It’s with a somewhat heavy heart that I write this post. A number of times I have written extolling the virtues of IRV (Instant Runoff Voting), including directly after the recent 2009 mayoral election in Burlington, VT. In fact, my previous post went as far as to declare IRV in Burlington’s 2009 mayoral election to be a success while simultaneously pointing out that a paradox- a failure of the system to be consistent, free and fair- occurred. I want to be very, very clear as I move forward here: plurality voting (i.e., whoever gets the most votes wins) is a drastically flawed system that only works when there are only two choices on the ballot. This is because of something called Arrow’s Impossibility Theorem, which is simply a mathematical concept which states (recognizes) that whenever more than two choices appear on a ballot the standards for a consistent, free and fair vote (the Fairness Criteria– developed and agreed upon by political scientists and mathematicians) cannot be met. In the past I have extolled IRV as being the best alternative because it generally is capable of creating paradoxes or violating the Fairness Criterion a reasonably low percent of the time. In my opinion, there may still be some truth to this, and it is entirely true that IRV out-performs plurality voting. However, what seems to be even more true is that there are other methods for voting which greatly out perform IRV, and at the very least a modified form of IRV is necessary to over-come its most glaring flaws.
Unfortunately, the 2009 Burlington mayoral race demonstrates pretty much everything that can go wrong with IRV voting, and the simple fact that it did so should give us all pause and make us re-consider our acceptance of this voting method. Again, I want to be as clear as I can: IRV is drastically better than plurality voting (the method we’re accustomed to using when electing the President or members of Congress). That IRV in this instance suffered from a number of paradoxes and most certainly proved itself to not be consistent, free and fair should in no way whatsoever be construed as a reason to revert to plurality; likewise, the other options on the table for possible ways to hold an election all carry with them their own strengths and weaknesses. Again, Nobel winning mathematician Arrow proved for us decades ago that no system is going to be perfect, and even those few since him who claim to have found or proven an “impossibility” (a voting system that doesn’t violate the Fairness Criterion) have not done so to the degree that they have gained widespread acceptance for their findings. This is, in the end, a terribly complicated issue and there are no easy answers. There is, on the other hand, at least one important question on the table: was Bob Kiss preferred by the majority of voters in Burlington this past Town Meeting Day? A careful examination of the ballots suggests the answer is no, absolutely not. And because this is the answer the ballots themselves give us, it’s important we re-examine IRV because Kiss is the candidate it handed over as the “clear” winner.
As for the 2009 mayoral race in Burlington, I’d like to draw your attention to the facts as pointed out by UVM political scientist Anthony Gierzynski and Temple University mathematician Warren D. Smith (*and please note, whatever political agendas or personalization of this issue may exist between either of these two men and IRV or ardent IRV supporters like Terry Bouricius is not of my concern whatsoever- I am interested only in the facts as the actual ballots and a thoughtful, scientific analysis of them reveals):
-As has already been pointed out in my previous post, Montroll was the “Beats-All winner” (aka the “Condorcet winner”) as he would have beaten both Wright (56% to 44%) and Kiss (54% to 46%) in head-to-head races, demonstrating that he was the preferred candidate by the majority of voters. This is called a “Thwarted-Majorities Paradox”.
-Despite claims that IRV eliminates the “spoiler” effect, the truth is that Wright was a spoiler to Montroll. If Wright had not run, Montroll would have won (which, incidentally, would have been preferable for the majority of Wright supporters who listed Montroll as their second choice 1513 to 495 for Kiss as their second choice). Anyone who voted for Wright without fear that doing so would cause Kiss to win was simply wrong, because if those Wright supporters had strategically voted Montroll (like if Nadar voters had strategically voted Gore in Florida 2000) Kiss wouldn’t have won. This is called “Favorite-Betrayal” (i.e., when voting for your favorite harms either that candidate or another one you would have otherwise preferred).
-This election also produced a “No-Show Paradox”, meaning that if 753 Wright supporters who preferred Montroll over Kiss had simply stayed home on Town Meeting Day they would have gotten a more desirable (in their view) outcome- Montroll would have beaten Kiss. For these folks, the election results would have been more to their liking if they didn’t bother to vote!
-The software used to tabulate the votes eliminated Smith (1306 votes), Simpson (35 votes), and “write-in” (36 votes) in the first round of counting because it deemed them all to be mathematically (“inevitably”) losers of the election. Yet the software provider has made no clarification (that myself or the other listed authors of this analysis can ascertain) as to what criteria defines this: using the information and “logic” that has been provided, we could very well say that Montroll (and even Wright) would have “inevitably” lost and eliminated them in the first round as well.
-Most problematic of all (and for me, once I realized this my entire opinion of IRV began to rapidly shift): this election featured non-monotonicity. If 753 people who voted Wright as their first choice (specifically, all 495 who voted Wright>Kiss>Montroll plus 258 of the 1289 people who voted Wright only) had instead voted for Kiss then Wright would have been eliminated instead of Montroll and Montroll would have beaten Kiss in the final round of IRV (4067 to 3755). Put another way, Kiss won, but if 753 Wright voters had instead voted for Kiss, then Kiss would have lost and Montroll would have won (Monotonicity is one of the four Fairness Criteria and it states that “if x wins an election, and in a do-over or re-election the only changes made are in favor of x, then x should still win”). Simply put: if things change in your favor, but those changes negatively impact your final outcome, that is not consistent nor fair.
In the end, there are a whole slew of voting methods to consider. As I said, they each have their strengths and weaknesses. In the past I have been a strong proponent of IRV. Unfortunately, this election should give us all pause. Consider voting methods that would have chosen Montroll (simply using the preferences indicated by the ballots cast by the voters themselves): Nanson-Baldwin, Black, Raynaud, Schulze-beatpaths, Simpson-Kramer minmax, BTR-IRV, Tideman-ranked-pairs, WBS-IRV, Copeland, Heitzig-River, Arrow-Raynaud, Borda, Dodgson, Bucklin and (probably) Range and Approval. Voting methods that would have elected Wright: plurality. Voting methods that would elect Kiss: IRV.
