Erroneous Treatment of Un-indicated Candidates

The purpose of this post is to argue against the variant of Score Voting (Range Voting) proposed by Donald Arthur Kronos, Ph.D..

The reason I argue is to provide you with the benefit of my knowledge and understanding.  The tool for doing this is reason.  I wouldn’t argue if I didn’t think you were in error.  I also wouldn’t argue if I thought that I had no chance to convince you and/or other readers of how to correct the error.  The reason this matters is that I care about democracy, and I believe that the greater the degree to which people become properly informed about those aspects that are essential and central to the possibility of democracy, the closer we all come to that happy day when it will be real in the US.

Since you communicate your ideas almost exclusively through tweets, there is a job of compilation required before our possible readers and I can understand for sure what you are proposing.  So, let’s get started.

Excerpt #1

@DonaldKronos Correct me if I’m wrong, but I believe you propose a scale from -x to x where x is positive. So zero on your scale is the mid.

@william_waugh Yes, that much you have correct. It is the REST you have wrong!

From this pair of tweets (one from me and your reply), we can establish for sure that your proposal is for a Score election where the range goes from -x to +x where x is some positive number.  The midpoint on such a scale is zero.

Excerpt #2

@DonaldKronos Let’s say 100 million vote O. +5, R. -5, and as many vote R. +5, O. -5, and 3 people vote Waugh +5. Who wins?

@william_waugh The tie breaker, obviously… since at least on the average that person was “liked” rather than undecidable.

From this two-tweet excerpt, we can see that a candidate who receives three votes out of 200 million three can win the election.  Here’s the math: O. receives a net score of zero on Dr. Kronos’s zero-centered scale, because (100 million) * 5 + (100 million) * (-5) + 3 * zero = zero.  Similarly, R. receives a score of zero, the middle of the scale.  But candidate Waugh receives a score of (200 million) * zero + 3 * 5 = 15.  So Waugh receives the highest score and so wins the election.

Dr. Kronos goes on:

@DonaldKronos @william_waugh But such an unlikely scenario deserves much less consideration than the stupid REALITY we see every election.

Today’s Argument

I provide a two-pronged argument against the latest statement above. In the first prong, I plan to convince you (all readers) that scenarios of the sort of which I gave a (perhaps ridiculous and extreme) example do deserve consideration.  In the second prong of my argument, I will point out that one or more voting systems that are alternatives to the one Dr. Kronos advocates for would reverse “the stupid REALITY we see every election”.  If you are convinced by the second prong of my argument, you will be able to see that the dichotomy Dr. Kronos draws between “such an unlikely scenario” and “the stupid REALITY we see every election” is a false dichotomy.

First Prong

Here’s the horror story that runs through my mind when I think about voting systems that put candidates not named on a ballot in the middle of the scale instead of at the bottom.  Any number of subsets of the population are communicating in their cliques and echo chambers in a way that they can reach a fair number of like-minded people without drawing a lot of attention from the rest of the population, those not affiliated with that particular clique.  For example, consider the preachers in some of the megachurches.  Some of these people can reach hundreds, maybe thousands of listeners or readers without drawing my attention or yours.  I suggest that it is plausible that in a close election, a few thousand votes for some preacher could make the difference and elect that preacher President, under the kind of voting system Dr. Kronos proposes.  We’re talking about a candidate whose positions are unknown to the vast majority of the electorate.  Perhaps that candidate wants to establish a theocracy, or kill everyone caught having certain kinds of sex. That candidate could be in favor of anything at all; almost no one would know, just a relative few followers.  Under the proposed system, a voter would not only have to weigh in on the candidates already listed on the blank ballot.  The voter would have the need and duty to address every write-in that a few thousand other voters might bring into the election.  I would consider the system repaired if the voter were allowed to provide a default score, and say that their ballot should give that score to everyone not specifically named on the ballot.  Would you accept, as a compromise, to offer voters that option, Dr. Kronos?

I think it may help readers think about the true meaning of the system Dr. Kronos proposes if we give a bit of discussion to the notion of applying a linear transform to the scores used in the proposed system. Applying a liner transform does not affect the result. Whoever receives the highest score still receives the highest score after the transform is applied.  Let us say “x” stands for the highest score in a given Kronosian voting system.  Then the lowest score is -x. Let “s” stand for the score that a voter gives to a candidate.  Let the transformed score s’ be (s + x) / (2x). After transformation, the range is zero to one, and the center point is 1/2.

