Blog Entry

WEEK 17 PLAYOFF SCENARIOS

Posted on: December 25, 2011 4:47 pm
 
Well...we finally made it through another tiebreaker season.  It's been interesting, especially in the AFC West and with all the potential ties at 9-7 in both the AFC and NFC.  I'm putting out Week 17 scenarios a little early since all AFC games are done for week 16 and the NFC is pretty cut and dry now.

Also...you should know that the scenarios below were simplified as two potential scenarios that would come down to Strength of Victory (SOV) tiebreaker are already locked up.  If BAL wins and NE loses, the battle for the #1 seed comes down to SOV as they would be tied in conf record and common opponents.  BAL has already secured the better record among teams they have defeated (SOV) over NE, otherwise we might have had some games with playoff impact that are not directly related to BAL-NE.  The other scenario where SOV may come into play is a 4-team tie at 9-7 between CIN-NYJ-TEN-OAK (needs CIN loss, NYJ win, TEN win, OAK win and DEN win).  CIN would drop out on conf record and NYJ-TEN-OAK don't have enough common opponents so it goes to SOV.  OAK has already secured better Win-Loss-Tied percentage among defeated opponents (assuming scenario above) over TEN and NYJ.

Also....TEN is probably the most interesting scenario to figure out.   If you look at TEN scenarios below, basically they need another team to match up with them and CIN at 9-7 to avoid losing H2H to CIN.  If Jets win...they win that 3-way with NYJ-CIN-TEN on common opponents over NYJ after CIN drops out on conf record.  BUT...they can't have OAK as Wild Card potential at 9-7 as well because in that case CIN drops out and not enough common opponents for NYJ-TEN-OAK and OAK wins that tiebreaker on Strength of Victory.    BUT....if OAK is there as potential WC at 9-7 (both DEN and OAK win), TEN can advance if NYJ loses as then it would be TEN-CIN-OAK and CIN would drop out on conf record and TEN beats OAK on common opponents.  So TEN gets in if they Win and CIN loses and either NYJ wins or OAK wins (and doesn't win division)...BUT NOT BOTH.  Weird.

So, we're left with below:

WEEK 17 PLAYOFF PICTURE (also includes remaining 2 Week 16 games):

NFC

  CLINCHED:    Green Bay Packers -- North Division and first-round bye.
San Francisco 49ers -- West Division.
New Orleans Saints -- wild card spot.
Detroit Lions -- wild card spot.
  ELIMINATED:  Arizona, Carolina, Minnesota, Philadelphia, St. Louis, Seattle, Tampa Bay, Washington.

 GREEN BAY PACKERS
  Green Bay clinches homefield advantage throughout NFC playoffs:
   1) one GB win or tie
   2) SF loss or tie

 SAN FRANCISCO 49ERS
  San Francisco clinches a first-round bye:
   1) SF win
   2) SF tie + one NO loss or tie
   3) one NO loss
  San Francisco clinches homefield advantage throughout NFC playoffs:  
   1) SF win + two GB losses

 NEW ORLEANS SAINTS
  New Orleans clinches NFC South Division:
   1) one NO win or tie
   2) one ATL loss or tie
  New Orleans clinches a first-round bye:
   1) two NO wins + SF loss or tie
   2) one NO win + one NO tie + SF loss

 NEW YORK Giants
  NY Giants clinch NFC East Division:
   1) NYG win or tie

 DALLAS Cowboys
  Dallas clinches NFC East Division:
   1) DAL win

 ATLANTA Falcons
  Atlanta clinches NFC South Division:
   1) two ATL wins + two NO losses
  Atlanta clinches a wild card spot:
   1) one ATL win or tie
   2) one CHI loss or tie

 CHICAGO Bears
  Chicago clinches a wild card spot:
   1) two CHI wins + two ATL losses

 AFC

  CLINCHED:    New England Patriots -- East Division and a first-round bye.
Houston Texans -- South Division.
Baltimore Ravens -- wild card spot.
Pittsburgh Steelers -- wild card spot.
  ELIMINATED:  Buffalo, Cleveland, Indianapolis, Kansas City, Miami, Jacksonville, San Diego

