Majority Judgment: Difference between revisions

Removed multiwinner category as it's not a multiwinner method.
No edit summary
(Removed multiwinner category as it's not a multiwinner method.)
 
(17 intermediate revisions by 6 users not shown)
Line 1:
{{Wikipedia}}
'''Majority Judgment''' is a single-winner [[voting system]] proposed by Michel Balinski and Rida Laraki. Voters freely grade each candidate in one of several named ranks, for instance from "excellent" to "bad", and the candidate with the highest [[median]] grade is the winner. If more than one candidate has the same median grade, a tiebreaker is used which sees how "broad" that median grade is. Majority Judgment can be considered as a form of [[Bucklin voting]] which allows equal ranks.
 
'''Majority Judgment''' is a single-winner [[voting system]] proposed by [[Michel Balinski]] and Rida Laraki.<ref>{{cite book|author= M. Balinski & R. Laraki|year=2010|title=Majority Judgment. |publisher=MIT Press|isbn=978-0-262-01513-4}}</ref><ref>{{Cite journal|last=de Swart|first=Harrie|date=2021-11-16|title=How to Choose a President, Mayor, Chair: Balinski and Laraki Unpacked|url=https://link.springer.com/10.1007/s00283-021-10124-3|journal=The Mathematical Intelligencer|language=en|doi=10.1007/s00283-021-10124-3|issn=0343-6993}}</ref> Voters freely grade[[Score voting|score]] each candidate in one of several named ranksqualities, for instance from "excellent" to "bad",. Each quality is associated with a numeric score and the candidate with the highest [[median]] gradescore is the winner. If more than one candidate has the same median grade, a tiebreaker is used which sees how "broad" that median grade is. Majority Judgment can be considered as a form of [[Bucklin voting]] which allows equal ranks.
 
==Voting process==
Line 9 ⟶ 11:
 
== Satisfied and failed criteria ==
Majority Judgment voting satisfies the [[Majority criterion for rated ballots|majority criterion]] for rated ballots]], and also a weak form of the [[mutual majority criterion]] (a majority giving only and all of their preferred candidates perfect grades will win), the [[monotonicity criterion]], [[reversal symmetry]], and [[later-no-harm|later-no-help]]. Assuming that ratings are given independently of other candidates, it satisfies the [[independence of clones criterion]] and the [[independence of irrelevant alternatives|independence of irrelevant alternatives criterion]]<ref>Badinski and Laraki, ''Majority Judgment'', p. 217</ref> - although this latter criterion is incompatible with the majority criterion if voters shift their judgments in order to express their [[preferential voting|preferences]] between the available candidates.
 
It fails the [[Condorcet criterion]],<ref group="nb">Strategically in the [[strong Nash equilibrium]], MJ passes the Condorcet criterion.</ref> [[later-no-harm]],<ref group="nb">MJ provides a weaker guarantee similar to LNH: rating another candidate at or below your preferred winner's median rating (as opposed to your own rating for the winner) cannot harm the winner.</ref> [[consistencyConsistency criterion for voting systems|consistency]], the [[Condorcet method|Condorcet loser criterion]],<ref group="nb">Nevertheless, it passes a slightly weakened version, the majority condorcet loser criterion, in which all defeats are by an absolute majority (for instance, if there aren't equal rankings).</ref> and the [[participation criterion]].<ref group="nb">It can only fail the participation criterion when, among other conditions, the new ballot rates both of the candidates in question on the same side of the winning median, and the prior distribution of ratings is more sharply-peaked or irregular for one of the candidates.</ref> It also fails the ranked or preferential [[majority criterion]], which is incompatible with the passed criterion [[independence of irrelevant alternatives]], and [[reversal symmetry]].
 
==Example application==
Line 17 ⟶ 19:
 
If there were four ratings named "Excellent", "Good", "Fair", and "Poor", and each voter assigned four different ratings to the four cities, then the sorted scores would be as follows:
 
 
<table>
 
<tr>
<td align=right>City&nbsp;&nbsp;&nbsp;</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td width=49%>&nbsp;</td>
<td width=2% textalign=center>\/</td>
<td width=49%>Median point</td>
</tr>
Line 37 ⟶ 38:
<td align=right>Nashville</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td bgcolor=green width=26%>&nbsp;</td>
<td bgcolor=YellowGreen width=42%></td>
<td bgcolor=Gold width=32%></td>
<td bgcolor=red width=0%></td>
</tr>
</table>
Line 51:
<td align=right>Chattanooga</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td bgcolor=green width=15%>&nbsp;</td>
<td bgcolor=YellowGreen width=43%></td>
<td bgcolor=Gold width=42%></td>
<td bgcolor=red width=0%></td>
</tr>
</table>
Line 65 ⟶ 64:
<td align=right>Knoxville</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td bgcolor=green width=17%>&nbsp;</td>
<td bgcolor=YellowGreen width=15%></td>
<td bgcolor=Gold width=26%></td>
<td bgcolor=redDarkOrange width=42%></td>
</tr>
</table>
Line 79 ⟶ 78:
<td align=right>Memphis</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td bgcolor=green width=42%>&nbsp;</td>
<td bgcolor=YellowGreenDarkOrange width=058%></td>
<td bgcolor=Gold width=0%></td>
<td bgcolor=red width=58%></td>
</tr>
</table>
Line 98 ⟶ 95:
<td>&nbsp;</td>
<td>
<table cellpadding=1 width=600 border=0 cellspacing=1>
<tr>
<td width=25%>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</td>
<td bgcolor=green>&nbsp;</td>
<td>&nbsp;ExellentExcellent&nbsp;&nbsp;</td>
<td bgcolor=YellowGreen>&nbsp;</td>
<td>&nbsp;Good&nbsp;&nbsp;</td>
<td bgcolor=Gold>&nbsp;</td>
<td>&nbsp;Fair&nbsp;&nbsp;</td>
<td bgcolor=redDarkOrange>&nbsp;</td>
<td>&nbsp;Poor&nbsp;&nbsp;</td>
<td width=50%>&nbsp;</td>
</tr>
</table>
Line 118 ⟶ 114:
</table>
 
