What category will the Google top Canadian News Story of 2016 belong to?
Politics/Elections: 25%
Terrorism/Crime: 20%
Sports/Entertainment: 20%
Business/Economics: 10%
Some Other Category: 25%
(see the "News Stories" list here for the 2015 stories. If Google changes their categories, a similar category will be used, if available; if not, this question will be voided.)
Of the top news stories of 2015, 3 were Sports/Entertainment stories, 2 were Politics/Elections, 2 were Terrorism/Crime, and 3 were Business/Economics. I think 2015 was an unusually good year for Sports/Entertainment stories. The US Presidential Election and the Olympics are obvious contenders for biggest story of 2016. An major terrorist attack could also become the biggest story. Business/Economics rarely seem to win in this sort of competition.
My classifications for 2015 were as follows:
Blue Jays Playoff Run: Sports/Entertainment
Federal Election: Politics/Elections
Pan Am Games: Sports/Entertainment
Paris Attacks: Terrorism/Crime
Charlie Hebdo Attack: Terrorism/Crime
Mayweather Pacquiao Fight: Sports/Entertainment
Donald Trump Candidacy: Politics/Elections
Greece Debt Crisis: Business/Economics
Oil Prices: Business/Economics
Ashley Madison Leak: Business/Economics
Showing posts with label Meta. Show all posts
Showing posts with label Meta. Show all posts
Saturday, 2 January 2016
Friday, 1 January 2016
Top Global News 2016
What category will the Google top Global News Search of 2016 belong to?
Terrorism/Crime: 30%
Politics/Elections: 25%
Disasters/Diseases: 20%
Riots/Unrest: 10%
Some Other Category: 15%
(see the "Global News" list here for the list for 2015. If Google changes their categories, a similar category will be used, if available; if not, this question will be voided.)
In 2015, the top global news search was for "Charlie Hebdo" which counts as Terrorism/Crime according to my categories. In 2014, the top search was "Ebola", which counts as Disasters/Diseases. Of the 18 topics listed in the last two years, 7 have been in Terrorism/Crime, 2 in Politics/Elections, 5 in Disasters/Diseases, 3 in Riots/Unrest, and 1 in Some Other Category. I'm going to give the highest weight to Terrorism/Crime, based on the historical data. Because of the US Election, I'll give Politics/Elections a higher weight than the past history numbers would suggest. I'll also overweight Some Other Category, because to some extent I chose my categories to fit all of the stories, so the historical data probably underestimates Some Other Category.
My classifications for 2015 were as follows:
Charlie Hebdo: Terrorism/Crime
Paris: Terrorism/Crime
Hurricane Patricia: Disasters/Diseases
Isis: Terrorism/Crime
Nepal: Disasters/Diseases
El Chapo: Terrorism/Crime
Greece: Politics/Elections
Baltimore Riots: Riots/Unrest
San Bernardino: Terrorism/Crime
Hurricane Joaquin: Disasters/Diseases
For 2014,
Ebola: Disasters/Diseases
ISIS: Terrorism/Crime
Malaysia Airlines: Disasters/Diseases
Crimea / Ukraine: Riots/Unrest
Ferguson: Riots/Unrest
Gaza and Israel: Some Other Category
Scotish Referendum: Politics/Elections
Oscar Pistorius trial: Terrorism/Crime
Terrorism/Crime: 30%
Politics/Elections: 25%
Disasters/Diseases: 20%
Riots/Unrest: 10%
Some Other Category: 15%
(see the "Global News" list here for the list for 2015. If Google changes their categories, a similar category will be used, if available; if not, this question will be voided.)
In 2015, the top global news search was for "Charlie Hebdo" which counts as Terrorism/Crime according to my categories. In 2014, the top search was "Ebola", which counts as Disasters/Diseases. Of the 18 topics listed in the last two years, 7 have been in Terrorism/Crime, 2 in Politics/Elections, 5 in Disasters/Diseases, 3 in Riots/Unrest, and 1 in Some Other Category. I'm going to give the highest weight to Terrorism/Crime, based on the historical data. Because of the US Election, I'll give Politics/Elections a higher weight than the past history numbers would suggest. I'll also overweight Some Other Category, because to some extent I chose my categories to fit all of the stories, so the historical data probably underestimates Some Other Category.
My classifications for 2015 were as follows:
Charlie Hebdo: Terrorism/Crime
Paris: Terrorism/Crime
Hurricane Patricia: Disasters/Diseases
Isis: Terrorism/Crime
Nepal: Disasters/Diseases
El Chapo: Terrorism/Crime
Greece: Politics/Elections
Baltimore Riots: Riots/Unrest
San Bernardino: Terrorism/Crime
Hurricane Joaquin: Disasters/Diseases
For 2014,
Ebola: Disasters/Diseases
ISIS: Terrorism/Crime
Malaysia Airlines: Disasters/Diseases
Crimea / Ukraine: Riots/Unrest
Ferguson: Riots/Unrest
Gaza and Israel: Some Other Category
Scotish Referendum: Politics/Elections
Oscar Pistorius trial: Terrorism/Crime
Saturday, 17 October 2015
Conservative Polling Error
What will the difference be between the national Conservative vote total and the final poll average at threehundredeight.com?