Another way we can look at this all: all of the above voting methods are unanimous in one thing: Wright was the least favorite of the top three choices (i.e., he was the “worst choice” or the least popular among them). If this is true, then at the very least the election should come down to Kiss versus Montroll, and we know the voters preferred Montroll 54% to 46% over Kiss.
Again, none of this should be mis-construed: in a plurality voting system, Wright would have won the election despite being the “Lose-To-All loser” (i.e., he would lose head-to-head against either Montroll or Kiss), which would have been a far greater dis-service to consistency, fairness, and democracy. As well, I am not particularly extrapolating from any of this a recommendation about what system should replace IRV. Simply put, the jury is still out. We do know, however, that plurality is pretty much the worst possible method for voting we could use. As well, IRV doesn’t necessarily fail on this level every time- but the fact that it has, no matter what the statistical probability, is deeply concerning. We also know that, according to the consensus of the political science and mathematical communities no known system completely satisfies the most basic and simple criteria for consistent, free and fair elections (and though there are a number of individuals or small sects which claim otherwise, these claims have thus far not been accepted by the wider academic community and therefore should be considered with care and a thorough examination of the facts; the public has generally demonstrated to have a weak stomach for having its democracy rest on the back of an unknown system). Still, the claims of IRV have proven dubiously inconsistent at best,and possibly far worse than that. Because of this the public needs to carefully, thoughtfully engage in a conversation about how best to move forward. Range voting, for instance, may be a worthwhile step forward.
In an effort to pre-empt at least some of the critics who will come out against these conclusions: I absolutely favored Kiss to win and if I had lived in Burlington I would not only have voted for him, but probably would have volunteered for his campaign. Likewise, the voters have approved IRV as the method they want to use and it is absolutely true that IRV produced Kiss as the winner, and I do not think any of this analysis calls into question the fact that he was the election and is mayor (Wright’s re-count, even if carried all the way through, would have reproduced Kiss as winner every time using IRV, which is the method Burlington currently uses, by law). However, analysis of the ballots strongly suggests that Montroll was the most preferred candidate by the majority of the voters and the fact that IRV (or any other system) fails to elect him is troubling; again, not because I personally would prefer it (just the opposite, actually) but because the democratic will of the people would have. Sound logic does not make value judgments, except against that which proves inconsistent. The results from the 2009 mayoral race in Burlington were inconsistent (the candidate most preferred by the voters failed to win), unfair (voters intending one thing and preferring yet another got neither) and unfree (some voters would have gotten results more to their liking if they voted insincerely, or not at all).
In my own opinion, the only reasonable conclusion from any of this is that representative democracy itself proves, well, un-democratic. If we were to reject the very notion of electing people to govern over our lives and our communities and instead move to a direct, participatory model of self-governance we could very nearly do-away with the problems of having a system which is neither consistent, free, or fair. Just as with the challenges of replacing mere pluralism with a better system such as IRV, and just as with the challenges of replacing IRV with a system which proves itself to meet our standards for fairness, this would of course be difficult. However, democracy itself is difficult, a fact which fails to make it something un-worthy of our goals.
35 comments
Comments feed for this article
March 16, 2009 at 9:29 am
wdh3
Afterthought: the more I’ve looked into it, the more I am inclined to agree with proponents that Range Voting may actually solve Arrow’s Impossibility Theorem. See for yourself: rangevoting.org
If it could be implemented on a large scale without causing tremendous confusion and headaches (which it would seem it may be) than it would certainly be the answer.
March 16, 2009 at 2:04 pm
Clay Shentrup
Regarding the inevitable losers problem, Warren Smith got this response from the voting software company:
“If a collection of candidates with the lowest rankings have a total number of votes between them that is less that the next highest candidate, and if their elimination leaves enough candidates to fill the open seats, then their elimination is mathematically inevitable. That is, even if they are eliminated one-by-one, none of them can catch the next candidate.”
–Steve Willett, Election Solutions Inc.
Warren responds: “I think Willett is correct and his software was therefore ok on this.”
As for your claim that IRV is better than plurality, that is highly speculative. Statistically speaking, IRV does pick better winners (better as in “more satisfying to the average voter” = less Beyesian regret). However it also increases spoiled ballots by a factor of 7, and it causes election integrity nightmares. It is plausible that it is therefore *worse* than plurality, at least in certain circumstances.
Some would argue that it is the only politically viable alternative however. This may be true, sadly.
March 16, 2009 at 5:54 pm
wdh3
Well that does seem to clear-up the question about the software and how that first round worked, thanks.
March 16, 2009 at 10:57 pm
Rob Richie
Good luck engaging with all of this. My last post here, just quickly.
1. This election did not have a non-monotonic result. It could have had a nonmonotonic result, but it didn’t — that is, Kiss in fact did win. The only way you have a nonmonotonic result here is if Wright, Montroll or Smith would have won if being raised on ballots in some new combination.
2. As said before, but I’ll say it just one more time — system choices reflect different values. The Condorcet system that would have elected Montroll here puts a particular premium on not being a last choice and no particular value on being a first choice. That would change how candidates campaign, and I don’t think for the better. But others would weight that as the highest value, and those folks should support Condorcet.
3. For students of Burlington politics, I think there’s simply no way to know if approval voting or range voting would have elected Montroll. I suspect not, based on what I know of the city’s politics and how the candidates and their supporters would have behaved in that system.
Again, good luck there.
