Other Scores for Sat. 2/24

[jtwcornell91]

[quote Trotsky][quote jtwcornell91]Brown 3, Union 1 F
Yale 4, RPI 3 F
Dartmouth 3, Colgate 1 F
Princeton 4, SLU 1 F
CCT 2, Q 0 (3rd)

If my calculations are correct, a Clarkson win would set up the ECAC infinite loop (Cornell-Q tie for 4th requires top 8; RPI-gate-Yale tie for 8th requires top 4).  Let's Go Tech![/quote]

OK, in that case, why are we sure we're in 4th?[/quote]

Tiebreakers Applied

  Tied for 4th place: Cr, Qn

   CAPTION: Head-to-head record

      Team    PF-PA Pct.
   Cornell    2-2   .500
   Quinnipiac 2-2   .500

   CAPTION: Record vs Top 3 (above tie)

      Team    PF-PA Pct.
   Cornell    3-9   .250
   Quinnipiac 3-9   .250

   Top 8 is not complete; cannot apply!

  Tied for 8th place: Cg, Ya, RP

   CAPTION: Head-to-head record

    Team   PF-PA Pct.
   Colgate 4-4   .500
   Yale    4-4   .500
   RPI     4-4   .500

   Top 4 is not complete; cannot apply!

  Implementing infinite loop fix

   CAPTION: Record vs Top 5 (above & including tie)

    Team   PF-PA Pct.
   Colgate 5-15  .250
   RPI     5-15  .250
   Yale    3-17  .150

  Tied for 4th place: Cr, Qn

   CAPTION: Head-to-head record

      Team    PF-PA Pct.
   Cornell    2-2   .500
   Quinnipiac 2-2   .500

   CAPTION: Record vs Top 3 (above tie)

      Team    PF-PA Pct.
   Cornell    3-9   .250
   Quinnipiac 3-9   .250

   Top 8 is not complete; cannot apply!

  Tied for 8th place: Cg, RP

   CAPTION: Head-to-head record

    Team   PF-PA Pct.
   Colgate 2-2   .500
   RPI     2-2   .500

   Top 4 is not complete; cannot apply!

  Implementing infinite loop fix

   CAPTION: Record vs Top 5 (above & including tie)

    Team   PF-PA Pct.
   Colgate 5-15  .250
   RPI     5-15  .250

   CAPTION: Record vs Top 9 (above & including tie)

      Team    PF-PA Pct.
   Cornell    16-16 .500
   Quinnipiac 14-18 .438

  Tied for 8th place: Cg, RP

   CAPTION: Head-to-head record

    Team   PF-PA Pct.
   Colgate 2-2   .500
   RPI     2-2   .500

   CAPTION: Record vs Top 4

    Team   PF-PA Pct.
   Colgate 5-11  .312
   RPI     3-13  .188

  Tied for 6th place: Ha, Pn

   CAPTION: Head-to-head record

     Team    PF-PA Pct.
   Harvard   2-2   .500
   Princeton 2-2   .500

   CAPTION: Record vs Top 4

     Team    PF-PA Pct.
   Princeton 9-7   .562
   Harvard   4-12  .250

[DeltaOne81]

[quote Trotsky][quote DeltaOne81]he's pretty sure[/quote]::looking::[/quote]

Heh, okay, he's sure. And I'm beginning to understand. Apparently this is all explicit somewhere in the bowels of ECACHL law.

[Al DeFlorio]

[quote DeltaOne81][quote Trotsky][quote DeltaOne81]he's pretty sure[/quote]::looking::[/quote]

Heh, okay, he's sure. And I'm beginning to understand. Apparently this is all explicit somewhere in the bowels of ECACHL law.[/quote]
Can Q appeal to the Supreme Court?

[Trotsky]

Well, we're probably safe either way.  Either the top 5, then top 9 routine is in the rules, or since, everybody from the coaches to the announcers uses John's script to figure out tie breakers, it becomes a fait accomplis.

[Trotsky]

[quote Al DeFlorio]Can Q appeal to the Supreme Court?[/quote]That was exactly where I was going with the fait accomplis comment, above.

[jtwcornell91]

Okay, let's try the human rather than computer explanation.

Cornell and Quinnipiac are tied for 4th.  Head-to-head is a wash.  So, on to record vs top 4.  We each have 3 points vs the top 3, which is what we do for ties for 4th.  On to record vs top 8.  Oh crap, Colgate, RPI and Yale are tied for 8th.  Have to break that tie first.

Colgate, RPI, and Yale each split their season series with each other, so they have 4 points each head to head.  On to record vs top 4.  Oh crap, Cornell and Quinnipiac are still tied for 4th. ::panic::

The ECAC actually spelled out a couple of years ago how to handle this infinite loop.  So.  The first step is to do the best you can in defining record vs top 4 for the 8th-place tie.  For that purpose we expand "top 4" temporarily to include everyone tied for 4th place as well, i.e., the top 3 plus Cornell and Quinnipiac.  Colgate and RPI have 5 points against the top 5, but Yale only has 3.  So Yale gets 10th place.

We've broken a tie, so we start over, with Cornell/Quinnipiac still tied for 4th and Colgate/RPI still tied for 8th.  Cornell/Q still needs the top 8, and Colgate and RPI split their series head-to-head, so they need the top 4, and we're back in the infinite loop. ::panic::

Okay, we know from before that Colgate and RPI each have 5 points vs the top 5.  So the next step is to compare Cornell and Quinnipiac's records versus the "expanded top 8" i.e., the top 7 plus Colgate and RPI.  Cornell has 16 points, Quinnipiac 14, so Cornell wins the tiebreaker and gets 4th place! ::banana::

With the 4th place tie broken we can start over on Colgate and RPI; against the actual top 4, which we now know to be CCT, SLU, Dartmouth and Cornell, Colgate has 5 points and RPI 3.  So Colgate gets 8th.

With the top 4 and top 8 sorted out, it's straightforward to break the 6th-place tie between Harvard and Princeton.

[Trotsky]

That does make sense, actually.  But I still bet that if you coded your script to break infinite loops however it would be most advantageous to Cornell, nobody from the ECAC would ever catch it.

[billhoward]

If the ECACHL actually thought this out in advance, they're smarter than we thought, on this one point.

[Al DeFlorio]

[quote billhoward]If the ECACHL actually thought this out in advance, they're smarter than we thought, on this one point.[/quote]
Blind squirrel...

[DeltaOne81]

[quote billhoward]If the ECACHL actually thought this out in advance, they're smarter than we thought, on this one point.[/quote]

Apparently you can thank the guy who runs CollegeHockeyStats, who was the league statistician or something a several years ago and actually worked this all out.

On occasion you actually do get someone intelligent who knows what they're talking about.

[jtwcornell91]

[quote billhoward]If the ECACHL actually thought this out in advance, they're smarter than we thought, on this one point.[/quote]

Looking back, it seems that they first explained the solution to the public in 1998:

http://lists.maine.edu/cgi/wa?A2=ind9803&L=HOCKEY-L&P=R3770&I=-3

Now, that year the infinite loop was actually a possiblity going into the final weekend.  So it's possible they just made it up then.  I don't know how long the Top4/Top8 (or Top5/Top10) thing has been around, but I remember talking about the infinite loop as an abstract possibility in 1995.  Maybe the ECAC has always had an explanation of how to sort out these details, but didn't publicize it before 1998.