 NEW ENGLAND PATRIOTS
  New England clinches homefield advantage throughout AFC playoffs:
   1) NE win or tie
   2) BAL loss or tie + PIT loss or tie

 BALTIMORE RAVENS
  Baltimore clinches AFC North Division and a first-round bye:
   1) BAL win
   2) BAL tie + PIT loss or tie
   3) PIT loss
  Baltimore clinches homefield advantage throughout AFC playoffs:
   1) BAL win + NE loss

 PITTSBURGH STEELERS
  Pittsburgh clinches AFC North Division and a first-round bye:
   1) PIT win + BAL loss or tie
   2) PIT tie + BAL loss
  Pittsburgh clinches homefield advantage throughout AFC playoffs:
   1) PIT win + BAL loss or tie + NE loss

 DENVER Broncos
  Denver clinches AFC West Division:
   1) DEN win
   2) DEN tie + OAK loss or tie
   3) OAK loss

 OAKLAND Raiders
  Oakland clinches AFC West Division:
   1) OAK win + DEN loss or tie
   2) OAK tie + DEN loss
  Oakland clinches a wild card spot:
   1) OAK win + CIN loss + TEN loss or tie
   2) OAK win + CIN loss + NYJ win
 
 CINCINNATI Bengals
  Cincinnati clinches a wild card spot:
   1) CIN win or tie
   2) NYJ loss or tie + OAK loss or tie
   3) NYJ loss or tie + DEN loss or tie
 
 NEW YORK JETS
  NY Jets clinch a wild card spot:
   1) NYJ win + CIN loss + TEN loss or tie + OAK loss or tie
   2) NYJ win + CIN loss + TEN loss or tie + DEN loss or tie

 TENNESSEE Titans
  Tennessee clinches a wild card spot:
   1) TEN win + CIN loss + NYJ win + OAK loss or tie
   2) TEN win + CIN loss + NYJ win + DEN loss or tie
   3) TEN win + CIN loss + NYJ loss or tie + OAK win + DEN win

Comments

Since: Dec 18, 2008
Posted on: December 29, 2011 2:09 pm
 

WEEK 17 PLAYOFF SCENARIOS

3 unrelated points....
1)  I believe all this discussion on hypothetical tiebreaking changes, and the personal likes and dislikes of each step, have been fair game all season. It's the discussion about the inefficiencies/redundancies/ambigui

ties of the current language that has been put on hold until after the season is over. Joe said he would set up a new blog for this after the season is officially over, which would be sometime after Sunday's games.

2)  Personally, I would not like it if the tiebreakers were modified so that non-conference games took on less importance in the tiebreaking procedure. I think all games should be equally important, however there is no way of getting around the fact that division games are more likely than other games to affect a team's position within its division.

3)  Agn said,
So [for divisional ties]:

1. Head-to-head (best won-lost-tied percentage in games
among the clubs.)
2. Best won-lost-tied percentage in games played within the
division.
3. Best won-lost-tied percentage in confernce common games
4 [new]. Best won-lost-tied percentage in all common games.
4 [old]. Best won-lost-tied percentage in games played within the
conference
...

Under the current scheduling format, using my process, the old step 4 is  superfluous, as I don't think you could be tied in all the previous steps without also being tied for conference record

Agn is correct. Not only would the "old" conference tiebreaking step automatically be tied (and be unnecessary to include), but this structure has the added bonus that the two non-common games are never used to break a tie. Currently, the non-common games are counted at the conference record step.

(The "strike-through" formatting got lost when I pasted the text.)








Since: Dec 27, 2006
Posted on: December 29, 2011 1:53 pm
 

WEEK 17 PLAYOFF SCENARIOS

I think either of my ways (but particularly the second: conf before common) preserves the notion that winning inter-conference games at the expense of intra-conference games should always hurt you, not help you as it does in the current  division tie-breakers.




Since: Dec 27, 2006
Posted on: December 29, 2011 1:47 pm
 

WEEK 17 PLAYOFF SCENARIOS

Better yet, since the final step is superfluous, reverse them, and this removes the common game step altogther. Simply replace the common games tie-breaker with the common conference games (which people and websites have an easier time following anyways)

1. Head-to-head (best won-lost-tied percentage in games
among the clubs.)
2. Best won-lost-tied percentage in games played within the
division.
3. Best won-lost-tied percentage in conference common games
4. Best won-lost-tied percentage in games played within the
conference
3. Best won-lost-tied percentage in all common games.
...