March 16, 2009 at 11:38 pm
Clay Shentrup
Rob,
This election was definitely non-monotonic. This is a simple and obvious fact, explained by Warren Smith (a Pinceton math Ph.D.) as follows:
“..this election also featured non-monotonicity. If 753 of the W-voters (specifically, all 495 of the W>K>M voters plus 258 of the 1289 W-only voters) had instead decided to vote for K, then W would have been eliminated (not M) and then M would have beaten K in the final IRV round by 4067 to 3755. In other words, Kiss won, but if 753 Wright-voters had switched their vote to Kiss, that would have made Kiss lose”.
— http://scorevoting.net/Burlington.html
In essence, Kiss won because he wasn’t liked enough. If voters had liked him more, he would have lost. That’s the definition of a non-monotonic election.
As for your second point, you’re contradicting yourself. If this had been a two-way IRV race between Montroll and Kiss, with the exact same voter preferences, you would have claimed that Montroll was the right winner (because IRV would have elected Montroll in that case).
March 17, 2009 at 1:21 am
Rob Richie
Okay, Clay, one more post… this was on my screen, and your post was there, and I read it, so… I’ll respond and then I’ll close the screen to avoid debating temptations….
1. Burlington COULD HAVE been non-monotonic. It WOULD HAVE been non-monotonic if Kiss had lost because he won additional first choices. That’s different than saying it was non-monotonic – in fact, it wasn’t.
2. It wasn’t a two-way race with Kiss and Montroll. And it wasn’t going to be such a race because Montroll didn’t speak to a lot of voters as a first choice, which is why nearly half of voters supported candidates as first choice other than Kiss and Montroll. Candidacies often flow out of the desires voters, and you can’t simply pretend them away.
3. You have no way of knowing whether range voting would have elected Montroll, so range quite easily would not have elevated a weak third-place finisher in first choices to victory.
Regardless, and this is a key point, you certainly will have to admit that range voting might not elect the Condorcet winner — in fact, you have to admit that range might not elect a candidate who won 51% of first choices. So given that you’re a range voting advocate, I wouldn’t bank too much on the Condorcet argument
March 17, 2009 at 3:08 am
Clay Shentrup
Rob,
Any political scientist or mathematician would acknowledge that this election was non-monotonic. You comment demonstrates confusion about what that means.
Like many election system criteria, monotonicity is about having consistency between two opposite “mirror image” scenarios. You assert that this election only WOULD HAVE been non-monotonic if an increase in support had caused Kiss to lose. But that’s the same as saying that it WAS non-monotonic because a DECREASE in support caused him to WIN.
Based on your reasoning, here’s what you would have said if Kiss HAD lost because he won additional support:
“Burlington COULD HAVE been non-monotonic. It WOULD HAVE been non-monotonic if Kiss had won because he lost first choices. That’s different than saying it was non-monotonic – in fact, it wasn’t.”
But that’s precisely what happened. The election was non-monotonic. And denying this obvious fact doesn’t help you. It just makes you seem unqualified and/or deceptive. Do yourself a favor and try to face reality when IRV’s flaws come to light. Then at least maybe educated people will empathize with your suggestion that IRV is at least better than what we have now.
As for your second point, you are acknowledging IRV’s failure of independence of irrelevant alternatives. If Montroll and Wright had done everything the same, and voters had had the exact same opinions of them, IRV would have made a different decision about which of them was the better candidate, based only on the presence or absence of the other candidates. Just like monotonicity, IIA is about consistency. When a voting method contradicts itself, that logically proves that it is occasionally making the wrong decision.
You’re right that we cannot say with certainty what would have happened with score voting or approval voting. But we can draw some reasonable conclusions with high statistical confidence. It is ironic for you to dismiss any conclusions that cannot be known with 100% certainty, considering you so often make blanket statements (like “No voters cast a strategic ballot in this election”) with absolutely no data whatsoever, as if you could read the voters’ minds.
The only way to really know voters’ exact thoughts is to use digital voters in Bayesian regret simulations. When you do that, score voting and approval voting certainly are superior to IRV. In fact, IRV comes in almost dead last.
http://scorevoting.net/UniqBest.html
March 17, 2009 at 4:24 am
Mister Guy
The idea that Montroll should have won this past election is just silly on the face of it. Sure, he was one of the main benefits of Smith’s votes once he was booted from the IRV process, but Kiss was the overwhelming winner of Montrol voters, which pushed him over the top in the end.
So, first people get upset because Wright lost, and now we’re supposed to be “concerned” that Montroll lost as well?? Please…
I personally don’t think that’s it possible to state simply that “Montroll was the ‘Beats-All winner’ (aka the ‘Condorcet winner’) as he would have beaten both Wright (56% to 44%) and Kiss (54% to 46%) in head-to-head races, demonstrating that he was the preferred candidate by the majority of voters” without twisting the intent of voters around. IRV is not about ranking one candidate against one other candidate only…it is a system of voting that really only should be used when there are more than two candidates in any given race.
“If Wright had not run, Montroll would have won (which, incidentally, would have been preferable for the majority of Wright supporters who listed Montroll as their second choice 1513 to 495 for Kiss as their second choice).”
But Wright DID run, and there’s really no way to accurately predict what Wright voters would have done had he not been in the race…stayed home, voted for another candidate, etc., etc….
“This election also produced a ‘No-Show Paradox’, meaning that if 753 Wright supporters who preferred Montroll over Kiss had simply stayed home on Town Meeting Day they would have gotten a more desirable (in their view) outcome- Montroll would have beaten Kiss.”
Again, would these people have bothered to show up had Wright not been in the race??
“The software used to tabulate the votes eliminated Smith (1306 votes), Simpson (35 votes), and ‘write-in’ (36 votes) in the first round of counting because it deemed them all to be mathematically (’inevitably’) losers of the election.”
Really…does anyone really think that these candidates could make that bad a showing in the first choice votes and THEN end up winning in the end?? Please…when has this kind of thing *ever* happened in IRV??
