"At a theoretical level, it is well known in game theory and social- choice theory that a larger menu of choices do not necessarily lead to better decisions, even when there is complete information. Indeed, additional choices often render a system more vulnerable to manipulation, especially by sophisticated players...
Even 10 levels is asking a lot of voters, who--psychological experiments show-- generally cannot make such fine discriminations. We do not say that 2 levels is always perfect--3 may be better in some instances..."
Yes, having too many choices can be bad. But more often, having too few choices is what is bad. Well, why is too-many bad? For three reasons: (1) you do not have enough time to consider them all and hence are prevented from making a decision at all, or (2) you are prevented from making a good decision, or (3) you are ignorant about some aspects of the choice and hence would prefer not to be forced to guess about those aspects.
Example: The Bush administration drug plan for seniors caused those seniors to have to consider over a hundred different "plans" they could select, written in legalese and advertising-ese.
However, none of these three reasons occur in range voting. Nobody is prevented from range-voting and range-voting well, because of their choices. If you increase the range from 0-9 to 0-99 on a 7-candidate ballot, people do not say "Holy cow, my number of possible choices is now 1014, I am so helpless, I cannot consider them all, I cannot even vote or hope to make a good vote." No. They simply instantly write down their numerical scores and are done with it. We've tried it many times.
Furthermore, range voters have the option of intentionally expressing ignorance or "no opinion" by leaving a candidate's score blank in which case that candidate's election result will not be affected by that voter. A lot of voters who are ignorant about a candidate prefer to do that. An apparently even larger number prefer to put "zero" as that candidate's score, thus "playing it safe" about unknown candidates. Another (apparently smaller) set of voters prefer to give unknowns an intermediate score like 35. All three choices are possible, and almost all voters decide how they want to play it in a matter of seconds, with nobody taking more than a minute.
This one surpised even me. It turns out that people actually vote faster with range voting (1 to 10 scale) than with approval voting (APPROVE versus NOT 2-way choice)!!
I got this from William Poundstone's book (Hill & Wang, Feb. 2008) Gaming the Vote. Poundstone interviewed Hot or Not co-founders James Hong & Jim Young. Hot or Not is a cult-favorite website in which you can rate members of the opposite (or same!) sex "hot" (10) or "not" (1) or anywhere in between – i.e. range voting. (You can also do other things, like submit your own photo – vanity central – or pay to join a matchmaking service.) So far, there have been over 12 billion votes cast on Hot or Not, which is far more votes than in all US presidential elections. Combined. All time.
Enough data for you?
They originally were worried than maybe scoring people on a 1-to-10 scale was going to take too long, which would repel customers. So they tested simpler schemes where you just had two buttons: "hot" or "not." They also tested another scheme where you saw two people at once, and pressed a button indicating who's hotter.
Guess what? It took the raters longer to make a less-informative decision! So Hot or Not abandoned both 2-choice vote schemes and went with the 10-point scale.
"Sometimes people can't even express the number," said Hong. "They just have a feeling and like having that bar" 'ah, it's kinda like here.'" They position the cursor where it feels right, and click.The rest was history – Hot or Not became one of the most successful web sites ever. If you try it, it should quickly become apparent to you that ranking the hotties in order, is a lot harder than (and will take you longer than) rating them. (You have to make sure no two have the same ranking, and make tough hotness-comparison judgements, and that takes time.) So again: it takes less time to rate people than to do other things which provide less information!
There have been complaints that Hot or Not is silly, superficial, and sexist. No one thinks it's hard to understand.
– (Quoted from Poundstone book, chapter 14.)
The "many choices paradox" range voting critics are actually correct that there is a "paradox"; it is just that it is a different paradox than they thought, and it yields exactly the opposite conclusion than they thought.
"Oh my god, I can't breathe! I'm going to die!"
"It is the vast number of air molecules in the room! There are over 1025 of them! I can't choose which!"
– "OK! Hang on! I'm going to transfer you to a vacuum chamber where there are only 10 air molecules!"
Return to main page