Talk:Random ballot

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

The math is wrong. 0.01^26 would the be probability that all 26 members of a 26 member body were from the 1% minority group. It doesn't take into account the other 24 seats in the congress. — Preceding unsigned comment added by 173.87.209.54 (talk) 20:16, 3 July 2016 (UTC)[reply]

The probability of a majority is 1 - F(25) where F is the binomial cdf with n = 50 and p = 0.01. The current math on the page is incorrect for the case of exactly 26 seats, let alone 26+.

216.171.8.23 (talk) 08:45, 3 November 2016 (UTC)[reply]

Are there any examples of this method being used by any organisation? LukeSurl

I was going to ask the same. Is this a purely academic concept? Wouter Lievens 15:03, 18 Mar 2005 (UTC)
Presumably an example of a random ballot (with 2 voters and 2 alternatives) would be the coin toss at the start of a sporting event which determines which of the two team captains decides whether to play first or second...? --213.232.66.5 00:37, 7 May 2005 (UTC)[reply]

"it undermines majority rule since there is a substantial possibility that the selected voter may be in the minority."

Well yes, but this is less likely to be a problem if the method is used to select memebers of a large group, rather than a single voter. Should this article cover this case, or is that a topic for another article? See also Voting system. This is the method used to select the Council of 500 prytanies in the Athenian democracy. ABostrom 22:30, July 14, 2005 (UTC)

It could be mentioned that this is the most proportional form of democracy you could ever get. In that, a party with 5% of the vote may get 5% of the seats, but that does not necessarily translate into 5% of the power - if able to hold the balance of power and gain significant concessions it could be more than this, but if only a tag-along to a major party, it could be much less power. In random ballots, a party with 5% of the vote has a 5% chance of power - exactly proportional. That does not make it necessarily desirable. -Nichlemn 02:12, 6 August 2006 (UTC)[reply]

Single Stochastic Vote[edit]

Random ballot was promoted - under the name "Single Stochastic Vote" (a parody of Single Transferable Vote) by Henry Potts in the Usenet group uk.politics.electoral around the time of the 1997 General Election. A search of the google archives would probably be productive —The preceding unsigned comment was added by 82.42.213.163 (talk) 18:33, 15 April 2007 (UTC).[reply]

Swedish football betting[edit]

When one of the 13 football games of the Swedish toto ("Stryktipset") is cancelled (mostly due to bad weather), the "outcome" of that game is decided by a random ballot. Ten daily newspapers have before the cancellation tried to foresee the result, and each guess is represented by one ballot. To these 10 ballots are added 2 ballots for each possible result (home win, draw or away win), for a total of 16 of which 1 is drawn. Although it is not a political election, it could be seen as a modified random ballot method... Fomalhaut76 (talk) 11:53, 20 December 2008 (UTC)[reply]

Manipulability when used in a multi-winner district[edit]

Dear Bmcollier,

the article says: "It is also strategy-free in that there is no advantage in tactical voting. (...) If the Random Ballot is used to select the members of a multi-constituency body, it can serve to retain the attractive features of both First Past the Post and Proportional Representation."

However, when random ballot is used in a multi-winner district, then it isn't strategyproof anymore.

Example:

Suppose there are 100,000 voters and 100 seats. Suppose my favorite candidate will get signifiantly more than 1000 votes, but my second favorite candidate will get only a handful of votes. Then it is a useful strategy to give my vote to my second favorite candidate, as my favorite candidate will be elected by some other voter with a high probability.

The only way, how random ballot in a multi-winner district could be strategyproof, is when a ballot is chosen randomly and the 100 favorite candidates on this ballot are declared elected. But this wouldn't be proportional anymore. Markus Schulze 18:23, 26 May 2014 (UTC)[reply]

Hi Markus Schulze

That makes sense, I'll amend accordingly. Bmcollier (talk)

There is a solution to that: either give candidates who get more than one random ballot extra voting rights in the representative body or allow them to appoint friends/clones as additional representatives. --Rumping (talk) 20:33, 24 July 2014 (UTC)[reply]
The form of random ballot that's appropriate for a multiple-winner election is a ranked-choice random ballot: if there are N seats open, each voter lists N choices in order of preference. Then N random ballots are drawn, and for each one, the highest-ranked candidate that hasn't already been seated is seated. This is still strategy-free. 107.203.108.56 (talk) 18:21, 13 April 2015 (UTC)[reply]

Random dictatorship[edit]

This voting rule seems to be equivalent to "random dictatorship" and is much older than 1984. Random dictatorship has been analyzed in dozens of academic papers at least since 1977 (Gibbard: "Manipulation of schemes that mix voting with chance", Econometrica 45(3)). The article should be renamed and rewritten accordingly. — Preceding unsigned comment added by Flix~enwiki (talkcontribs) 18:19, 21 May 2015 (UTC)[reply]

I would disagree. The major difference is that dictatorship is a form of government where as random ballot is a system of voting. Also - "Random Ballot" is the common name referring to it in reliable sources, and therefore should be the name per Wikipedia title policy. CoffeeGiraffe (talk) 14:39, 31 October 2015 (UTC)[reply]
Yes, it's the same thing. Added to the article. — Omegatron (talk) 19:00, 12 October 2019 (UTC)[reply]

Poor example on page[edit]

The example of proportionality given is poor. No one examining this system would be seriously worried about a 1% party getting a majority. What they would be worried about would be more like real world issues: 1. A party with 48% beating a party with 52% in a 600 seat election 2. A neo-nazi party with 0.5% of the vote in every seat getting 1 or 2 seats Can we have the probabilities of these please? Fig (talk) 19:02, 17 September 2018 (UTC)[reply]