“If 753 people who voted Wright as their first choice (specifically, all 495 who voted Wright>Kiss>Montroll plus 258 of the 1289 people who voted Wright only) had instead voted for Kiss then Wright would have been eliminated instead of Montroll and Montroll would have beaten Kiss in the final round of IRV (4067 to 3755).”
But they didn’t vote this way, and I’m sure all of those voters had valid reasons why they voted the way that they did. Woulda, coulda, shoulda…
“Another way we can look at this all: all of the above voting methods are unanimous in one thing: Wright was the least favorite of the top three choices (i.e., he was the ‘worst choice’ or the least popular among them). If this is true, then at the very least the election should come down to Kiss versus Montroll”
And this is supposed to be a surprise to anyone?? This is Burlington, VT people…GOPers, in general, aren’t valued too highly in that city yanno…sheesh…
March 17, 2009 at 5:12 am
Clay Shentrup
Mister Guy said:
“IRV is not about ranking one candidate against one other candidate only…it is a system of voting that really only should be used when there are more than two candidates in any given race.”
That is exactly the point. IRV probably picked the wrong winner because it calculates the result in a sub-optimal self-contradictory way. If Kiss really was the “best candidate”, then he must still have been the best candidate if Wright had dropped out on the day of the election (and no voters’ preferences had changed as a result). But that’s not what would have happened. IRV would then have elected Montroll.
If you do Bayesian regret calculations, you find that IRV is one of the poorest-performing voting methods — meaning that it makes voters statistically less satisfied with election results:
http://scorevoting.net/UniqBest.html
March 17, 2009 at 5:27 am
Mister Guy
“IRV probably picked the wrong winner because it calculates the result in a sub-optimal self-contradictory way. If Kiss really was the ‘best candidate’, then he must still have been the best candidate if Wright had dropped out on the day of the election (and no voters’ preferences had changed as a result). But that’s not what would have happened. IRV would then have elected Montroll.”
Again, woulda, coulda, shoulda…that’s not a vaild argument for one system over another. The voters of Burlington voted the way they did for a set of reason that is completely & totally unknown to you from afar, period.
You’re an advocate for “Range Voting”…fine. But what passes for “analysis” of this past election in Burlington is rather laughable in quite a few spots. You make several assumptions about the suposed will of the voter where there is no evidence that making those kind of assuptions is in any way, shape, or form logical or reasonable.
It appears to me that you just have it in for IRV……
March 17, 2009 at 5:42 am
Clay Shentrup
Mister Guy:
You’re not understand the point. In a case of two contradictory “mirror image” election scenarios, one of them is the one that DID happen, and the other is the one that “coulda would shoulda” happened. Logically, at least one of them had the wrong result (and possibly both). There is therefore a 50% or better chance that the wrong candidate was elected in the election that DID happen.
Once you understand this, the analysis isn’t laughable at all. It demonstrates serious problems that IRV proponents commonly dismiss as statistically insignificant blips. If anything, this analysis is academically valuable as a tool for exposing the blatant deception of groups like FairVote.
There are also bigger concerns.
1) The poor performance of IRV, per extensive Bayesian regret calculations (where you actually CAN read voters’ minds and know exactly what they wanted).
2) The severe election integrity problems that occur as a result of IRV, since it cannot be counted in precincts, and may be conducive to the increased adoption of fraud-prone (electronic!) voting machines — AND causes the rate of spoiled ballots be about 7 times as high.
Do we “have it in” for IRV? Well, in many respects yes. But that’s as a RESULT of the actual scientific evidence. Whereas FairVote, for instance, has a pre-defined political agenda of implementing proportional representation. They see IRV as nothing but a useful stepping stone to that, thus they don’t particularly care whether it is good or not. Here you see Rob Richie, in this very message board, claiming that this election wasn’t non-monotonic. That’s just a blatant lie. That is not a total indictment of IRV, but it is telling of his credibility and/or expertise.
March 17, 2009 at 7:41 pm
buckles
While the mathematics and science is fascinating, whats really important is that conclusion! Direct participatory democracy in our communities and workplaces is the only solution to the repeated problems and disappointments of representative democracy. Thanks for this report.
March 18, 2009 at 3:03 am
Mister Guy
“There is therefore a 50% or better chance that the wrong candidate was elected in the election that DID happen.”
That’s sheer nonsense.
“The poor performance of IRV, per extensive Bayesian regret calculations (where you actually CAN read voters’ minds and know exactly what they wanted).”
LOL…sure you can…look, you might be able to fool some people that don’t have much knowledge of the scientific method or how mathematics actually works, but you’re not fooling me with this nonsense.
“The severe election integrity problems that occur as a result of IRV, since it cannot be counted in precincts, and may be conducive to the increased adoption of fraud-prone (electronic!) voting machines — AND causes the rate of spoiled ballots be about 7 times as high.”
The so-called “spolied ballot” issue has been dealt with very nicely here:
vermontdailybriefing.com/?p=1215
There was, indeed, a LOT of smoke blown both before & after the election about the “fear” of “integrity problems” with the electronic voting machines THAT LEAVE A PAPER BALLOT trail in Burlington. Almost all of those issues (if not all of them) were proven to be false once they were investigated further, and I would direct you to the discussion here for more on these smoke without any fire accusations:
http://www.burlingtonpol.com/2009/02/irv-and-election-integrity.html
http://www.burlingtonpol.com/2009/02/irv-and-election-integrity-ii.html
There are city workers that have pledged to conduct elections in Burlington fairly, there are volunteers at the polls that have sworn the same oath as well, and the ballots & voting chips can be (and I think are actually) escorted from the polling places by these officials with a police escort.
“Well, in many respects yes.”
Thanks for admitting the obvious…now, if you could, tell me where range voting is actually being used (or has been used) as a voting method in the modern world.
“Whereas FairVote, for instance, has a pre-defined political agenda of implementing proportional representation. They see IRV as nothing but a useful stepping stone to that, thus they don’t particularly care whether it is good or not.”