When we view the proposal under the transformation, we can see that for write-in candidates that a voter doesn’t know about and doesn’t mention, the system awards that candidate a score of 1/2.  That is the system’s error, and that is how it can quite plausibly elect an extremist unknown to almost 100% of the electorate.  It registers a tremendous level of support from voters who don’t know about the extremist, amounting to a major proportion of those voters’ political power in the election.

Note that any number of these extremists could be simultaneously sneaking around the periphery of the election. Voters might learn about two or three and make a point of voting against them, but some others could slip under the radar.  Some could be put up specifically as distractions.  Extremists can try many times, and it takes just one candidate to go unnoticed, plus a close election among the well-known candidates, for the extremists to win.

Prong the Second

I submit that Approval Voting will suffice to reverse “the stupid REALITY we see every election.” And please understand that I share Dr. Kronos’s ardor for reversing said reality. I consider this reality a key, or linchpin, aspect of the means by which the current dictatorship maintains control, keeps the rest of us out of having any political influence, and promotes a false narrative that says the US is a democracy.  For more information on Approval Voting, including the should-be-famous New York poll that indicates what a dramatic effect Approval Voting can have on an election outcome, please see the list of links I publish at http://wp.me/p23U97-bd .

Thanks for reading.

__
William Waugh

Appendix — Citing This Post

Erroneous treatment of candidates not specifically indicated on ballot http://wp.me/p23U97-bn

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

8 Responses to Erroneous Treatment of Un-indicated Candidates

  1. Dale says:

    If you don’t want to use min.score, there’s the soft quorum.

    http://www.rangevoting.org/BetterQuorum.html

    • What bugs me about the “Better Quorum” is its use of a number without any notion of where the number would come from. In particular with this algorithm, the arbitrary number is the count of artificial votes to include. I think simply adding up the scores to determine the winner is fine.

  2. Donald Arthur Kronos says:

    Yes, I agree that it would be good to make an adjustment for people not on the ballot, if the current means of getting on the ballot are maintained, although I don’t know how much it would be fair for that adjustment to be. However, I would still say that the system we have now puts a known BAD CHOICE ahead of an unknown, which would not be corrected by anything I have seen you supporting. It is good though that you actually seem to care. Ideally, we should have ongoing open elections that never actually stop, with people able to adjust their vote any time, to allow for tthings like adding in votes against a previously unknown candidate or changing one’s mind based on newly acquired information, rather than having the booths open and close arbitrarily like they do now.

    Also, I would like to point out that a system which uses an average of votes rather than a cumulative total also runs the risk you mention, as it would be possible for a well known candidate who was not UNWANTED but simply not very WANTED to be beat out by a small number of high rating votes for someone who is practically unknown. I know some of the proposed systems out there work around this by requiring a minimum number of votes or by starting with a score intended to offset that possibility, but such measures are also arbitrary, and to the extent possible I think it would be best to eliminate the need for guesswork.

    The MAIN THING that I am trying to combat is the ability for slander campaigns to win elections, and the main reason for this is simple. If someone with an outrageous amount of money to throw into buying an election picks a PLAN A candidate and a PLAN B candidate, I would like to eliminate their ability to SCARE PEOPLE INTO VOTING FOR THEIR PLAN A by convincing them that if they don’t, their PLAN B will win. This is how it’s currently done, because from their perspective the worst their slander ads can backfire is if people see through the deception and vote for their PLAN B to keep out their PLAN A… which is generally fine with them, but still hurts the rest of is. Note that this happens in the primary elections just as it does in general elections, and an approval voting system is LESS VULNERABLE to it than a plurality system or runoff system, but is still vulnerable. If you can see what I’m talking about… perhaps you can find a solution that you can stand behind which would eliminate the problem as my system would, but without the need to arbitrarily adjust for the most likely valid assumption that known candidates are on the average better than the worst case unknown candidates. As it stands right now, worst case KNOWN CANDIDATES are about all we ever get, so it shouldn’t be hard to improve on that.

  3. Well, one thing that satisfies me a lot about the state of the dialog between us is that I feel I have finally made myself understood about the substance of my objection to the system (or class of systems) you propose. So, thanks for having communicated back in such a way that I can see that you have understood.

    Also, I appreciate your leadership in shifting the tone of our dialog so it is civil now. We were sort of going at each other on Twitter there for a while.

    “Yes, I agree that it would be good to make an adjustment for people not on the ballot, if the current means of getting on the ballot are maintained, although I don’t know how much it would be fair for that adjustment to be.”