This restores the importance of conference play, and is a "ever-expanding subset": 2-6-10-12. Not the "replace these games with those games" that we have know.

Even my other way (common before conf) is expanding 2-6-10-14. with only teams added none removed

Currently we do 2-6-14-12, which is not expanding, and involves replacing teams.




Since: Dec 18, 2008
Posted on: December 29, 2011 1:39 pm
 

WEEK 17 PLAYOFF SCENARIOS


nygsb42 said:
So in 4 way the 2-1 2-1 1-2 1-2 scenario comes into play. What's your take Brett, since you created it?
I intentially did not comment on this because I figured whatever is decided for the "head-to-head only" step would also be the practice at the "head-to-head/common games" step.

But, since you brought it up, let me throw out a 3rd possibility here.
Like Big 10 Football, the NFL could ignore tied games (ie. treat them as if they were never played). The NFL could not count head-to-head results in the [1-0-1, 1-1-0, 0-1-1] case by reducing these records to [1-0-0, 1-1-0, and 0-1-0]. BUT, at the same time apply the head-to-head step in the 4-way [2-1, 2-1, 1-2, 1-2] case.

Wording this is tricky because we want a "complete set of H2H games" to count in the 4-team example, but not in the 3-team example. How about:
head-to-head record. .....for 4-team ties: A team is eliminated if it has a loss against 2 or more of the teams it is tied with.
This wording has the feel and appearance of consistency with when head-to-head is applied in 3-team cases. I believe this wording also handles all of Jerry's "group sweep" cases without needing to define a new and potentially confusing tiebreaking term
(Jerry, I would prefer to keep the language as simple as possible. Ie. not worry about 5 and 6 team ties until they become possible.)

Let's look at a few (1 in a billion) 4-team cases to see if the wording handles them the way we intend it to.


Teams eliminated are in boldface.

a)  2-0-1, 2-0-1, 1-2-0, 1-2-0

b)  2-1-0, 1-1-1, 1-1-1, 1-2-0

c)  2-1-0, 2-1-0, 0-2-1, 0-2-1

d)  2-0-1, 2-1-0, 1-2-0, 0-2-1

e)  2-0-1, 1-1-1, 1-2-0. 1-2-0

Even though removing a tied game has the consequence of there no longer being a "complete set" of head-to-head games, I believe that removing the tied game should still result in the elimination of the 2-loss teams. This should be acceptable since traditional sweeps in 3-team ties often do not have a complete set of head-to-head games.

Also, despite the tied result in these cases, I believe applying head-to-head records (in accordance with the above wording) is consistent with not applying head-to-head records in the 3-way [1-0-1, 1-1-0, 0-1-1] case. Consider the following argument:

In the 3-way case, replacing the tied result with a different result could make the head-to-head records [1-1-0, 1-1-0, 1-1-0]. However, in all the 4-way cases above, replacing the tied result with a different result could not make any of the 2-loss teams better than 1-2-0. And, since we would be eliminating the 1-2 teams in the [2-1, 2-1, 1-2, 1-2] case, we would want all 0-2-0, 1-2-0 and 0-2-1 teams to be eliminated in all cases.



Since: Dec 18, 2008
Posted on: December 29, 2011 1:38 pm
 

WEEK 17 PLAYOFF SCENARIOS


Vito said:
Actually, all teams have exactly the same schedule, if you only count the standings rank from the previous year.  A last place team does not play itself, so it's divisional opponents are 1, 1, 2, 2, 3, and 3.  Then they play two 4th place teams, as mentioned.  It obviously works the same for any given ranking.
I also agree this is a great point by Vito. And good suggestion by Agn to keep this in mind when discussing potential 17 or 18 game schedules.

Agn said:
2. Head-to-head (apply only if one has defeated each of the others, one club has lost to each of the others, or -- if all teams have played each other -- best won-lost-tied percentage in games among the clubs.)
Very nice wording.