Look, there are a more than a few people that would like VT to go to a more parlimentary system of government. Those kind of changes would require MASSIVE changes to not only this state’s Constitution, but maybe even the U.S. Constitution. I’ve used IRV more times than anyone in Burlington has through the union that I used to be a member of in a former job. It works just fine tyvm.
March 18, 2009 at 10:22 am
Vermont News Guy» Blog Archive » The Way We Vote - Real News for Real Vermonters
[…] (Confused? For details, see Hamilton’s full article). […]
March 18, 2009 at 7:45 pm
Terry Bouricius
I don’t want to allow the false assertion that the Burlington IRV election suffered a “non-monotonicity failure” to settle into the realm of “known fact,” simply by repetition. In fact there was no such failure.
The originator of this erroneous claim appears to be range-voting advocate Warren Smith, who states incorrectly states that Burlington’s IRV election suffered from a monotonicity failure on his web site (which Wes and Tony Gierzynski simply seem to accept as true).
To understand the context, here is the situation…Republican Wright was ahead in first choices in the initial tally with 2951 votes, next came Progressive Kiss at 2,585, and the Democrat, Montroll, was in third with 2063. There were also two other candidates with 1,306 and 35 votes respectively (and some write-ins). In the final runoff tally, Republican Wright had 48.5% and Progressive Kiss won with 51.5% (with some voters sitting out the runoff by not ranking either of the finalists).
What his analysis actually shows is that non-monotonicity could have affected the election, in theory, but did not. …If just over 25% of the supporters of Republican Wright had abandoned their true first choice and instead voted for any other candidate (although to meet the non-monotonicity definition they would need to switch to Kiss), they could have kept their favorite candidate, Wright, from making it into the runoff, and allowed the Democrat to face off against Kiss in the final runoff, where the Democrat would beat Kiss. But this did not happen, and there is nobody who thinks it was a sensible strategy for any voters or candidates to advocate. Certainly it had no impact on how candidates campaigned nor ever would have.
Another way that non-monotonicity could have occurred would be if Wright or Montroll lost the election because they got too many first choices that might have gone to some other candidate instead. Since Montroll didn’t even make it into the final runoff there is no way this could apply to him (any fewer first choices would just confirm his elimination). That leaves Wright. Again, even if some of his supporters had voted for any other candidate first, Wright would still lose the runoff between either Kiss or Montroll. Smith would need 1,279 first choices that actually went to Wright, to get Smith into the final runoff. But at that point Wright would be in fourth place and have no chance of advancing to the final runoff. So it is mathematically impossible for a switch of first choices away from Wright to have made him a winner. Thus, despite the Smith analysis, there was, in fact, no non-monotonicity event in the Burlington election.
It is also worth noting that this same kind of non-monotonic strategy (Republicans conspiring to help elect the Democrat to block the Progressive) could be pursued under the old separate runoff system just as well. In fact, non-monotonicity is a much bigger risk under two-election runoffs because with a separate runoff, strategic manipulators can change their first choice on their ballot between rounds, which can’t be done with IRV.
In another point in the analysis, Smith and Gierzynski attack IRV for failing to elect the apparent compromise Condorcet-winner. This is disingenuous because Warren Smith himself dismisses the Condorcet-criterion, since his favored method also fails to meet this criterion — in fact, range voting can elect the Condorcet-loser, which IRV never can, and could quite possibly allow the defeat of a candidate who won an absolutely majority of 51% or more of voters’ first choices.
He observes that in the Burlington election, the candidate who came in third in the initial tally was a compromise choice who could have beat either Wright or Kiss in a head-to-head election. This is called a Condorcet winner, and is a mathematically valid calculation, though little regarded in American politics. The point is that this compromise candidate would come in third and lose badly under the plurality system most IRV opponents support, and would also be eliminated in a traditional two-round runoff election system. In fact, such compromise candidates have a better chance under IRV than any other voting system used by any government anywhere in the world. But some anti-IRV, pro-plurality activists illogically use IRV’s failure to elect the third-place candidate in this election as an excuse for attacking IRV and supporting plurality elections.
This criticism of IRV is legitimate when coming from Condorcet advocates, (which Warren Smith is not, since his favored system also fails the Condorcet criterion). It comes down to a matter of what values one feels are most important in an election process – -both in who should win and what kind of campaign you want to see run. I and many other experts feel the “mutual-majority” and “later-no-harm” criteria are far more important than the Condorcet criterion, for example.
Condorcet voting methods discount the relative importance of first choices, to the extent that a candidate who came in last place in terms of first choices, or even wins no first choices at all, but who is a broadly acceptable second choice, can win a Condorcet election. One concern is how that might affect candidate policy discussion, where the avoidance of alienating any voters becomes more important than the earning of any first choices. But this, at least is an area of legitimate disagreement over what values to reward in an election process, where reasonable and honest people can hope to resolve their differences through open discussion.
So the Smith analysis shows that IRV is better than plurality (and in fact better than two-election runoffs), yet plurality voting advocates twist the story to claim it shows how bad IRV is. The lead author of the report favors other un-tested voting methods that he naively thinks are better than any other system, and constantly attacks IRV in hopes of winning support for his favorite theoretical system. The net result is to help maintain the status quo.
March 18, 2009 at 9:31 pm
Clay Shentrup
Terrill Bouricius,
This election certainly exhibited a non-monotonicity failure. This is explained by Warren Smith as follows:
“If 753 of the W-voters (specifically, all 495 of the W>K>M voters plus 258 of the 1289 W-only voters) had instead decided to vote for K, then W would have been eliminated (not M) and then M would have beaten K in the final IRV round by 4067 to 3755. In other words, Kiss won, but if 753 Wright-voters had switched their vote to Kiss, that would have made Kiss lose!”