The mediansmedian rating for Nashville and ChatanoogaChattanooga would beis "Good"; for Knoxville, "Fair"; and for Memphis, "Poor". Nashville and ChatanoogaChattanooga are tied, so "Good" ratings wouldhave to be removed from both. After removing 16% of the votes from each, Chatanooga'suntil mediantheir wouldmedians roundbecome downdifferent. to "Fair", while Nashville would remain at "Good"<ref>After removal,removing Chatanooga would have 42% of the initial electorate at "Fair", 2716% "Good", andratings 15%from "Excellent",the whilevotes Nashvilleof would have 32% "Fair"each, 26%the "Good",sorted andratings 26%are "Excellent"</ref> and so '''Nashville''' would win.now:
 
<table>
 
<tr>
<td align=right>City&nbsp;&nbsp;&nbsp;</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td width=49%>&nbsp;</td>
<td width=2% textalign=center>\/</td>
<td width=49%>Median point</td>
</tr>
Line 138 ⟶ 134:
<td align=right>Nashville</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td bgcolor=gray width=8%>&nbsp;</td>
Line 153 ⟶ 149:
<td align=right>Chattanooga</td>
<td>
<table cellpadding=0 width=600500 border=0 cellspacing=0>
<tr>
<td bgcolor=gray width=8%></td>
Line 173 ⟶ 169:
<td>&nbsp;</td>
<td>
<table cellpadding=1 width=600 border=0 cellspacing=1>
<tr>
<td width=20%>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</td>
<td bgcolor=gray>&nbsp;</td>
<td>&nbsp;Removed&nbsp;ratings&nbsp;(sorted&nbsp;to&nbsp;both&nbsp;ends&nbsp;evenly&nbsp;for&nbsp;easy&nbsp;comparison&nbsp;of&nbsp;medians&nbsp;with&nbsp;above).</td>
<td width=20%>&nbsp;</td>
</tr>
</table>
</td>
</table>
 
Chattanooga now has the same number of "Fair" ratings as "Good" and "Excellent" combined, so its median is rounded down to "Fair", while Nashville's median remains at "Good"<ref group="nb">After removal, Chattanooga has 42% of the initial electorate at "Fair", 27% "Good", and 15% "Excellent", while Nashville has 32% "Fair", 26% "Good", and 26% "Excellent"</ref> and so '''Nashville''', the capital in real life, wins.
 
If voters from Knoxville and Chattanooga were to rate Nashville as "Poor" and/or both sets of voters were to rate Chattanooga as "Excellent", in an attempt to make their preferred candidate Chattanooga win, the winner would still be Nashville.
 
==Variants==
 
Variants of majority judgment have been described. Fabre considers three: the typical judgment, usual judgment, and central judgment.<ref name="Fabre20">{{Cite journal |first=Adrien |last=Fabre |title=Tie-breaking the Highest Median: Alternatives to the Majority Judgment |journal=[[Social Choice and Welfare]] |year=2020 |volume=56 |pages=101–124 |url=https://github.com/bixiou/highest_median/raw/master/Tie-breaking%20Highest%20Median%20-%20Fabre%202019.pdf |doi=10.1007/s00355-020-01269-9|issn=0176-1714}}</ref>. He argues that all of these are less sensitive to noise than the majority judgment, with the usual judgment being the most robust, though the calculation that determines the winner is more complex.
 
==See also==
 
* [[List of democracy and elections-related topics]]
* [[Voting system]]
* [https://www.youtube.com/watch?v=ZoGH7d51bvc&t=917s Reforming the presidential election! (in French)]
== Notes ==
{{reflist|group=nb}}
{{Reflist}}
 
== References ==
{{Reflist}}
*Balinski, Michel, and Laraki, Rida (2010). ''Majority Judgment: Measuring, Ranking, and Electing'', MIT Press
 
{{DEFAULTSORT:Bucklin Voting}}
[[Category:Non-proportional multi-winner electoral systems]]
[[Category:Single winner electoral systems]]
[[Category:Preferential electoral systems]]
[[Category:Monotonic electoral systems]]
[[Category:Graded Bucklin methods]]
[[Category:Single -winner electoralvoting systemsmethods]]
[[Category:Cardinal voting methods]]
[[Category:NonNo-proportional multifavorite-winnerbetrayal electoral systems]]
1,196

edits