Factoring in SOS to statistics

[Mikie2times]

This is long, but will be worth it to die hards and handicappers so just bear with me.

 

NFL statistics are extremely biased, much of the value statistics have gets lost without factoring in the strength of your opponent. So for my own handicapping purposes I started a stat project that factors in strength of schedule. After all how can we say a certain area of this team is performing well if the competition being played has performed far below NFL averages.

 

Rather then try and go threw a lengthy explanation of how I reached these statistics I will just give a quick example. Let’s say Buffalo is scoring 25 points per game, and the opponents they played have averaged allowing 20 points per game. The NFL league average for points allowed is 17, so overall Bills opponents have allowed 3 more points then the league average. Since Buffalos opponents have allowed three more points then average three points gets deducted from the Bills total. So against an average team Buffalo would score 22 points, and not the 25 points the scored against slightly inferior competition.

 

That example is fictitious, but I have done this process with over 25 NFL statistics and if anybody wants the excel file feel free to PM me. The following is how the Bills stack up across the league in terms of ranking when everything is averaged out. You will also notice I didn’t include rush yardage or passing yardage, both these statistics get manipulated by how good or bad other areas of your team are. Instead I use averages which are a much better indication of the overall performance.

 

If an asterisk is by the ranking it means it was not averaged out league wide.

 

Offense

Points Scored- 30

Pass %- 26

Yards Per Pass- 32

QB Rate- 29

Sacks Per Pass Attempt – 28 *

Sacks Allowed- 24

Yards Per Rush- 11

3rd Down Conv %- 9

Penalties- 23 *

Giveaways- 10

 

Defense

Points Allowed- 15

Pass %- 17

Yards Per Pass- 8

QB Rate- 4

Sacks Per Pass Attempt – 7 *

Sacks- 25

Yards Per Rush- 31

3rd Down %- 26

Penalties- 28 *

Takeaways- 4

 

ST/ MISC

Kick Return average- 1

Opponents Kick Return average-5

Punt Return Average- 12

Opo Punt Ret Avg- 25

FG%- 8 *

Punt Avg- 5 *

Turnover margin- 4

[Mikie2times]

I was going to post the match up with Oakland and some analysis if you guys want, but it doesn't appear like people care that much about these stats.

[Matt in KC]

I was going to post the match up with Oakland and some analysis if you guys want, but it doesn't appear like people care that much about these stats.

I care, I'm just not quite sure what to make of these.

 

 

Pass %- 26

So 26% of plays were passes? ..or 26% of passes were completed? I'm not sure I get it. (sory if I'm thick)

 

Yards Per Pass- 32

Really? This is yards per completion adjusted based on how many yards per completion these teams allowed elsewhere?

 

QB Rate- 29

Negative twenty-nine seems a bit low....

Sacks Per Pass Attempt – 28 *

So for every pass we were sacked 28 times?

 

Are you sure some of these aren't the reciprocals?

 

I like the concept, and could geek-out looking at the Excel, but I'm having a hard time engaging with these numbers.

 

Okay, I should apologize even before I even post... I just realized those are the ranks! Wow, I'm usually a bit quicker... What does the defensive QB rating mean? That we made opposing QBs look worse than all but 3 teams did? Also, did you look at time of posession?

 

Please do post about the Oakland matchup!

[Lothar]

I care, I'm just not quite sure what to make of these.

So 26% of plays were passes? ..or 26% of passes were completed?  I'm not sure I get it. (sory if I'm thick)

 

Really?  This is yards per completion adjusted based on how many yards per completion these teams allowed elsewhere?

 

Negative twenty-nine seems a bit low....

So for every pass we were sacked 28 times?

 

Are you sure some of these aren't the reciprocals?

 

I like the concept, and could geek-out looking at the Excel, but I'm having a hard time engaging with these numbers.

 

Okay, I should apologize even before I even post...  I just realized those are the ranks!  Wow, I'm usually a bit quicker...  What does the defensive QB rating mean?  That we made opposing QBs look worse than all but 3 teams did?  Also, did you look at time of posession? 

 

Please do post about the Oakland matchup!

478665[/snapback]

 

I think he's just quoting relative rank with respect to the NFL - 29th out of 31 etc.

[Mikie2times]

Thanks for a response

 

I will post the Oakland match up, I did one for the Colts game tonight and it looks very accurate. Like I said I created this to handicap games, because the stats get averaged out the weak and strong teams become very apparent. This basically is measuring in every phase of the game how strong your competition has been compared to the average team, and how well you faired.

 

One discrepancy I noticed with Buffalo is the pass defense. We lead the league in yardage allowed by a decent margin, but in reality were not as good as it appears. The passing offenses we have faced have been way below the NFL average statistically, and they haven't been throwing the ball that often. Buffalos run defense has been so bad teams just don't pass, but when they do the average per pass puts buffalo at 8, and not 1. Eight spots in that category is a big jump, and may describe the reason for our 3rd down defense.

[HopsGuy]

I'd be interested in your statistical analysis. My friend believes that he can look at 3 stats in a game and tell you who won: turnovers, red-zone efficiency, and 3rd down conversion rate). I think that's probably true.

 

Of course, I also believe that past performance is no indication of future results.

[Lothar]

Interesting stuff Mike.

 

One thing that may add more credence to your analysis is to use a weighted average or a trend function. For example, the Bills after the 1st 4 games last year sucked in every offensive category. But as Willis got going, the points per game, quarterback completion rates, 3rd down conversions - all started improving.