Greater than 2%: 35%
Between 0% and 2%: 30%
Between -2% and 0%: 20%
Less than -2%: 15%
(A positive difference means that the Conservatives have outperformed the poll average. The final poll average is the one that takes into account all polling done before election day, but does not take into account exit polling or actual returns.)
The vote total for the conservative party has been somewhat underestimated in some recent elections (the most notable being the recent UK election). In the past 4 Canadian elections, the Conservative vote total has been underestimated by 2.5% (2011), 2.9% (2008). -0.8% (2006), and -0.9% (2004). The errors in the 2008 and 2011 elections were much larger than the statistically calculated error.
Results in previous years suggest that the final poll average is likely to underestimate the Conservative vote total, possibly substantially. However, this might not happen to the extent that it has happened in previous elections.
Greater than 2%: 35%
Between 0% and 2%: 30%
Between -2% and 0%: 20%
Less than -2%: 15%
(A positive difference means that the Conservatives have outperformed the poll average. The final poll average is the one that takes into account all polling done before election day, but does not take into account exit polling or actual returns.)
The vote total for the conservative party has been somewhat underestimated in some recent elections (the most notable being the recent UK election). In the past 4 Canadian elections, the Conservative vote total has been underestimated by 2.5% (2011), 2.9% (2008). -0.8% (2006), and -0.9% (2004). The errors in the 2008 and 2011 elections were much larger than the statistically calculated error.
Results in previous years suggest that the final poll average is likely to underestimate the Conservative vote total, possibly substantially. However, this might not happen to the extent that it has happened in previous elections.
Friday, 25 September 2015
Good Judgement Closure
By the end of 2016, will the end of the Good Judgement Open Tournament be announced?
Yes: 25%
(If the tournament is actually closed by the end of 2016, then this question will resolve as a yes. If the announcement of the closure of the tournament is conditional (e.g. on number of forecasters) then this question will close if the tournament does actually close at the time specified. Announcements that the tournament will close after the end of 2017 do not count for this question. Temporary interruptions before the end of 2016 are also not sufficient.)
The Good Judgment Open Tournament is a recently launched forecasting tournament, based on the Good Judgment Project (see site here, and Wikipedia article here). There have been no announcements about how long the tournament will last, but no questions currently in the tournament close after March 1st, 2017 (and only one question closes that far in the future).
The success of the tournament may depend on the success of the parent company, Good Judgment Inc. If the parent company is unable to find business, the tournament might close down fairly quickly. Good Judgment Inc. does not appear to be dependant on a single source of funding, which means it is not as vulnerable as it might be. Still, the future of the tournament is fairly uncertain; it is quite possible that it will close down after the current set of questions end.
Yes: 25%
(If the tournament is actually closed by the end of 2016, then this question will resolve as a yes. If the announcement of the closure of the tournament is conditional (e.g. on number of forecasters) then this question will close if the tournament does actually close at the time specified. Announcements that the tournament will close after the end of 2017 do not count for this question. Temporary interruptions before the end of 2016 are also not sufficient.)
The Good Judgment Open Tournament is a recently launched forecasting tournament, based on the Good Judgment Project (see site here, and Wikipedia article here). There have been no announcements about how long the tournament will last, but no questions currently in the tournament close after March 1st, 2017 (and only one question closes that far in the future).
The success of the tournament may depend on the success of the parent company, Good Judgment Inc. If the parent company is unable to find business, the tournament might close down fairly quickly. Good Judgment Inc. does not appear to be dependant on a single source of funding, which means it is not as vulnerable as it might be. Still, the future of the tournament is fairly uncertain; it is quite possible that it will close down after the current set of questions end.
Sunday, 5 July 2015
Scicast Re-Opening
Q172: Before the end of 2015, will SciCast Predict re-open?
Yes: 60%
(SciCast must reopen as an active platform, not just a data repository. Substantial changes in the platform are acceptable, as long as it remains a freely accessible science and technology forecasting site)
SciCast Predict, the science and technology forecasting platform, has gone on "hiatus" after losing funding from its major sponsor, IARPA (see post here). They seem to be in discussions with a new sponsor, and hope (and expect) to restart in the fall.
I think that SciCast probably will re-open, but this is far from a sure thing. In this sort of situation, organizations often have a bit of an optimistic bias (the canonical example is a restaurant "closed for renovations"). Their last post does have a bit wrap-up feel to it that is not entirely consistent with a temporary closure of the site.
Yes: 60%
(SciCast must reopen as an active platform, not just a data repository. Substantial changes in the platform are acceptable, as long as it remains a freely accessible science and technology forecasting site)
SciCast Predict, the science and technology forecasting platform, has gone on "hiatus" after losing funding from its major sponsor, IARPA (see post here). They seem to be in discussions with a new sponsor, and hope (and expect) to restart in the fall.
I think that SciCast probably will re-open, but this is far from a sure thing. In this sort of situation, organizations often have a bit of an optimistic bias (the canonical example is a restaurant "closed for renovations"). Their last post does have a bit wrap-up feel to it that is not entirely consistent with a temporary closure of the site.
Subscribe to:
Posts (Atom)