Definition of tactical voting

36 views
Skip to first unread message

Clay Shentrup

unread,
May 21, 2016, 3:44:06 PM5/21/16
to The Center for Election Science
Wikipedia says:

In voting systems, tactical voting (or strategic voting or sophisticated voting or insincere voting) occurs, in elections with more than two candidates, when a voter supports a candidate other than his or her sincere preference in order to prevent an undesirable outcome.

It cites a 1969 book called Theory of Voting.

I contend this is an antiquated and very wrong definition. Here's a simple example.

Suppose your utilities for three candidates are X=10, Y=1, Z=0.

Now you can choose between a guarantee of getting Y, or a 99/1 chance of getting X or Z respectively.

Obviously the better strategy that virtually any human being will take is the lottery, even though it increases one's risk of getting the most undesirable outcome in Z.

Warren D Smith

unread,
May 21, 2016, 4:59:42 PM5/21/16
to electio...@googlegroups.com
On 5/21/16, Clay Shentrup <cl...@electology.org> wrote:
> Wikipedia says:
>
> In voting systems, tactical voting (or strategic voting or sophisticated
> voting or insincere voting) occurs, in elections with more than two
> candidates, when a voter supports a candidate other than his or her sincere
>
> preference in order to prevent an undesirable outcome.
>
> It cites a 1969 book called *Theory of Voting*.
>
> I contend this is an antiquated and very wrong definition.

--it certainly was wrong. Insincere tactical voting is when your vote
is dishonest.
Period. If the system is based on rank-order ballots, that means your ballot's
rank-ordering is not identical to your honest ordering.

In general systems, which are based on an arbitrary bitstring as a ballot,
it is not necessarily obvious what "honesty" even means, and therefore
what dishonesty means either. Fortunately, designers of voting systems
usually try to make there be a fairly obvious intended
notion of the meaning of "honesty"
but it does not have to be that way.

Clay Shentrup

unread,
May 21, 2016, 5:36:21 PM5/21/16
to The Center for Election Science
On Saturday, May 21, 2016 at 1:59:42 PM UTC-7, Warren D. Smith (CRV cofounder, http://RangeVoting.org) wrote:
Insincere tactical voting is when your vote is dishonest. Period.

I think of a distinction between a vote that's _intended_ to be tactical vs. one that actually is. Perhaps that's wrong or simply not a helpful distinction.

William Waugh

unread,
May 22, 2016, 12:17:40 AM5/22/16
to The Center for Election Science
In regard to the conversation at https://groups.google.com/forum/?fromgroups#!topic/electionscience/nba2jgeimfk about "tactical voting", I don't find any intellectual value in using the term. I think that a criticism of some voting systems that is related to at least part of what people are trying to get at when they talk in terms of tactical voting, is that some voting systems fail to give good information about the relative popularities of the losing candidates. They do this by taking up the voter's bandwidth on just deciding the winner and leaving none for information about the losers. The vote-for-one system, for example, falls prey to that criticism, but Score{100, 99, 90, 50, 10, 1, 0} does not.

Steve Cobb

unread,
May 22, 2016, 8:26:00 AM5/22/16
to The Center for Election Science
"a voter supports a candidate other than his or her sincere preference in order to prevent an undesirable outcome."

This is only half the story--tactical voting includes voting down a candidate (e.g. his second choice), e.g. bullet voting. I describe it as voting a candidate insincerely 1) higher or 2) lower, with the expectation that the candidate will a) win or b) lose. Four categories.

I agree with Warren. Votes combine distributed knowledge and/or interest. Imagine a zero-interest decision, or at least one where you the voter had no interest, or at least where your interest was to determine the correct answer (e.g. jellybeans in a jar). You would vote sincerely. Tactical voting is your deviation from your zero-interest vote.

William Waugh

unread,
May 24, 2016, 10:27:12 PM5/24/16
to The Center for Election Science

This assumes that a vote has a particular semantics or meaning in the first place, like the estimate of a jelly-bean count. Can that be supported? I say a vote has no such denotation. The true semantics of a vote is similar to the semantics of an accelerator cable in a car, the cable that connects the accelerator pedal to the throttle. It conveys a command, not knowledge about the world.

Toby Pereira

unread,
May 25, 2016, 4:12:17 AM5/25/16
to The Center for Election Science
And commands can be tactical can't they? I think I understand your point - that a command can't be honest/dishonest or sincere/insincere, because it is just a command. However, whatever you want to call it, I think there is a useful distinction to be made between a vote that matches a ranking/rating etc. of your true preferences and one that doesn't.
Reply all
Reply to author
Forward
0 new messages