 

That was one reason why I felt JP would be ok this year. Bledsoe ranked ~26th in the league last year - if JP picked up the offense quickly (re: not looking overwhelmed) - we shouldn't have missed a beat. Now with Takeo's loss and our somehat wimpy Run D, I still think we can be successful, albeit it may take a mediocre 15-20th ranked QB to keep us in the playoff hunt. I think Holcomb fits the bill.

 

Thanks for a response 

 

I will post the Oakland match up, I did one for the Colts game tonight and it looks very accurate. Like I said I created this to handicap games, because the stats get averaged out the weak and strong teams become very apparent. This basically is measuring in every phase of the game how strong your competition has been compared to the average team, and how well you faired.

 

One discrepancy I noticed with Buffalo is the pass defense. We lead the league in yardage allowed by a decent margin, but in reality were not as good as it appears. The passing offenses we have faced have been way below the NFL average statistically, and they haven't been throwing the ball that often.  Buffalos run defense has been so bad teams just don't pass, but when they do the average per pass puts buffalo at 8, and not 1.  Eight spots in that category is a big jump, and may describe the reason for our 3rd down defense.

478694[/snapback]

[Mikie2times]

Here is the Bills at Oakland Match up using the adjusted statistics. Rather then post a unreadable chart I just did a write up including some of the highlights.

 

Raiders on Offense

 

The raiders with Randy Moss didn’t pass with much consistency, without him these stats plummet further. Oakland completes only 56% threw the air, and when matched up against our secondary should be held to a passer rating under 70. While the outlook is rosy for the pass defense the same can't be said for the run defense. Oakland isn't a good rushing team, but matched up against us they should average around 4.3 a clip. With Moss out expect Oakland to try and establish the ground game. They saw us let one of the leagues worst ground games torch us, and will certainly try play to our ground weakness over our pass strength.

 

Key to winning: Force turnovers and stop third down conversions. Oakland is ranked first in not turning the ball over, Buffalo is ranked fourth in forcing takeaways, something’s got to give. Despite how horrible Buffalo's third down defense is, Oakland has been equally bad in that department. Oakland average of 27.6% will kill an offense, and they don't pass with much efficiency, but Buffalo is 26th on 3rd down defense so both coaches are looking to right the ship this week.

 

Bills on Offense

 

This weekend will be a real test for this offense. We run the ball well with an average of 4.14, but Oakland holding teams to 3.75 per rush. Hopefully the addition of a CV and MW will pay dividends in the run game for the second straight week. As for our passing game the games in which JP played have really pulled down our averages, so it’s difficult to measure how effective this unit is. I have Oakland’s pass defense as the 17th best in the league, and 21st in QB sacks. They certainly aren’t world beaters, but they play around the league average, and being on the road should make it a challenge for the Bills passing attack.

 

Key to Winning: Keep the completion % high in the passing game. Holcomb has been able to keep our offense is manageable third downs by hitting on a high rate of his short passes. Oakland’s pass defense, while average, is 5th in the league in completion %. The ground game won’t be effective all day, so Holcomb needs to continue to complete short throws and win the 3rd down battle against Oakland who ranks 21st in third down defense.

 

Special Teams

Oakland ranks 20th in Kick Returns, 30th in kick returns allowed, 22nd on Punts returns, and 21st in Punt returns allowed. Janakowski is also slumping badly as their 28th in the league in FG accuracy. Bobby April should be salivating at this match up as the Bills hold a major advantage in almost every ST category. Look for Bills ST to play a major role in this one, and Terrance to possibly return his first kick for a TD.

 

Final Thoughts

Offensively and defensively these two teams match up pretty evenly, so anybody expecting an easy win could be in for a rude awakening. Being on the road will make it a challenge, and our ST needs to take advantage of this obvious mismatch. Expect a low scoring, tightly contested game, one that will go in Buffalos direction if they take advantage of the ST mismatch. If not Oakland probably wins by a field goal and the Sunday night game in New England becomes a must win.

[stuckincincy]

Here is the Bills at Oakland Match up using the adjusted statistics.  Rather then post a unreadable chart I just did a write up including some of the highlights.

480660[/snapback]

 

Good read, Mike. I appreciate the effort.

[Mikie2times]

Good read, Mike. I appreciate the effort.

480663[/snapback]

 

Thanks Cinci, I wasn't sure many people would read the preview with it being a few posts down. For what it's worth the first game I have used these stats with was the St Louis/Colts game and it had Indianapolis -17. That was spot on, but I will admit it was pretty far off on the total. Who would have thought the Rams would throw up 17 in the first few minutes of that one?

 

For all you gamblers here is my lock of the week (if such a thing exists)

 

 

49ers at Redskins- The Redskins -12.5.

 

These statistics are being influenced positively in the 49ers direction because of Rattay starting a few games, kind of like our passing stats our being dragged down by Losman starting a few games. Despite that the Redskins still show being a favorite of 20+ points. Statistically the Redskins hold a major advantage, but as any good handicapper knows it’s not all about statistics. The skins are coming off back to back road losses so this one is a must win for them. To make matters worse for the 49ers Alex Smith makes his first road start against a dominating Washington pass defense. In Smiths only other start of the year he threw 5 INT’s at home against the Colts. Don’t be afraid of the big spread, the Redskins will cover by a TD or more.