Saturday, September 29, 2007

Week 3 Sample Analysis

Tennessee went into New Orleans for a monday night game that was to be a must-win game for the Saints to avoid an 0-3 start. The Saints prepared that week for what they knew would be a battle that they had to bring their 'A' game. Despite the hostile environment and a rabid opponent with their backs to the wall, it was clear before the game that Tennessee was for real this season so far with hard-fought victories won with solid play at the line of scrimmage. This is how this piece of 20-20 hindsight was apparent before the game played out on monday night of week 3:


Tennessee opened the season with a solid victory over Jacksonville that was not assisted by turnovers or any overt breaks involving the proverbial 'bounce of the ball.' The quotient result is based specifically on the opponent played. What's more, the better an opponent does at the line of scrimmage on offense and defense, the more positive that team's result AND the lower an opponent is pushed down towards negative numbers. Therefore, a spread is created between the winner and loser that is wider with greater dominance in results of the quotient. (A spread that doesn't initially seem that close but involves two ppositive numbers usually indicates a fairly competitive game. (...like a 700.00 to 200.00 result, for example).

The quotient result for Tennessee was the best of week 1 suggesting a dominant performance:

--->Tenn.......1584.44333
vs. Jax..........-459.64

Not only did Tennessee's result go up with solid play,
but the Jaguar's resulting quotient was lowered to the minimum result of that week.


Tennessee's Week 2 quotient shows that they gave Indianapolis a solid battle up front (although Indy pushes the envelope with game-breaking plays on defensive secondaries---often resulting in larger final scores, ultimately winning the game):

---> Tenn........612.4433
vs. Ind......505.54

Regardless of the final result on the score board, this was a hard fought battle at the line of scrimmage with both Tennessee and Indy rushing the ball with success at various points during the game.

Given Tennessee's early success with their game plans, it was clear that the Saints would be in for a battle regardless of the ultimate outcome that left New Orleans 0-3.

NOT a Power Ranking!

One of my earliest posts emphasizes that this quotient is NOT a power ranking by any stretch of the imagination, and there are a number of reasons why. Everybody has one developing in their head, but short of being able to break down film and analyze the entire play of the league based on multiple angles This algorithm helps give me a unique angle to do just some of that analysis along with film breakdown, injury status, and performance analysis. In the end, my own hunch about who is what in a power rankings type of setup (like a chess ladder) at any given point in time from week to week---is just that, a hunch!

...With that in mind, we, hopefully, will attempt to run an analysis of the first quarter---the September football season results and try to get a handle on the constant changing landscape of the NFL from week to week. The quotient only adds one more element or Angle--if you will--to view that landscape and---most of all---try to filter out the loud noise of the media and public perception, who are quick to write off certain teams...and "crown" others (a la Denny Green) based soley on Records, Points Scored, and those all to familiar Offensive and Defensive Rankings which are gleaned solely from Yardage results.

Sample Category Breakdown

The Various Sections of the Algorithm

E-C,BG/100
E- D,L,AY
E- E,K
E-F,U,X,AG Just a Summation of the Various Categories
E- G,P
E-H,AH,AV
E-I,AO,AQ
E- J,AI



E-def rsh
E-rsh att
E-sacks
E-scks for Defensive Pressure and Defensive Rushing Results
E-given
scks
sck-given


E-rsh 1st
rsh 1st
op.rsh 1st Rushing
E- rsh avg
rush avg
op-rsh avg


E- 3rd dwn
E- Y
E-X & Y
3rd Some minor Contributors (3rd Dwn conversion and Time of Possession)
down
top
opp-top


rush
opp-rush
E-top Rushing, TDs, and Defense
E-rush td
E Def Rsh 1st
0
0
wk 1
wk 2
drives
lengthdrives
wk 5
drives Mounting and Sustaining Drives

drives
drives
wk 11
wk 12
wk 14
wk 15
wk 16
for byes
td
rush td
curr. Total
curr. Total Previous Week Totals for Cumulative Results

Sample Quotient

Without going into the boring number-crunching, here is a more detailed look at a sample quotient:

Column Headers

E-C,BG/100 E- D,L,AY E- E,K E-F,U,X,AG E- G,P E-H,AH,AV E-I,AO,AQ E- J,AI E-def rsh E-rsh att E-sacks E-scks for E-given scks sck-given E-rsh 1st rsh 1st op.rsh 1st E- rsh avg rush avg op-rsh avg E- 3rd dwn E- Y E-X & Y 3rd down top opp-top rush opp-rush E-top E-rush td E Def Rsh 1st 0 0 wk 1 wk 2 drives drives wk 5 drives length drives drives wk 11 wk 12 wk 14 wk 15 wk 16 for byes td rush td curr. Total curr. Total

Sample Result (from Week One)

Tenn 1584.44333 158444.3 137433.3 112933.33 98125 70000 58750 43250 23250 24500 7636 66 -20 2 2 28125 Tenn 15 5 3600 5.8 4 -2291.67 4.5 5.416667 5 12 36.55 23.05 49 18 13500 4500 20000 Tenn 41.751 41751 6000 5500 9500 14 8 5 3 6750 22 13 13375 570 -35 19 7 1 0 0 Tenn

So What Does It All Mean?

The Quotient is the result of an algorithm which attempts to measure performance on the line of scrimmage from a statistical analysis of the game performance. It is kind of like an ERA in baseball in that it puts things like rushing and stopping the run ahead of broken plays and turnovers which are interpreted partially as sort of an unearned run in baseball.
Specifically it looks at power stats such as rushing first downs and touchdowns as well as defensive pressure from sacks, hurries, and the ability to control the opponents run game as well as the ability to mount and sustain drives.
Football is a wacky game that doesnt necessarily produce a winner from the team that controls the game although both sides attempt to do just that. Wierd bounces of the ball, turnovers, and open field plays resulting from broken plays and missed tackles play an important role in determining the winner. Moreover, the team that controls the line of scrimmage usually slows down the game--lowering the score and their margin of victory---leaving them a big play or two of putting the game in jeopardy. This leaves a back door open for a loss, and many close games are still up for grabs eventhough one team or another is usually treading water or playing bend but dont break defense. The dominant team may not always win.
The media focus, on the otherhand, is clearly focused primarily on the resulting score of the game and a team's record. It assumes that a team is only as good as its record--for lack of deeper analysis--which helps to simplify and break down the contenders from the pretenders from the media. But today's NFL is in far greater flux than 20 years ago. Teams' performances change from week to week and the League is in a tremendous flux based on injuries, how hard teams practice, talent, stability, and coaching among many other factors.
Last season, it is widely believed that the Indianapolis Colts emerged from nowhere to put together a Super Bowl run at the last minute in the playoffs. But did they show any signs of steady improvement in their performance on the field from week to week before this apparent burst?
Even the perception surrounding the New England Patriots during their first Super Bowl run over the St. Louis Rams was one of a flukey, no-name performance, bolstered largely by luck and a "tuck rule" break in the playoffs to put Adam Vinateri in position to win it. But were there signs before that --that they were turning it around in the weeks of a long season that these days are full of a week to week constant flux of teams rising and falling?
The analysis and formulas suggest that any team's chance of winning on "Any Given Sunday" is much more complex than the perception the media loudly parlays from the final scores and the win-loss records that make up the weekly standings.

Wednesday, September 26, 2007

Week Three Results

Week Three Results


Phi..........191386.3
NE..........137020.3
Jax.........124913
Bal..........95475.5
Oak........91840.86
Pitt.........91115.17
TB..........89648
Dall........84407.33
Tenn.....81514.67
NYG.....76689.75
SD.........76675.77
Cinn......70209.33
Ind........62973.77
NYJ.......57172.17
Minn.....55800.5
KC.........55190.83
Mia........49755.5
Sea........26159.89
Chi.........24771.5
Was......23894.75
Hou.......18891.57
StL........17550.5
Atl...........6109
Cleve.......4787.25
GB...........4550.773
SF............4302.667
Arz...........-974.667
NO..........-4740.23
Buff.........-24697
Den.........-35391
Det..........-45043.5


Week Three Cumulative Results



NE
4177.62955
Pitt
3829.59095
Phi
3109.37119
Tenn
3012.03333
Bal
2654.22722
Dall
2271.21333
Was
2141.2225
Hou
1977.60935
Minn
1924.78
Car
1817.81179
Jax
1803.55
Den
1768.43364
Ind
1709.00091
Sea
1604.47746
Oak
1593.01975
Arz
1592.05833
SD
1535.82136
TB
1455.47222
Cinn
1370.00333
Chi
1019.44632
NYG
913.012045
KC
887.454762
GB
833.336616
SF
786.890238
NO
747.420455
Mia
508.323571
StL
493.116667
Det
401.345556
Atl
375.080238
NYJ
212.535556
Cleve
186.111698
Buff
-481.8725

