Talk:Tactical voting

From electowiki

This is the discussion page (the "Talk:" page) for the page named "Tactical voting". Please use this page to discuss the topic described in the corresponding page in the main namespace (i.e. the "Tactical voting" page here on electowiki), or visit Help:Talk to learn more about talk pages.


Beware the "What if?" Scenario

One avid EM analyst (who may choose to remain anonymous) has fired the following scenario at several method proposals to apparently illuminate vulnerability to tactical voting. It used to trouble me until I thought about what it would mean in the real world. Once I did that, I realized that it is a preposterous red herring that is not worth worrying about. It is now important to me to explain why so that others encountering this scenario (or its ilk) are not tripped up by it.

The scenario:

Sincere votes
45 B|A>C
49 A|B>C
6 C|
"A" is the Condorcet winner
Tactical votes
45 B|C>A (insincerely ranking A beneath C)
49 A|B>C
6 C|
Cycle A > B > C > A is created. Some methods might now choose "B" or "C".

Let's translate this into real world factions so we can see what is really being proposed here:

Sincere beliefs:
45% Rep | Dem > C
49% Dem | Rep > C
6% C

Thought 1: What kind of candidate is 'C' to be sincerely ranked bottom by both Republican and Democrat voters simultaneously? "C" is not a Libertarian or Green. Considering how disdainfully the Reps and Dems (voters) view each other, 'C' must be someone truly odious and unamerican for Rs and Ds to both prefer each other to him. Therefore, we can deduce that 'C' is something like an unapologetic Communist or Nazi.

Proposed Republican Strategy, Convince the Rep 45% to rank 'C' higher
45% Rep | C > Dem
49% Dem | Rep > C
6% C

Thought 2: What would voters do if the Republican (or anyone in that campaign) publicly asked voters to up-rank a Communist or Nazi? What political hay would the Democrat make to fuel such a backlash? With such political hay available, would the Democrat have any reason to counter the strategy with a similar ploy that might see 'C' actually elected?

Thought 3: How much effort does it take to move a whole 45% of a general electorate in lock-step when many are unenthusiastic swing voters? How much of that 45% would be lost simply because valuable resources had been diverted from straightforward campaigning? Wouldn't it be easier to win some swing votes from the other camp or to win the unused 2nd place votes from the 'C' voters?

Moral of the story: If you're not an ivory tower theoretician working on an academic proof, then when confronted with a "What if?" scenario, always pause a moment to consider what it would really mean in the real world. You may discover that it's not worth designing around. Jrfisher 12:47, 20 Aug 2005 (PDT)

I offered the above scenario to suggest strategic differences between Definite Majority Choice and Condorcet//Approval. Unfortunately, User:Jrfisher has interpreted my presentation of this scenario as an attack on one or the other method, and believes that I believe the above scenario is realistic. He has since requested that I stop confusing people, and I have agreed. Kevin Venzke 21:08, 20 Aug 2005 (PDT)


C4ES Tactical voting page

The Center for Election Science recently posted an article/whitepaper/whatever about tactical voting on their LinkedIn feed titled "Tactical Voting Basics".[1] I have had a chance to read the whitepaper yet, but I'm curious: what is it that C4ES gets right that the Tactical voting article on electowiki is silent about? Is the C4ES article clearer because it's shorter and doesn't get lost in the weeds? Is there anything we need to do in order to make the Tactical voting article as useful as C4ES's article? -- RobLa (talk) 03:49, 2 September 2020 (UTC)


Finding clearer names for strategies

"Compromise" has a nice, positive connotation that makes it easy for FairVote et al. to downplay it as "not that bad." It even makes it sound like it promotes "moderation" or "compromise" (when it does the opposite)! I prefer "decapitation" (cutting the "head"/top candidate off your list) or "lesser evil"—I've seen both of these in the literature.

Burial has a sufficiently-evocative name, I think (sounds like burying a dead body); which probably contributes somewhat to people's skittishness around Condorcet methods.

Pushover needs a better name (preferably one that highlights just how insane it is that this strategy works in IRV). I kind of want to call it "RNG hacking"...

Compression has an easy-to-understand name; it sounds less bad, but that's because it *is* less bad.

Pied-piper is more familiar to Americans by the name of "raiding." --Closed Limelike Curves (talk) 02:53, 22 March 2024 (UTC)