Blog Entry

Suggestions for Topics....

Posted on: January 4, 2010 2:11 am
 
Please reply with a topic you would like to discuss. - I'll then create a thread devoted specifically to your topic.


Possible topics :

1) How to expand the NFL schedule to 18 games.

2) Interpretation of the rules in the NFL Tiebreaking Procedures. / Q&A.

3) The odds of rare scenarios not referred to specifically in the NFL Tiebreaking Procedures.
(a) Head-to-head games possibly counting in a 3 or 4 way Wild-Card tie without a sweep!
(b) Strength of Victory (SoV) tiebreaker used between teams with different win totals.
4) Discussing ideas for new NFL tiebreakers.

5) Discussing the merrits of existing NFL tiebreakers.
(a) Should the common games tiebreaker include games against repeat opponents?
(b) Should a head-to-head game be completely ignored in a 3 or 4 way Wild-Card tie?
6) Historical NFL tiebreak info.
Comments

Since: Dec 9, 2008
Posted on: February 2, 2011 11:50 am
 

No ties

iTS WORKING yAY!!!!

aNYWAY, i THINK THERE SHOULD NEVER BE A TIE!



Since: Dec 18, 2008
Posted on: November 28, 2010 4:02 pm
 

Suggestions for Topics....

Falconbrook,

At first casual glance, I didn't think your 2nd idea could work (idea posted on Feb. 22, 2010). However, I now see how it works for a given year.

If the East has split 1 vs. the South, then the South has split 2 vs. the East - (I see now that you mention this yourself).


Example:


Year 1:

East: all vs. North, split 1 vs. South, split 2 vs. West

North: all vs. East, split 1 vs. West, split 2 vs. South

South: all vs. West, split 1 vs. North, split 2 vs. East

West: all vs. South, split 1 vs. East, split 2 vs. North



Now I'm stuck trying to work out the yearly rotation. - ok got it - it would have to be a 6 year rotation. (3 year rotation does not work.)
For this reason, and for all the reasons you metioned, I also like your 1st idea better (posted on Jan 18, 2010).

To get years 4, 5, and 6, and I copied years 1, 2, and 3, then switched the splits.

Year 1:

East: all vs. North, split 1 vs. South, split 2 vs. West

North: all vs. East, split 1 vs. West, split 2 vs. South

South: all vs. West, split 1 vs. North, split 2 vs. East

West: all vs. South, split 1 vs. East, split 2 vs. North


Year 2:

East: all vs. South, split 1 vs. West, split 2 vs. North

North: all vs. West, split 1 vs. East, split 2 vs. South

South: all vs. East, split 1 vs. North, split 2 vs. West

West: all vs. North, split 1 vs. South, split 2 vs. East


Year 3:

East: all vs. West, split 1 vs. North, split 2 vs. South

North: all vs. South, split 1 vs. West, split 2 vs. East

South: all vs. North, split 1 vs. East, split 2 vs. West

West: all vs. East, split 1 vs. South, split 2 vs. North


Year 4:

East: all vs. North, split 1 vs. West, split 2 vs. South

North: all vs. East, split 1 vs. South, split 2 vs. West

South: all vs. West, split 1 vs. East, split 2 vs. North

West: all vs. South, split 1 vs. North, split 2 vs. East


Year 5:

East: all vs. South, split 1 vs. North, split 2 vs. West

North: all vs. West, split 1 vs. South, split 2 vs. East

South: all vs. East, split 1 vs. West, split 2 vs. North

West: all vs. North, split 1 vs. East, split 2 vs. South


Year 6:

East: all vs. West, split 1 vs. South, split 2 vs. North

North: all vs. South, split 1 vs. East, split 2 vs. West

South: all vs. North, split 1 vs. West, split 2 vs. East

West: all vs. East, split 1 vs. North, split 2 vs. South





Since: Dec 13, 2009
Posted on: February 22, 2010 9:37 pm
 

Suggestions for Topics....

Thanks for the support, Brett.

There were actually two different ways to implement my 18-game plan that I could see.  One is the plan that I had explained above, where each division plays 1-2 and 3-4 against each other against one division and 1-3 and 2-4 against each other against the other division.

The other way is to have one division play another division in the following manner:
1 plays 1 & 2
2 plays 2 & 4
3 plays 1 & 3
4 plays 3 & 4

That division then plays a different division as:
1 plays 1 & 3
2 plays 1 & 2
3 plays 3 & 4
4 plays 2 & 4

So in the end, each team plays all the same ranked teams I had originally mentioned.  The difference is that for two divisions playing each other, it's not really symmetric anymore.  Instead, the two divisions view each other as the two different match-up patterns.  But in the end, the head-to-head match-up and common games chart comes out to be identical to the one I had shown for the other implementation.

This second implementation turns out to be easier to expand to different conference make-ups, if the NFL were to later change the number of teams per division or divisions per conference.  But I don't think that's a big enough reason to choose it over the first one.  I think that the lack of symmetry in this one makes the first implemention I had explained easier to follow for the casual NFL fan.  So I would suggest the first scheme I had shown if they do switch to an 18 game schedule.