    I’m not sure that anything I advocated would amount to a global “adjustment” of the people not on “the ballot” or even those not on a given ballot after the voter may have written in some names. Unless you think of it as an “adjustment” to give the lowest possible score to any candidates whose names are neither printed on the ballot nor written in on that particular ballot. So, that’s one of my proposals, to give the lowest score. But the other proposal I asked your thought on, is what if the voter is allowed to specify the score that such candidates should receive when that voter’s ballot will be tallied? The printing on the ballot for that option could say “All others”, or “Any others”. There would be a blank for filling in a score, or a series of circles to blacken in, if it’s a coarse-grained Score election. Anyway, the options available to the voter for marking “All Others” would be the same as for marking any of the names printed or written in. What do you think of that?

    Going back to the qualification you included, “…if the current means of getting on the ballot are maintained …”:

    Of course I advocate changing the means of getting on the printed ballot. However, I think write-ins should still be allowed. So, even if there are fairer means of getting on the printed ballot, if write-ins are still allowed, the question would continue to matter, of what happens if voter V1 writes in candidate C1 and voter V2 does not write in C1, then what score does the system assign to C1 on behalf of V2.

    “However, I would still say that the system we have now puts a known BAD CHOICE ahead of an unknown,”…

    Which system do you mean by “the system we have now”?

    …”which would not be corrected by anything I have seen you supporting.”

    I don’t see it as a problem that needs correcting. Suppose I’m Voter V2 and I don’t know about Candidate C1. Candidate C1 COULD be the devil himself in regard to my values as a voter. Even as bad as whichever Repugnicrat I hate the most. So, it seems to me that a candidate that I don’t know about should receive maximum rejection from me, because that may be what he deserves, and I have no way to know, and as a voter I should make the most pessimistic guesses as regards the intentions, behaviors, and values of other human beings. I see that you don’t agree with that. Given your disagreement, what do you think about the idea of letting every voter decide? Then when you, for example, cast your ballot, you can specify that you are going to give the unknowns the benefit of the doubt, and assign them the score that is the midpoint of the range.

    …”we should have ongoing open elections that never actually stop, with people able to adjust their vote any time, “…

    Yes, I like that. But for stability, it should take some kind of supermajority to replace the president/governor/Senator in the middle of her term.

    “Also, I would like to point out that a system which uses an average of votes rather than a cumulative total also runs the risk you mention, as it would be possible for a well known candidate who was not UNWANTED but simply not very WANTED to be beat out by a small number of high rating votes for someone who is practically unknown. I know some of the proposed systems out there work around this by requiring a minimum number of votes or by starting with a score intended to offset that possibility, but such measures are also arbitrary, and to the extent possible I think it would be best to eliminate the need for guesswork.”

    I share your askance-looking at systems that have arbitrary-seeming numbers and thresholds in them. That’s why, for example, I haven’t been promoting the so-called “better quota” for Reweighted Range Voting (RRV); I don’t include the “better quota” in the computer program I publish for tallying RRV, because “better quota” would use an arbitrary constant number for which I have seen no justification or reasoning (that being a count of fake votes to include in the tally).

    The coercive tactic you describe in the last paragraph of your comment is what I call “Good Cop, Bad Cop”. It is also related to the “prisoner’s dilemma” in Game Theory. And as you allude to, Approval (or more broadly Score) lessens the expected rewards of that tactic. I expect that with Score/Approval and a few other reforms, we shall see many more candidates on the ballot, and certainly a higher count of VIABLE candidates, than we have had up to now in the major elections. Voters will see that they don’t have to vote for Good Cop to keep Bad Cop from torturing them; they can elect a candidate who isn’t in the cops at all. The motivation for negative ads should plummet. When the voters by and large believe there are only two viable candidates, all the negative ads have to do is target the one “opposite” candidate to the one they are trying to get elected. But now there will be 10 or 15 opponents. The negative ads would have to say, for example, Jill Stein this, Gary Johnson that, Peta Lindsay the other, Virgil Goode such-and-such. Too many targets. It starts to make more sense to talk UP the candidate that the ad is FOR rather than concentrating on the opponents.

    While composing this reply, I’ve had a thought that makes me think maybe I don’t have really so very much reason after all to oppose your proposal that unknowns get the middle score. Where this comes from is remembering when I was trying out an example of using Score Voting and thinking about the 2012 candidates for US President and testing what strategies would work best for a voter with my values. What I found was that given how little support my favorite candidates had from other voters, my best strategy with regard to my Compromise candidate (the lesser evil) was to give him 99% of the maximum score (assuming the minimum is zero; I trust you can see how to map that figure to where its equivalent would fall on a range such as you advocate where the minimum is negative). So, given my strategy and the conditions that indicated it, giving a candidate 50% or the middle of the range would probably have had almost the same effect as giving them the bottom of the range.