— http://scorevoting.net/Burlington.html
Monotonicity is about the consistency between two “mirror image” election scenarios. If a candidate does BETTER in the scenario where he was ranked LOWER (or, alternately phrased, does WORSE in the election where was ranked HIGHER), then that means those two scenarios are non-monotonic.
Of course only ONE of those two mirror images can actually happen at once, so you will always say that the one that HAPPENED wasn’t non-monotonic, instead it was the mirror image scenario (that didn’t happen) that was the non-monotonic one. That’s clever, in that it confuses people who don’t have the time to think about it for more than a few seconds.
Here’s what you would have said if the if the mirror image scenario (that you now claim WOULD HAVE been non-monotonic) would have actually happened. If 753 of the W-voters had instead decided to vote for K (causing M to win), Smith would have pointed out that the election was non-monotonic, because 753 voters could have caused Kiss to win by LOWERING their rankings for him (i.e. what actually DID happen). But you would argue: “No no no. THIS election in which Kiss lost is NOT non-monotonic. It WOULD HAVE been non-monotonic if 753 voters had lowered Kiss below Wright on their ballots, thus causing Kiss to win.”
Now there’s the more minor concern you bring up, about whether the non-monotonicity was strategically utilized. For instance, say that 753 of the W-voters actually preferred K>W>M, but either voted W>K>M or W because that got them their first choice instead of their third. I don’t think it’s particularly likely, but it’s possible. That you dismiss even the possibility strikes me as presumptuous and un-objective.
Finally, Smith’s point about IRV’s failure to elect the majority winner isn’t disingenuous. He’s simply pointing out that IRV doesn’t live up to the claims of its supporters (e.g. you) who repeatedly claim that it always elects a “majority winner”. He’s simply demonstrating that you’re being hypocritical. And what methods Warren Smith supports are besides the point. We are talking here about IRV’s performance in Burlington, so to even bring that up is a red herring.
March 19, 2009 at 2:17 am
Mister Guy
“Republicans conspiring to help elect the Democrat to block the Progressive”
I would note that this time around, all of the candidates that I’m aware of in this race shyed away from “endorsing” another candidate (by saying that you should vote for me first & some other guy second). The last time IRV was used in Burlington, the GOP candidate then actively made it aware that he did not want the Dem that was in the race for mayor to win…so he advocated for his voters to vote for him first & Bob Kiss second. This same Republican bragged after the election that his voters (and I’m paraphrasing here) “helped to not elect a Democrat as Mayor of Burlington”. For quite a while now in Burlignton, GOPers & Progressives have gotten along much more than Dems & either other major Party in the city.
“The lead author of the report favors other un-tested voting methods that he naively thinks are better than any other system, and constantly attacks IRV in hopes of winning support for his favorite theoretical system. The net result is to help maintain the status quo.”
I suspected that this was true, which is why I asked this question (that’s gone unanswered BTW): “tell me where range voting is actually being used (or has been used) as a voting method in the modern world.”
“Kiss won, but if 753 Wright-voters had switched their vote to Kiss, that would have made Kiss lose!”
Again, as stated above, this didn’t happen because these voters most likely really, really wanted Wright to win (which is why they ranked him first!). You can’t just assume that the objective of these mostly Republican voters was to merely unseat Kiss…as many of them would very likely have MORE of a problem with the City of Burlington being run by a Dem than by Kiss!
“That’s clever, in that it confuses people who don’t have the time to think about it for more than a few seconds.”
Baloney…this is EXACTLY the tactic that you are using…trying to fool people with a completely phony “analysis” of the result…that completely ignores the reality of how voters chose to express their will in the election, period.
“Here’s what you would have said if the if the mirror image scenario (that you now claim WOULD HAVE been non-monotonic) would have actually happened.”
Here’s yet another example of you wrongly assuming what motivates people that support IRV. Many IRV opponents incorrectly assume that IRV advocates are merely trying to ensure that a more progressive (that’s with a small p) candidate gets elected. The reality is that in many parts of the country, IRV would help elect the exact opposite type of candidate.
“And what methods Warren Smith supports are besides the point.”
Nonsense! You’ve even admitted that you have it in for IRV!
March 19, 2009 at 3:16 am
Clay Shentrup
Mister Guy,
You continue to have the same misunderstanding about monotonicity as Rob Richie and Terrill Bouricius have shown. I’ll try, once again, to explain it more clearly.
Smith/I said:
“Kiss won, but if 753 Wright-voters had switched their vote to Kiss, that would have made Kiss lose!”
You responded:
“Again, as stated above, this didn’t happen because these voters most likely really, really wanted Wright to win (which is why they ranked him first!)”
Yes, we know that it DID NOT HAPPEN. The fact that you are saying that shows that you do not understand the point. Non-monotonicity is about TWO contradictory election scenarios – like the two equal but opposite parallel realities. In one reality, Kiss loses. In the other reality, everything is the same except that some voters rank Kiss LOWER, causing him to paradoxically WIN.
In the parallel reality where Montroll wins, Warren Smith is there arguing that the election is non-monotonic, because if some voters who voted W>M had instead voted M>W, then Montroll would have lost (OUR reality). And your doppelganger is in that parallel reality saying, “Yes, this election WOULD have been non-monotonic if some of these voters had instead preferred Montroll over Wright, and then Montroll would have LOST — but that didn’t happen! Montroll won!”
In the parallel reality where Kiss wins, Warren Smith is here arguing that the election is non-monotonic, because if some voters who voted M>W had instead voted W>M, then Montroll would have WON (the OTHER reality that you say WOULD have been non-monotonic). And you are in this reality saying, “Yes, this election WOULD have been non-monotonic if some of these voters had instead preferred Wright over Montroll, and then Mintroll would have WON — but that didn’t happen! Montroll lost!”
So you are saying that the election in that parallel reality was non-monotonic, not this one. And your doppelganger in that parallel reality is saying that YOUR Burlington election was non-monotonic, not his. The trick is that no matter WHICH reality you are in, you claim that the OTHER reality was the non-monotonic one.
