ECAC Byes and Home-Ice 2011-12

[burgie12]

Re: ECAC Byes and Home-Ice 2011-12

Any special reason for running 15,491 trials?

That's when I hit stop

My computer decided to be stupid and go into hibernate last night, that's why there weren't more.

[joecct]

Re: ECAC Byes and Home-Ice 2011-12

Here, I'll make it easy: Stay at the casino just outside Massena. You have a 40% chance of being there, with about the same chances of winning something.

I think the casino is on the wrong side (East) of Massena. Makes for a long drive to Potsdam.

lugnut, start checking out the hotels in the beautiful metropolis of Potsdam.

How much are you willing to spend for a hotel in the North Country? In Potsdam, all but one are road motels (the Smalling is pretty good). In Canton there are, I think, 3 chain motels all of which are (for the life of me why?) pricey.

[lugnut92]

Re: ECAC Byes and Home-Ice 2011-12

The hotel we go to will be up to the managers. I just wanted to know where I would be conducting. Unfortunately, my worst fears turned out to be true and we're most likely heading to Potsdam.

[FlagDUDE08]

Re: ECAC Byes and Home-Ice 2011-12

I think the casino is on the wrong side (East) of Massena. Makes for a long drive to Potsdam.

How much are you willing to spend for a hotel in the North Country? In Potsdam, all but one are road motels (the Smalling is pretty good). In Canton there are, I think, 3 chain motels all of which are (for the life of me why?) pricey.

It is east of Massena (in Franklin County actually), but it isn't too bad given Potsdam is south of Massena (33 miles according to Google Maps). Ogdensburg is about 20 miles or so from Canton, and that had some OK options the one time I drove through there. Since I grew up on the east side of the 'dacks, I would generally approach the area from the other direction.

[Professor]

Re: ECAC Byes and Home-Ice 2011-12

Good call. I originally had it as a three-way tie between Colgate, Yale, and (Clarkson / Harvard). Harvard wins that three-way tiebreaker and would push Yale down to 5th. Yale would finish second in a three-way tie that involved Clarkson (Clarkson loses the three-way on wins and then Colgate wins the head-to-head matchup against Yale on Points vs Top 4) and therefore grab the #4 seed. That was the original intent of my post and then I changed around the words and presented false info. Thank you for catching that.

Great work guys. If Yale sweeps it ends up with 23. If Colgate loses both (RPI and Cornell) it ends up with 23. If Clarkson ties Harvaadh but loses to Dartmouth it ends up with 23. If Harvaadh ties Clarkson and loses to St. Lawrence, it ends up with 22. If Quinnipiac loses to Yale and Yale beats Princeton the night before, Quinnipiac can only end up with 22. If St. Lawrence beats Haavadh but loses or ties Dartmouth, it ends up with 21 or 22. Under this scenario, you have Colgate, Clarkson and Yale tied at 23 and, if I understood correctly, Yale actually ends up with the bye with Colgate and Clarkson gets the 5th seed. I don't use computers, still on a slide rule. Does this make mathematical sense?

[LynahFan]

Re: ECAC Byes and Home-Ice 2011-12

Great work guys. If Yale sweeps it ends up with 23. If Colgate loses both (RPI and Cornell) it ends up with 23. If Clarkson ties Harvaadh but loses to Dartmouth it ends up with 23. If Harvaadh ties Clarkson and loses to St. Lawrence, it ends up with 22. If Quinnipiac loses to Yale and Yale beats Princeton the night before, Quinnipiac can only end up with 22. If St. Lawrence beats Haavadh but loses or ties Dartmouth, it ends up with 21 or 22. Under this scenario, you have Colgate, Clarkson and Yale tied at 23 and, if I understood correctly, Yale actually ends up with the bye with Colgate and Clarkson gets the 5th seed. I don't use computers, still on a slide rule. Does this make mathematical sense?

No - Colgate cannot lose to Cornell this weekend.

[Professor]

Re: ECAC Byes and Home-Ice 2011-12

No - Colgate cannot lose to Cornell this weekend.

As a Cornell fan how can you wish that Colgate would not not lose to Cornell at Lynah?

[LynahFan]

Re: ECAC Byes and Home-Ice 2011-12

As a Cornell fan how can you wish that Colgate would not not lose to Cornell at Lynah?

I wish Colgate could lose to Cornell EVERY night, but it's only a realistic possibility on the days they're actually playing, unfortunately.

[Professor]

Re: ECAC Byes and Home-Ice 2011-12

I am appropriately chastised and embarassed. I meant that Colgate were Colgate to lose RPI and Union (not Cornell). Cornell and Union have the lock - so the only play for the other two byes among, others, is with RPI, Colgate, Yale, Clarkson, Harvard, Dartmouth and Quinnipiac over the weekend and we need to see how all of these games play out

[FlagDUDE08]

Re: ECAC Byes and Home-Ice 2011-12

I am appropriately chastised and embarassed. I meant that Colgate were Colgate to lose RPI and Union (not Cornell). Cornell and Union have the lock - so the only play for the other two byes among, others, is with RPI, Colgate, Yale, Clarkson, Harvard, Dartmouth and Quinnipiac over the weekend and we need to see how all of these games play out

I know it was broken last year, but there's still that matter of the Starr curse with Union...