Since: Dec 18, 2008
Posted on: February 12, 2010 3:35 am
 

Suggestions for Topics....

Falconbrook,

This seems to be perfect solution to satisfying the NFL's desire to have an unbalanced schedule. I like it a lot.
Thank you for thinking this through and explaining its feasibility. It could definitely work.

I'm still looking through the common games chart, but I predict that I will agree with it 100%.





Since: Dec 13, 2009
Posted on: January 20, 2010 12:29 am
 

18 game schedule

I've made a table of the number of common games for ties for the 18 game schedule I'd defined above, similar to the table that Brett had made for the "paired" 18 game schedule in a Jan 6th post.  The interesting thing here is that for the 3 and 4 team conference tie-breaker scenarios, the number of common games can be determined right from the number of head-to-head match-ups that exist betwen the N teams.  The reason for this stems from the fact that for any two teams from divisions where each team plays each other, there's always exactly one common team that they play in each of the other two divisions.

I've arranged the table based on the number of H2H match-ups of each team in the scenario, showing the common games for each team when the number is uneven.  Note that the number of common games is greater when there are less H2H match-ups, which is when the common game tie-breaker would be used most often.  Overall, the number of common games reaches the minimum requirement of 4 for each team to use as a valid tie-breaker in more of the scenarios than the "paired" schedule (3/4 vs 1/2 of 3-team scenarios, and 7/16 vs 1/4 of 4-team scenarios).  And the scenarios where the common games do not meet that requirement are all more likely to have a H2H sweep prior to the common game tie-breaker, as they all involve at least one the the teams having played each other team, unlike in the "paired" schedule.


Division:

# of Teams     H2H     &n
bsp; Common Games   Frequency
1)   2     &nbs
p;     &nb
sp;  2-2     &n
bsp;     14     &nb
sp;     &n
bsp;           2/3    (1-2, 1-3, 2-4, 3-4)
2)   2    &nbs
p;     &nb
sp;   2-2  
;         12     &nb
sp;     &n
bsp;     &
nbsp;     1/3    (1-4, 2-3)
3)   3    &nbs
p;     &nb
sp;   4-4-4     
   10     &nb
sp;     &n
bsp;     &
nbsp;     All
4)   4    &nbs
p;     &nb
sp;   6-6-6-6 &
nbsp;    8     &nbs
p;     &nb
sp;     &n
bsp;     &
nbsp;All


Conference:

# of Teams     H2H     &n
bsp;  Common Games    Freque
ncy
5)   2     &nbs
p;     &nb
sp;  1-1     &n
bsp;     1
1     &nbs
p;     &nb
sp;     &n
bsp;     1/3    (Division all teams played)
6)   2    &nbs
p;     &nb
sp;   1-1  
;     &nbs
p;    7     &nbs
p;     &nb
sp;     &n
bsp;      1/3    (Other two divisions)
7)   2    &nbs
p;     &nb
sp;   0-0  
;     &nbs
p;   10     &nb
sp;     &n
bsp;     &
nbsp;     
1/3
8)   3     &nbs
p;     &nb
sp;  2-2-2     
   4-4-3  
      
;     &nbs
p;     1/4
9)   3    &nbs
p;     &nb
sp;   2-1-1 &nb
sp;      6-7-6     
      
;     &nbs
p;  1/2
10) 3     &nbs
p;     &nb
sp;  1-1-0    
    8-8-7 
      
;     &nbs
p;      1/4
11) 4     &nbs
p;     &nb
sp;  3-3-3-3  &
nbsp;  0   
;     &nbs
p;     &nb
sp;     &n
bsp;    1/16
12) 4     &nbs
p;     &nb
sp;  3-3-2-2    &nbs
p;2-2-3-3    &n
bsp;     &
nbsp;     1/4
13) 4     &nbs
p;     &nb
sp;  3-2-2-1     3-4-4-4   &nbs
p;     &nb
sp;     &n
bsp;1/4
14) 4     &nbs
p;     &nb
sp;  2-2-2-2   &nbs
p; 5     &nbs
p;     &nb
sp;     &n
bsp;       1/8
15) 4     &nbs
p;     &nb
sp;  2-2-1-1     5     &nbs
p;     &nb
sp;     &n
bsp;     &
nbsp; 1/4
16) 4     &nbs
p;     &nb
sp;  1-1-1-1     5     &nbs
p;     &nb
sp;     &n
bsp;     &
nbsp; 1/16




Since: Dec 13, 2009
Posted on: January 18, 2010 8:02 pm
 

18 game schedule

If the NFL were to go to an 18 game schedule, I believe that the fairest way would be the keep each team playing every team in one other conference division, and then for the 4 games against opponents in the other two divisions to be against the teams ranked as such:

1st place team plays 1, 1, 2, 3 (average rank of 1.75)
2nd place plays 2, 2, 1, 4 (average rank of 2.25)
3rd place plays 3, 3, 1, 4 (average rank of 2.75)
4th place plays 4, 4, 2, 3 (average rank of 3.25)

This would allow teams to continue to play other teams from every division each year, which would not be possible if they played each team in two other divisions, as others had suggested.  It also prevents to large incentive to finish 3rd over 2nd to get an easier schedule, as many noted would be the case if they paired 1-2 and paired 3-4.  Instead, the schedule would get gradually easier for each place down.