So this is not a “phony analysis”. This is what any mathematician would say is the obvious point of monotonicity.
March 19, 2009 at 6:17 pm
wdh3
The dusty old text book sitting here on my shelf, “Excursions in Modern Mathematics (Fifth Edition)” (Peter Tannenbaum, California State University-Fresno, Peason Education, Inc 2004) seems to side with me, Smith, et al:
“Monotonicity: If choice X is a winner of an election and, in a reelection, the only changes in the ballots are changes that only favor X, then X should remain a winner of the election.”
Wikipedia siting Douglas R. Woodall (from “Monotonicity and Single-Seat Election Rules”, Voting Matters, issue 6, 1996) likewise states:
“A candidate x should not be harmed [i.e., change from being a winner to a loser] if x is raised on some ballots”
Again, if a certain number of voters who didn’t vote Kiss as their first choice had instead put Kiss’ name as their number one choice, Kiss would have lost. It seems pretty clear that this scenario meets exactly the definition of non-Monotonicity.
March 20, 2009 at 5:33 am
Mister Guy
“Yes, we know that it DID NOT HAPPEN.”
But you’ll keep talking about your “parallel reality” outcome as if that outcome is in any way, shape, or form logical given the way the will of the voters has already been expressed or would likely be expressed. This excerise that you’ve been a part of is a textbook example of trying to use “complex-sounding” words to obscure the real issues that are at hand, period.
“In the parallel reality where Kiss wins”
…which is otherwise known as the real world as we know it…sheesh…
“So you are saying that the election in that parallel reality was non-monotonic”
When have I ever use that phrasiology?? Oh yea…it as never…because I’m not out to intentionally confuse anyone…
“So this is not a ‘phony analysis’.”
Yes, it certainly is, and (as has been pointed out before) it is an “analysis” that is meant to run-down a highly-successful method of voting that has been used the world over (IRV), period.
“Again, if a certain number of voters who didn’t vote Kiss as their first choice had instead put Kiss’ name as their number one choice, Kiss would have lost.”
Sheesh…and let’s complete & totally ignore the REALITY of WHY these voters didn’t do this or would likely NEVER do this…this has got to be one of the most pointless running conversations on the Internet today…no wonder I stopped reading this blog a long time ago…
March 20, 2009 at 9:23 am
Clay Shentrup
To Mister Guy:
I’ve tried my best to simplify and clarify the concept of monotonicity, so I’ll just just use one last simple analogy, and then summarize.
Say you have two note cards with contradictory statements written on them, along with some proven fact, like:
Card 1: In 1990, Smallville was the most populous city in Funland.
Card 2: In 2000, Springfield was the most populous city in Funland.
Proven fact: Between 1990 and 2000, Springfield’s population decreased, while Smallville’s population increased.
These two cards are contradictory – which is analogous to having two mirror image elections that are non-monotonic. Based on the “proven fact” (analogous to ballot rankings), it is obvious that at LEAST one of these two cards is wrong. So regardless of which one you pick, you know there’s a very good likelihood that it has a false statement. It COULD be the other card that has the false statement, but you do not know. You just know that when people walk up and choose a card, they often pick one that is false.
Likewise, when you have a non-monotonic election, you know that the wrong candidate won in AT LEAST one of the two “mirror image” election scenarios. It could have been the one that actually occurred, or it could have been the hypothetical one that COULD have occurred. Or BOTH. (In this case, the evidence pretty overwhelmingly says it happened in the election that actually occurred; but regardless, BOTH elections — the one that did happen, and the one that hypothetically COULD have — are non-monotonic, just like both note cards in the above analogy are contradictory.)
I said: “In the parallel reality where Kiss wins”
Mister Guy said: “…which is otherwise known as the real world as we know it…sheesh…”
Exactly. The real world as we know it experienced a non-monotonic election in Burlington. And it was probably THIS election that produced the wrong winner, not the other one (that could have happened but didn’t).
“When have I ever use that phrasiology?? Oh yea…it as never…because I’m not out to intentionally confuse anyone…”
I didn’t say you used that phrasing. I was just conveying the gist of what you said, using an analogy that I hoped would clarify, not confuse. I’m sorry if it was more confusing, but clearly just the “non-confusing” text book definition of monotonicity clear to you. I was trying another angle.
I said: “So this is not a ‘phony analysis’.”
Mister guy said: “Yes, it certainly is, and (as has been pointed out before) it is an “analysis” that is meant to run-down a highly-successful method of voting that has been used the world over (IRV), period.
Again, the analysis is just a statement of concrete fact. It can’t be phony, because we have the ballot data to prove every claim that was made. You can personally verify it with Burlington city officials if you’d like.
For you to say that this is an attempt to “run down” IRV is like saying that Al Gore is trying to run down the “global cooling” concept by presenting data that Earth is actually heating. You’re unnecessarily adding some emotional or ideological context to what is a mere statement of scientific fact. I don’t know what purpose that serves.
I said: “Again, if a certain number of voters who didn’t vote Kiss as their first choice had instead put Kiss’ name as their number one choice, Kiss would have lost.”
You said: “Sheesh…and let’s complete & totally ignore the REALITY of WHY these voters didn’t do this or would likely NEVER do this…”
We’re ignoring it because it doesn’t matter. What matters is that IRV likely picked the wrong winner. We don’t care WHY voters voted the way they did – we just care that they probably didn’t get what they most wanted. Likewise, I don’t care WHY you pick the note card you pick (maybe you like yellow better than pink). I just care that there’s a high statistical probability that your card has a false statement. In this case of the Burlington election, voters likely created a set of ballot rankings that caused IRV to make a false statement of who the best candidate was.
“this has got to be one of the most pointless running conversations on the Internet today…no wonder I stopped reading this blog a long time ago…”
I think having a government run by the people that the voters actually WANTED is extremely important, not pointless.
