Ice-hockey update

In the previous article about ice-hockey, another piece of evidence that the Bayesian reasoning is flawed was offered by the second anonymous "fast commenter" who has made the following prediction:

• The Bayesian probability of [Switzerland] winning gold is definitely higher now, whereas a frequentist wouldn't be allowed to put his two cents into the debate

Indeed, I said, the frequentist approach to probability does not allow us to say much, and if it does - and if it is combined with common sense - the result will be, of course, that the probability of Switzerland getting the gold is tiny. Most likely, Switzerland would be eliminated in the quarterfinals because it used to be frequently eliminated. ;-)

Needless to say, the Bayesian inference has been proved wrong once again and I was right - despite several fascinating previous matches, Switzerland has lost to Sweden 2:6. Canada won't defend its 2002 olympic gold because it lost to Russia 0:2. Finland defeated the U.S. 4:3.

The last quarterfinal match guaranteed that exactly one Czechoslovak team would make it to the semifinals. Slovakia has won all five matches at the beginning of the olympic tournament while Czechia was only able to beat two underdogs, namely Italy and Germany. Nevertheless, it seems that the Slovak players still view the Czechs as "older brothers" who are supposed to win: Czechia remains the most likely team to beat Slovakia. So eventually the Slovaks lost 1:3.

The previous link also explains that the Czechs had to play with the goalie #3 Milan Hnilička because the goalie #1 Dominik Hašek returned home with an injury while the #2 goalie Tomáš Vokoun was identified, much like loop quantum gravity, to be inconsistent.

In the semifinals, Czechia will play Sweden and Russia will face Finland. Czechia is the only surviving team in the tournament that will defend the glory of North American ice-hockey. The previous sentence should settle the question whom the citizens of the 1st and 3rd most civilized country in the world, according to the data in the right column of this blog, should root for. ;-)