    At http://articlev.org/proposals/index.php/116/score-voting I describe (and advocate) a coarse set of scores that gives finer options near the top of the scale than near the bottom. I don’t know how to PROVE this mathematically, but my feeling, after fooling around with numbers some, and also seeing a similar proposal from someone on the discussion forum of the Center for Election Science, is that such a deliberately skewed distribution of coarseness and fineness would fit well with the strategies that third-party and independent voters would usually have reason to employ.

    Anyway, thanks for the discussion, and I don’t expect I am going to harp on your system anymore, objecting to it, anyway. I have made my reservation about it clear, and even maybe am backing off from it some, as being all that important. However, if you have any further question or ideas to run by me, I would be glad to respond to such.

    William Waugh

    • In further support of what I said here about money in politics, here’s a better argument by Andrew Jennings, Clay Shentrup, and Warren D. Smith. They point out:

      “Money matters far too much in today’s political process. And efforts to curb that with typical campaign finance reform are inherently unstable, as cheaters will be more likely to win elections, and then just make their cheating retroactively legal, and/or intimidate government officials who dare to try to prosecute them. We believe it may be more effective to try to reduce the inherent importance of cash, than to wage a potentially futile battle to level the playing field. With score and approval voting, a candidate need not prove his electability in order to earn your vote.”

      http://asitoughttobe.com/2010/07/18/score-voting/

  4. Donald Arthur Kronos says:

    Sorry it’s been so long without me getting around to following up on this. Yes, I tried to tell you I have understood you quite well all along, and I appreciate you acknowledging that you can see now that I have understood. Thanks for that. I still think that if you do all the math based only on collected supportive votes it short changes the people who had anti-supportive votes they would have liked to cast, and adding the ability to represent a range from NON-supportive to positively supportive does little to help matters. Yes, it would be very sad if an election was held in which all candidates receiving large numbers of votes got exactly as many votes against them as for them and the win went to some candidate with a small number of positive votes and even less negative votes, but that does not mean that the winner was any worse of a choice than those candidates which people expressed EXACTLY MATCHING LEVELS OF SUPPORT FOR AND OPPOSITION TO, which of course becomes less likely as their total number of votes either way increases. The odds of such a scenario actually taking place are indeed quite negligible, but the odds are actually quite high in any given election without an #antivote option of enough votes meant to express opposition going TOWARD AN UNSUPPORTED CANDIDATE INSTEAD to effectively nullify the chances of any better choice winning the election. This is why so much money goes into slander ads rather than into positive campaign ads. I personally have no problem with someone resourceful enough to fund a campaign with huge amounts of money using such resources to advertise how much good they believe or would like the rest of us to believe their candidate can do. Such ads would be viewed with reasonable skepticism for the most part, which is exactly what tends NOT to be applied to ads which trigger hate and fear causing the fight-or-flight response to kick in.

  5. You say that Score where unindicated candidates receive the lowest score “shortchanges the people who had anti-supportive votes they would have liked to cast”. But anti-supportive votes are exactly scores that are below the midpoint, are they not?

  6. Dr. Kronos:

    Our dialog above took a pause from 2014 up to now. However now again I am receiving Tweets from you. You talk about “antivotes” and “honest voting”.

    I’ll start by addressing “honest voting” as a concept.

    In my opinion, the concept of “honest voting” is misleading. To judge a communication as honest or dishonest, we must find meaning in it as descriptive of the world or math or something, a description that we could in theory find to be false by running experiments etc. A vote is not a description, and it is not an expression of opinion. A vote is a command. Another example of command semantics is that of the signal carried by the cable connecting the accelerator pedal of a car to the throttle. When a voter casts her vote, she is commanding the tallying process to take her vote into account. The voter is selecting from within her freedom of movement in regard to the content of the ballot, in order to affect the tally according to the rules of that voting system in use for that election. The voter may judge what vote to cast based on her estimate of what gives the best expected value for affecting the outcome of the election, relative to the valuations she feels for the possible outcomes, and given the voter’s degree of partial or full ignorance concerning the behavior of other voters. This is similar to a car driver’s decision to adjust the throttle. The driver has an understanding of what doing that does to the motion of the car, has some goal to change or maintain that motion, and judges how to work the controls to bring that about. This is neither honest nor dishonest. Of course, a car usually has only one driver working it, and an election by contrast has many, and the drivers of the election are not usually in agreement about which way it should go; they are in fact competing about that. However, I stand by the analogy in the sense that we are talking about commands here and not opinions, and there is no dishonesty.

