Wow!!!!!!!

[melissa]

yup. moguls. was actually thinking about it this morning and the few cornellians present in the olympics have fared much better than avg. hope nieuwy gets gold. that'd be 3 golds and 1 silver out of 5 chances. not too shabby at all.

[Ben Doyle 03]

Let's all hope for a Canada/USA gold medal game. . .that would be pretty neat for both teams, there would be basically a home crowd for everyone.:-D

[Greg Berge]

> The system will award points based on completion of elements with attention to degree of difficulty.

Of course, this is already supposed to be the basis of the technical marks.  Generally speaking (I know, the pairs was an exception), the technical marks isn't where the controversy occurs.


> In order to make this more objective and reduce possible vote trading there will now be 14 judges allocating points but only 7 (randomly chosen by computer) will count.

Sounds awful.  In a close contest, the computer's randomization will effectively determine the winner.  That's hardly fair.

I would prefer that they extend the concept of dropping the high and low scores.  Take the current 9 judges.  For each set of marks (tech and presentation), drop the top 2 scores and the bottom 2 scores.  Take the mean of the remaining 10 scores.  Get rid of the ordinals -- simply score by that mean.

What is wrong with that simple solution?

[Al DeFlorio]

Yep, this would be real progress, but one of the keys is getting rid of the ordinals.  Last night, if all the skaters besides Hughes had crawled around the ice on all fours for four minutes, and if Kwan had been judged to have crawled the best by the majority of the judges, she would still have won the gold medal based on ordinals (i.e., she would have finished second in the long program and ended with 2.5 ordinals to Sarah's 3.0).  

Even the ECAC playoff system isn't this stupid. ::nut::

[Greg Berge]

The ECAC playoff system makes the #4 and #5 skaters from the short program skate a run off immediately before skating again in the long program.

[melissa]

that is my hopes. that way i'll be happy no matter who wins (tho  i DO have a preference!):-D

[jeh25]

Greg Berge wrote:

> In order to make this more objective and reduce possible vote trading there will now be 14 judges allocating points but only 7 (randomly chosen by computer) will count.

Sounds awful.  In a close contest, the computer's randomization will effectively determine the winner.  That's hardly fair.

Greg-

I think you are missing the point. In a blind random system, any horse trading for votes is *immediately* eliminated.  I'm not going to try to trade votes with you if I can't be sure you actually followed through on our deal.

If someone better versed in history can correct me, please do, but I seem to remember that the US moved to the secret ballot system in the mid-1890's to prevent  people like Boss Tweed from buying votes.  If a buyer cannot confirm that the seller actually voted as promised, the entire vote fixing scheme falls apart.

-j

[Jordan 04]

"That's just sour grapes," said Dan Tomaselli, a high school music teacher in Great Neck, N.Y. "The Russians, of all people, should keep their mouths shut."

Uh-oh.

The Russians have gone and done it.  They've pissed off Mr. Tomaselli, and you do NOT want to get on his bad side!!

May they feel the wrath!

"Get....in...my....belly!" ::worry::

[CUlater \'89]

So are you saying that you prefer having the short program count the same as the long program?

[CUlater \'89]

It's interesting to note that when Sarah performed poorly during the short program (and even before skating) there was some criticism of her decision to not stay in the Salt Lake City area earlier, similar to the criticism of Kwan at Nagano.  When Kwan won the silver in Nagano, it was blamed in part on her decision to live offsite etc.  

Now that Hughes has won the gold, what happened to all the naysayers who said she blew it by her choice of living arrangements?  Maybe the experts should wake up and realize that where you stay or where you practice or whether you partake of the Olympic experience is mostly irrelevant and that on some days you have it, and some days you don't.  Unfortunately for Michelle Kwan (and Slutskaya), she just didn't have it yesterday.

[Greg Berge]

> Greg- I think you are missing the point. In a blind random system, any horse trading for votes is *immediately* eliminated. I'm not going to try to trade votes with you if I can't be sure you actually followed through on our deal.

Nope, I think you are missing the point.  ("No, you."  "No, you!"  "Oh yeah... step outside.")  I'll explain.

Whether vote trading will or will not disappear is beside my point (although you are also not correct in saying that randomization guarantees the end of vote trading, because if the odds of having your own vote count are reduced by the same ratio as the person with whom you are negotiating, then an attempted deal is still attractive -- that is how the prisoner's dilemma problem in game theory resolves to the negotiators' maximum advantage).

But my point isn't about vote trading, it's about how winners get picked.  Random selection of results will make ranking close competitors a total crap shoot.  The slight variance in the range of votes will be overwhelmed by the large variance in the potential outcomes of judge selection.

(Of course, it's perfectly valid to say that close competitions are inherently not resolvable because there's such a high degree of subjectivity, but I think I'd rather have an artificial human resolution than a coin flip).

[Al DeFlorio]

Of course not.  Why would you think that?  Simply weigh the long program twice as much in accumulating total points from the two programs.  

Just get rid of ordinals.  What makes no sense is having the outcome depend on whether Slutskaya finished between Hughes and Kwan in the long program (Hughes wins) as opposed to finishing below both (Kwan wins).  What does that have to do with the relative merits of Hughes's and Kwan's skating both nights.

[CUlater \'89]

How does throwing out two more scores fix a problem?

[Greg Berge]

It minimizes the effect of vote trading and unilateral vote skewing.  I'm the US judge and you're the French judge (sorry).  I have a #2 world ranked US woman (#1 is a Russian) in the womens and you have a #2 world ranked French team  (#1 are Jamaican -- but they're really good) in the pairs.

So we cut a deal.  I boost the French pair by .2 points and cut the Jamaicans down by .2 points, while you boost the US woman .2 points and cut the Russian woman .2 points.

As of now, we really do affect the result.  But if you cut the high and low pair of scores, if all of our scores are out of line with the other judges, then our inflated and defalted scored miss the cut when the high and low scores are tossed out.  The result is that the skaters with the best scores among the honest judges will win.  If it turns out our scores aren't out of line with the judges and other scores are cut, then nothing was lost -- our "fix" just happened to accidentally reflect the real quality on the ice.

And like Al says, you double-count the long program mean: S+2L, when you compute the skaters final score.

[min \'97]

sorry to get off-topic, but i just watched a video clip on cnn.com and it was reported that the russians had filed a complaint to ISU and they demanded that another gold medal be awarded to irina.

are they serious?