Thursday, September 20, 2007

Cumulative Score after Week 2

cumulative result after week 2

Pitt......................2918.43929
NE.......................2807.42636
Tenn....................2196.88667
Den......................2122.34364
Was.....................1902.275
Hou.....................1788.69364
Bal.......................1691.08056
Arz......................1601.805
Dall.....................1427.14
Minn..................1366.775
Sea.....................1342.87857
Phi......................1195.50833
Ind.....................1079.26318
Car.......................863.056786
Det.......................851.780556
NO.......................794.822727
GB........................787.828889
Chi........................771.731324
SD.........................769.063636
SF.........................743.863571
Oak......................674.611176
Cinn.....................667.91
TB........................558.992222
Jax.......................554.42
KC........................335.546429
StL.......................317.611667
Atl........................313.990238
NYG....................146.114545
Cleve...................138.239198
Mia........................10.7685714
Buff...................-234.9025
NYJ..................-359.18611

Week Two

week 2

Pitt 143754.5
NE 133728.3
Den 130224.5
Hou 104806.5
Jax 101406
Bal 91392.56
Chi 84535.75
Dall 79281
GB 78299.33
Was 76529
Det 67571.56
SF 63990.86
Tenn 61244.33
StL 59019.5
Minn 58829.5
Arz 58241.5
Ind 50554
TB 48251.56
Cleve 46992.77
NO 41354.77
Cinn 38927.5
Phi 24847.17
KC 24471.5
Sea 21025
Mia 16815.86
Atl 16109.86
NYJ 14365.56
Oak -897.382
NYG -2426.77
SD -30706.5
Car -31343.6
Buff -47070.8

Friday, September 14, 2007

Week One Results!

I will first publish the raw results and then comment later given the opportunity. I must caution that these are NOT Power Rankings. One must play their own hunches based on a variety of factors including injuries, matchups, coaching, talent, etc. I will insert my roundup of the rest of the league as we proceed into the season since I focused primarily on the NFC East (which is currently local to my area) in the preseason. So without much further ado, here are the RAW (and I emphasize that) results from Week One:


Tenn 1584.44333
Pitt 1480.89429
NE 1470.14318
Car 1176.4925
Was 1136.985
Sea 1132.62857
SD 1076.12864
Arz 1019.39
Phi 947.036667
Den 820.098636
Minn 778.48
Bal 777.155
Hou 740.628636
Oak 683.585
Dall 634.33
Ind 573.723182
NO 381.275
Cinn 278.635
Buff 235.805
Det 176.065
NYG 170.382273
Atl 152.891667
SF 103.955
KC 90.8314286
TB 76.4766667
GB 4.83555556
Chi -73.626176
Mia -157.39
StL -272.58333
Cleve -331.68853
Jax -459.64
NYJ -502.84167


Keep in mind these are not absolute rankings based on numbers that measure a teams performance against the entire league but are only significant for the conditions of that particular sunday (a la "any given sunday..." and other cliches...) AND that particular sunday's matchup......for ex., Minn played Atl---> The Viking's Week One numbers can only be judged on how they performed on 9/9/07 versus the Falcon's Sept. 9th performance. We don't know how good or bad the Falcons are...so we CANNOT reach any conclusions about how good Minnesota is versus other teams in the league. (If we could, we'd all be rich!). For example, 100 points scored against the Falcons would not be equal to a 100 points scored against the Bears, etc.

(The only way to have numbers from one game that could be compared to another would be to have everyone play the same opponent at the same time under the same conditions ... say any team---take Kansas City---were cloned and played every other team under exactly the same team, weather, health, and other considerations and conditions.......then the results of the games could be compared not only in terms of numbers but actual performance on the field etc....).
The BCS runs into this problem every week in trying to quantify rankings while NFL analysts are continually having to adjust to a league where teams change in performances from week to week and are not static --- neither in terms of their performances, lineups, and health---but their talent level as well--which changes with coaching, practice, experience, health, and other factors.

...........
...........
...........

(I may give my own power rankings based on my own hunches and a detailed review of last minute Sunday Injury Report Analysis from time to time after analysis of week to week results and trends).