    Now to “antivote”, and your mention above that suggests that the systems I advocate fail to respect the voters’ rights to oppose candidacies. (“I still think that if you do all the math based only on collected supportive votes it short changes the people who had anti-supportive votes they would have liked to cast, and adding the ability to represent a range from NON-supportive to positively supportive does little to help matters.”) Score Voting does not shortchange the people who want to oppose a candidacy. It gives them equal power to the same count of people who support that candidacy.

    In a Score election, a vote consists of a “grading” for each candidate coming from the voter who casts that vote. I introduce the term “grading” in order to attach some term to that thing so I can talk about it. Some Some people call it a “vote”, but to be precise, it’s necessary to distinguish a grading from the whole ballot.

    Can we characterize a grading as supporting or opposing a candidacy? Clearly we can if it is at the top of the range or at the bottom of the range. Otherwise, we cannot know whether it supports, opposes, or has neutral effect, without knowing something about the other votes in the election. Therefore, any vote at the bottom of the range is an “antivote” in your term, and no other necessarily is.

    Since you are advocating some kind of Score Voting, which of course in general I strongly agree with, I just want to give a note here about what I think is the most important consideration when choosing a voting system (this is not a critique of your position and I am not saying you would violate this consideration), and then I will go on to describe my current favorite system to suggest that States should enact (for single-winner elections, if we don’t find a way to eliminate those). (I am not smart enough to argue about multiwinner elections).

    Agreeing with the Frohnmayers, father and son, I judge, at least tentatively, that the key required characteristic of a voting system that would permit the people to establish, for single-winner elections, should such continue, characteristics consistent with a democratic republic, is equality of influence on the outcome. The Frohnmayers point out that the traditional test for equality of weight is balance. They propose a balance test to which any voting system can be subjected. A system passes the test iff for all possible sets of votes A and all possible votes _c_, there exists a vote _d_ such that the electoral outcome with votes A bag-union {c} bag-union {d} is equivalent to the outcome with only votes A. What this says is that if a voter based on his values casts _c_, another voter whose values are exactly opposed can cast _d_ to cancel the effect of the first voter’s vote. So the logic here is that if the voting system allows one of these voters to sway the outcome, and does not allow the other voter to sway it back to exactly where it was in the first place, then the system is allowing one of those voters to have more power than the other, over the outcome. So that would be undemocratic by definition, I would say.

    When I refer to “equivalent” outcomes, what I mean is that they would elect the same person, and would have the same meaning to the public regarding how well the losing candidates did relative to one another and relative to the winner. Shifting all the numbers up or down by a constant does not change this meaning.

    Currently here’s what I’m promoting for a single-winner election.

    1. For every candidate in the race, a voter has a right to give that candidate a “grade” of A, B, F, or G.

    So for example, my ballot might look like this:
    Nader – A
    Gore – B
    Bush – G

    The ballots should of course be made of paper.

    2. Each entry on a ballot, where a voter assigns a grade to a candidate, shall be called a “grading”.

    3. The tallying process accumulates an _overall score_ for each candidate. At the start of the tally, the candidate’s accumulated score is zero. The tallying process examines the gradings that the candidate received from the voters. For every grade of “A” found among those, the process adds 50 points to that candidate’s score. For every grade of “B”, it adds 49 points. For every grade of “F”, it deducts 49 points. For every grade of “G”, it deducts 50 points. When all the gradings for that candidate, found on the ballots, have been taken into account, the accumulated score for that candidate is that candidate’s overall score.

    The tallying process should of course be carried out by human beings, with witnesses to check to assure accuracy.

    4. The candidate with the highest overall score wins the election.

    Under this system, my working hypothesis about the best voting strategy for those who favor Nader and strongly oppose Bush, is to vote as in my example above, i. e., Nader A, Gore B, and Bush G. Someone may come along with mathematical reasoning, or made-up example elections, and cause me to reject that hypothesis. Or I could come up with such reasoning or examples myself. But for now, it is my best notion of how they should best vote for their own interests in this scenario.

    If you took the above system and eliminated the possibilities of the voters awarding B and F grades to the candidates, keeping only A and G as the possibilities, you’d end up with Approval Voting.
    Under Approval Voting, would an equally good strategy be available to those who favor Nader and strongly oppose Bush, as good as the strategy available to them under my ABFG grading system?
    This is a rhetorical question. I believe I know the answer. What do you think, readers?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s