This scheme may appear non-symmetric and difficult to implement at first, but it's actually pretty simple and manages to retain a measure of symmetry to the schedule.  It can be implemented by having the teams each play every team in one division, for the next division have the 1-2 teams play the 1-2 teams and the 3-4 teams play the 3-4 teams, and for the last division have the 1-3 play the 1-3 and the 2-4 play the 2-4.  Let's call the 1-2, 3-4 pairing Split 1 and the 1-3, 2-4 pairing Split 2.  Each year, each division could rotate which division they would play with each split type as they rotate which division they play each team in.  For example, one year the match-ups might be as such:

      
;     East  West  North   South
East:    N/A    All 
;    Split1   Split2
West:   All     N/A    Split2   Split1
North:  Split1 Split2 N/A   &n
bsp;  All
South: Split2 Split1  All     &n
bsp; N/A

For division tie-breakers, two teams would now have 14 common games with 2/3 frequency and 12 common games with 1/3 frequency.  For Wildcard tie-breakers, teams would have 11 common games against teams in the division they played all teams in (1/3 frequency), 7 common games against teams they played in the other two divisions (1/3 frequency), and 10 common games against teams they have not played (1/3 frequency, but the most important since they have no H2H).  I haven't dug into 3 or 4 team tie-breakers.  In the end, I see no reason to change any of the tie-breaker hierarchy with this schedule, and the strength of schedule diversity remains what it is today.



Since: Dec 18, 2008
Posted on: January 12, 2010 3:54 pm
 

Suggestions for Topics....

Agn,
 If you can't be respectful, please don't post comments on my blog. I will continue to delete all disrespectful comments. You did make a good point though, which I will restate for you.

Agn's point:
"1 win + 1 tie = 1 win" for 2 games against the same common opponent, is inconsistent with "1 tie = 1 half win + 1 half loss" in all other cases.

My reply:
I respect this point. My reply is that "1 win + 1 tie = 1 win" would be part of a group of equations with each equation in the group being an exception to how wins, losses, and ties are treated in all other cases; nowhere else would 2 wins = 1 win (or 2 losses = 1 loss). This group of equations would simply apply only when combining the results from two games into one single result. I think, from this perspective it would be ok if "1 win + 1 tie = 1 win". (Feel free to disagree, if you can do so respectfully).

Since Baltimore played Indy only one time, from my point of view, the 2nd game against Indy for Jacksonville should not be considered a common game, not in anyway that I can see it. By combining the results of two games into one single result, we have a more fair way of comparing records in common games.

New Idea
Another idea would be to take Jacksonville's best result against Indianapolis and count that single result in their common games record.

One drawback to the new idea
The one drawback is that Baltimore could complain that they got only one shot to beat Indy, while Jacksonville got two chances. However, I still like this new idea a lot better than counting both of Jacksonville's games against Indy.

Could all be moot anyway
If the NFL adopts Noob's/Matt's 18 game proposal, there would never be repeat common opponents between teams in different divisions!

For teams in the same division , the repeat opponents would be the same for all the teams, just as they are in the current 16 game schedule.



Since: Dec 27, 2006
Posted on: January 12, 2010 9:48 am
 

Suggestions for Topics....

Wow. Now we see whet kind of commish brett would make.



Since: Dec 27, 2006
Posted on: January 11, 2010 11:57 am
This comment has been removed.

Post Deleted by Administrator




Since: Dec 18, 2008
Posted on: January 10, 2010 9:00 am
 

Suggestions for Topics....

Maybe for Jacksonville only the home game against Indianapolis should count, as Baltimore only played the Colts at home.
Same reason for Baltimore only the away game against Cleveland should count, as Jacksonville only played the Browns on the road. That's a pro.
This is a very clever idea. I like it a lot on one hand, but then again I don't like it. Consider what Jacksonville would think if they had beaten Indy on the road (in the common game that didn't count).

Also, I would have a hard time with this rule for this reason:
- In all other instances, a game is still considered common if it's home for one team but away for the other.


Is there anything wrong with this?
  • 2 wins = 1 win
  • 2 losses = 1 loss
  • 1 win + 1 loss = 1 tie
  • 1 win + 1 tie = 1 win
  • 1 loss + 1 tie = 1 loss
  • 2 ties = 1 tie


The views expressed in this blog are solely those of the author and do not reflect the views of CBS Sports or CBSSports.com