[LynahFan]

Re: ECAC Byes and Home-Ice 2011-12

I am appropriately chastised and embarassed. I meant that Colgate were Colgate to lose RPI and Union (not Cornell). Cornell and Union have the lock - so the only play for the other two byes among, others, is with RPI, Colgate, Yale, Clarkson, Harvard, Dartmouth and Quinnipiac over the weekend and we need to see how all of these games play out

I know - just messin' with ya. Thanks for being a good sport!

[Rainman]

Re: ECAC Byes and Home-Ice 2011-12

Great work guys. If Yale sweeps it ends up with 23. If Colgate loses both (RPI and Cornell) it ends up with 23. If Clarkson ties Harvaadh but loses to Dartmouth it ends up with 23. If Harvaadh ties Clarkson and loses to St. Lawrence, it ends up with 22. If Quinnipiac loses to Yale and Yale beats Princeton the night before, Quinnipiac can only end up with 22. If St. Lawrence beats Haavadh but loses or ties Dartmouth, it ends up with 21 or 22. Under this scenario, you have Colgate, Clarkson and Yale tied at 23 and, if I understood correctly, Yale actually ends up with the bye with Colgate and Clarkson gets the 5th seed. I don't use computers, still on a slide rule. Does this make mathematical sense?

In this scenario, Colgate, Clarkson, and Yale would all be tied for 3rd place at 23 points. Colgate would win the tiebreaker and the #3 seed based on record vs. top-4 (their 2 wins vs. Cornell would give them the tiebreak), then the tie-breaking procedures start over again between Yale and Clarkson. Yale wins this on total wins and gets the #4 seed; Clarkson would get the #5 seed.

[RHamilton]

Re: ECAC Byes and Home-Ice 2011-12

Here's all 531,441 trials, first weighted by KRACH, then by straight ## of scenarios. Using Burgie's 13.2% tie percentage.

Union               0.686183     0.313817            0            0            0            0            0            0            0            0            0            0
Cornell             0.313817     0.686183            0            0            0            0            0            0            0            0            0            0
Colgate                    0            0     0.641162     0.293867     0.048017    0.0169531            0            0            0            0            0            0
Clarkson                   0            0     0.162648     0.325268     0.232337     0.225301    0.0485888   0.00585654            0            0            0            0
Harvard                    0            0     0.184376     0.219664     0.200154     0.195617     0.115894    0.0748487   0.00944698            0            0            0
Quinnipiac                 0            0    0.0110722     0.132703     0.299187     0.268549      0.15904     0.121415   0.00803449            0            0            0
St.Lawrence                0            0  0.000741505    0.0200543     0.118443     0.159924     0.376225     0.148108     0.176505            0            0            0
Yale                       0            0            0   0.00844376     0.101668     0.125353     0.236666     0.473911    0.0539584            0            0            0
Dartmouth                  0            0            0            0  0.000193787   0.00830284    0.0635866     0.175861     0.641053     0.109482   0.00152056            0
Princeton                  0            0            0            0            0            0            0            0     0.109657     0.697331     0.183391   0.00962097
RPI                        0            0            0            0            0            0            0            0   0.00134537    0.0101844     0.500546     0.487924
Brown                      0            0            0            0            0            0            0            0            0     0.183003     0.314542     0.502455
Union            413343  118098       0       0       0       0       0       0       0       0       0       0
Cornell          118098  413343       0       0       0       0       0       0       0       0       0       0
Colgate               0       0  360612  140496   27822    2511       0       0       0       0       0       0
Clarkson              0       0   90396  210681  131706   86751   10935     972       0       0       0       0
Harvard               0       0   77802  141303  130086   88857   65691   25272    2430       0       0       0
Quinnipiac            0       0    2430   27702  111780  157950  109107  113530    8942       0       0       0
St.Lawrence           0       0     201    5265   75291  100440  182504   80503   87237       0       0       0
Yale                  0       0       0    5994   54513   89829  125955  226719   28431       0       0       0
Dartmouth             0       0       0       0     243    5103   37249   84445  369409   32805    2187       0
Princeton             0       0       0       0       0       0       0       0   32805  384912   89667   24057
RPI                   0       0       0       0       0       0       0       0    2187   28431  301806  199017
Brown                 0       0       0       0       0       0       0       0       0   85293  137781  308367

Interestingly, no ties will be broken based on goals and there's just a 1% chance that ties will be broken based on record vs. Top 8 teams.

EDIT: Source, for anyone curious. (Required files: team_stats.txt, games.txt)

[burgie12]

Re: ECAC Byes and Home-Ice 2011-12

RHamilton, what about the following scenario?
Qpac loses out, Dartmouth gains 3 points, SLU finishes in the top 8, Colgate and Harvard grab the final two bye spots?

By my spreadsheet, that tiebreaker goes down to Goals vs Top 4.

[RHamilton]

Re: ECAC Byes and Home-Ice 2011-12

RHamilton, what about the following scenario?
Qpac loses out, Dartmouth gains 3 points, SLU finishes in the top 8, Colgate and Harvard grab the final two bye spots?

By my spreadsheet, that tiebreaker goes down to Goals vs Top 4.

Yep, found the bug. I was accidentally using old h2h values from the current standings, rather than final standings.

I now see 3,267 scenarios where tiebreaks are settled with goals, which makes more sense. I updated my previous post, randomly determining the outcome of anything based on goal differential. I'm not interested in storing goals scored, nor do I want to predict goals scored.