NFL point spreads.

You might think you do, but you don't.

So if what you are saying about feeding your family is true, you must be very lucky.

 

[Sophandros]

I love stats as well. But your stats don't explain ATL.
OYPP
23 Atlanta (10-2) 5.2
OPPP
13 Atlanta (10-2)0.367
DYPP
28 Atlanta (10-2)
DAPPP
12 Atlanta (10-2)0.327

Well, Atlanta doesn't turn the ball over and they have a healthy lead in first downs due to penalty, for starters.

 

[BadMoonRising]

instead of per play do it per drive.

If you do that it will reflect the top teams in the league including the Saints. All you are doing is showing what teams allow the fewest yards per play.

So you are showing big play offenses, and few biggest play defenses that doesn't always correlate with winning.

Defense should be about stopping scores, yardage doesn't mean nearly as much.

 

[GameBreaker]

So you are what your stats show you are?

Go figure.

 

[LSSpam]

I love stats as well. But your stats don't explain ATL.
OYPP
23 Atlanta (10-2) 5.2
OPPP
13 Atlanta (10-2)0.367
DYPP
28 Atlanta (10-2)
DAPPP
12 Atlanta (10-2)0.327

I can explain why

#1 in third down conversions on offense
#4 in turnover differential

 

[Sophandros]

I'm not going through the entire thread again yet, but there seems to be some conflict based on what people see statistics for. Statistics are either predictive, explanatory, or descriptive.

People who try to find a formula to help in gambling are trying to determine predictive stats.

People who just want to know what the score was in the game or who has the most yardage, etc. are looking for descriptive stats.

People who want to analyze games and figure why a team is winning or losing or figure out who's the best QB or most efficient RB or strong defense seek out explanatory stats.

I find that much of the disagreement regarding the usefulness of a particular metric or group of metrics revolves around a misunderstanding between parties in WHAT they're talking about.

Stats DO NOT lie. Those who interpret stats may "lie", either intentionally or by accident by not using them properly.

 

[Surviving Saint]

With all due respect, you don't have a formula that uses stats to beat NFL point spreads.

You might think you do, but you don't.

So if what you are saying about feeding your family is true, you must be very lucky.

That is BS. I have one and it has worked for years. Don't think people just get lucky to beat the spread. The handicappers who are consistently winning have a formula. You can't just go off your gut.

 

[SaintRob39]

This is true. If Vegas can get so close to the actually spread of a game then I am sure there are guys out there with formulas that work.

 

[Surviving Saint]

This is true. If Vegas can get so close to the actually spread of a game then I am sure there are guys out there with formulas that work.

You are right. The spread only matters in about 18% of all games played every year though.