Since: Dec 27, 2006
Posted on: December 29, 2011 1:29 pm
 

WEEK 17 PLAYOFF SCENARIOS

So:

1. Head-to-head (best won-lost-tied percentage in games
among the clubs.)
2. Best won-lost-tied percentage in games played within the
division.
3. Best won-lost-tied percentage in confernce common games
4. Best won-lost-tied percentage in all common games.
4. Best won-lost-tied percentage in games played within the
conference
...

Under the current scheduling format, using my process, the old step 4 is  superfluous, as I don't think you could be tied in all the previous steps without also being tied for conference record



Since: Dec 27, 2006
Posted on: December 29, 2011 1:19 pm
 

WEEK 17 PLAYOFF SCENARIOS

Maybe the step after division record should be "common conference opponents".
Then we get to the all common vs all confernce debate. But at least we can then say that, right from the onset of a season a team knows that its games against division opponents are supremely important followed by the "other division within conference". That takes care of 10 games. And these have the same importance whether you are talking division or wild card. Then you can take your pick which should be the next in importance: the 2 non-common conference games, or the 4 common non-confernce games.

I would think the 4 non-confs are the next most important, which makes common higher ranking than confernce, just as it is currently configured. However adding the common conference step softens that pardign shift that pillbox mentions.




Since: Dec 20, 2010
Posted on: December 29, 2011 12:50 pm
 

WEEK 17 PLAYOFF SCENARIOS

Thanks everyone for their thoughtful comments.
Vito and agn, I can't really wrap my brain around how they are the same schedules in terms of division rankings. I just see 14 common games for each team within a division, and then the 2 non-common games which are based on the previous season's rankings. I may very well be missing the point! As I said -- I am not that analytical.

In any case, agn summarized  my key beef (and sorry I don't know how to cut and paste to make things blue):

"That is a problem, given that, if you do not win your division, conference games take on supreme importance. In other words, be sure to win your conference games, unless you are in a division struggle , in which case be sure to win your non-conference games."

Indeed, the current tiebreaking system basically contains a paradigm shift, explained in agn's quote above. You know, for years, it was *always* drummed into us that an inter-conference game was not as important as a conference game. And now, for divisional ties, that is not the case. While Joe and others here have explained the rationale, I think the paradigm shift has been largely glossed-over.

Oh and nygsb42, I wasn't so much commenting on SoS as an applied rationale for my beef, since the matchups were obviously determined from standings the year before. It just seems wrong to my un-scientific brain!

At any rate, I appreciate the feeback. As I Jets fan I couldn't help but look one more time this morning at how their Denver loss flipped the season. Had they won that, it could have set up an incredibly close SoV, and maybe even SoS situation with the Titans this week that this blog would have been all over! Oh well, maybe next year!






Since: Nov 21, 2011
Posted on: December 29, 2011 12:30 pm
 

WEEK 17 PLAYOFF SCENARIOS

Maybe arbitrary is the wrong word. But that is not really my (unpopular) point. There is clearly a system, so by definition of the word it is not arbitrary. What it is is non-ideal. There is no way to do the tiebreakers that is without downside or clearly better than another way. I would bet that everyone here would want unique minor changes to different sections of the tiebreakers (not so drastically as my rejected SoS overhaul thoughSmile). So I'll stand by my point that since everyone has a POV (again, not complaining about lack of influence on the rules committie or whoever determines tiebreakers), there's no good way to do it.

Pillbox, having read your point about the 2 standings determined games not counting in common games, the only thing I can say is that the tiebreakers do adhere to "largest comparable subset" priority (which I obviously am not a fan of), and your problem regards SoS, which everyone likes to reject as a good tiebreaker. So while I agree with your problem, I think there is no solution within the restraints of the outlines of the current tiebreaker system. 



Since: Dec 27, 2006
Posted on: December 29, 2011 10:14 am
 

WEEK 17 PLAYOFF SCENARIOS

Actually, all teams have exactly the same schedule, if you only count the standings rank from the previous year.  A last place team does not play itself, so it's divisional opponents are 1, 1, 2, 2, 3, and 3.  Then they play two 4th place teams, as mentioned.  It obviously works the same for any given ranking.
Great point Vito.

In fact all teams in the league play 4 games againt opponents that finished 1st, 2nd, 3rd, and 4th place.

It would be worth keeping this in mind when discussing 17- or 18-game schedules.





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