March 21, 2009 at 4:10 am
Mister Guy
“I’ve tried my best to simplify and clarify the concept of monotonicity”
I under the concept just fine…it’s a purely theoretical mathematical concept that some try & apply to hypothetical forms of voting. IMHO, it’s simply not relevant to the way the will of the voters is expressed under an IRV system.
I just love all the comments above about the so-called “wrong winner” or “best candidate”…as if one can pick & choose which candidate “should” have won…it’s really hilarious…lol… In any voting system, one can try and pick & choose who they *personally* thought should have won, but that decision is ultimately left up to the voters in that election to decide. Either you believe in representative democracy or you don’t. Sometimes the system isn’t pretty, but, in the long run, I firmly believe that the people are smart enough to choose wisely about who should be operating their govt. for them.
“the analysis is just a statement of concrete fact.”
It’s pretty clear that you are merely going to repeat this same baloney over & over again as if it’s true. I think it’s very clear to outside viewers who is blowing a lot of smoke to try & obscure the fact that they have it in for IRV & who’s not.
“For you to say that this is an attempt to ‘run down’ IRV is like saying that Al Gore is trying to run down the ‘global cooling’ concept by presenting data that Earth is actually heating. You’re unnecessarily adding some emotional or ideological context to what is a mere statement of scientific fact.”
LOL…don’t try & change the subject to something else (you really, really don’t want to get into a discussion about climate change with the likes of me), and please don’t try & backpeddle from the position that we already know is true (because you already admitted it far above)…you & your colleagues have it in for IRV, period end of story.
“We’re ignoring it because it doesn’t matter.”
Thanks for admitting this as well…it’s been very apparent from the very beginning that you don’t care about the will of the voters as was expressed very, very clearly by the way they actually voted. IRV gives everyone a huge amount of insight (when compared to “traditional” voting) into how voters really feel about a given set of candidates. The type of voting system that you advocate for isn’t used by anyone apparently.
“We don’t care WHY voters voted the way they did – we just care that they probably didn’t get what they most wanted”
…by making a large amount of untrue assumptions about what was likely factoring into why they voted the way that they did in the first place. We’ve been over this before. I live in the area where this vote actually occurred…you obviously don’t…
“I think having a government run by the people that the voters actually WANTED is extremely important”
…and the voters used the IRV process to pick Bob Kiss for Mayor of Burlington this time around, and the VAST majority of voters in Burlington are very pleased with that choice tyvm!
March 23, 2009 at 6:41 am
Clay Shentrup
Here’s a helpful simplified explanation of the non-monotonicity problem for Mister Guy, and others. Here are two simplified IRV election scenarios. You can mathematically prove that the wrong winner is elected in at least one of them.
=========
#voters their vote
8 B>A>C
5 C>B>A
4 A>C>B
(winner = C)
#voters their vote
4 B>A>C
6 C>B>A
5 A>C>B
2 B>C>A
(winner = C)
=========
So if your election looks like one of these, you know there’s effectively a 50% or greater chance that the wrong winner was elected.
In the Burlington election you have the same thing. The election that happened, and its “mirror image”. At least one of them elects the wrong winner. (The evidence actually shows that it was probably in the one that actually occurred, since Montroll was probably the candidate that best represented the will of the voters.)
March 23, 2009 at 3:45 am
Joyce McCloy
It astonishes me to hear people say that IRV is better than plurality, since most often, IRV contests produce a plurality winner:
Saturday, March 7, 2009 No Majority Winner in Instant Runoff Voting election in Burlington Vermont Mayoral Contest
December 7, 2008 2 out of 3 Pierce County RCV “winners” don’t have a true majority
Consistent Majority Failure in San Francisco’s Instant Runoff Voting Elections. A review of the results for San Francisco Ranked Choice Voting elections shows that IRV elects a plurality winner: These results are remarkably consistent. Out of 20 RCV elections that have been held since the referendum establishing it passed, when IRV was used, it elected a plurality winner
IRV failed to produce a majority winner in the Cary North Carolina experiment in October 2007.
March 23, 2009 at 11:36 pm
Mister Guy
“The election that happened, and its ‘mirror image'”
…which is otherwise known, in this case anways, as the election that didn’t & would never happen, period.
“since Montroll was probably the candidate that best represented the will of the voters.”
More utter nonsense.
“March 7, 2009 No Majority Winner in Instant Runoff Voting election in Burlington Vermont Mayoral Contest”
Wrong again…Bob Kiss won that election with over 51% of the vote.
March 23, 2009 at 11:47 pm
Clay Shentrup
Mister Guy said:
>> “The election that happened, and its ‘mirror image’”
>…which is otherwise known, in this case anways, as the election that didn’t & would never happen, period.
No one is claiming it would have. The point is that either THAT election or the one the one that happened (or BOTH) elected the wrong winner. And you can mathematically PROVE it.
Here’s a “brain teaser” puzzle that demonstates this.
http://scorevoting.net/PuzzClay2Irv.html
That puzzle also demonstrates that you could swap the candidates in this alternate election scenario, to make it more plausible to your sensibilities, and it would still be equally statistically likely to have elected the wrong candidate.
>> “since Montroll was probably the candidate that best represented the will of the voters.”
> More utter nonsense.
There’s good evidence that it is true, not nonsense. But either way, we have mathematical CERTAINTY that in either this election, or the mirror image (that “didn’t & never would happen, period”), IRV elected the wrong candidate.
>> “March 7, 2009 No Majority Winner in Instant Runoff Voting election in Burlington Vermont Mayoral Contest”
> Wrong again…Bob Kiss won that election with over 51% of the vote.
Nope. Montroll was preferred to Kiss by a huge 54% to 46% majority.
March 26, 2009 at 9:28 am
Mister Guy
“No one is claiming it would have”
…which is why it doesn’t matter what would have happened in that purely hypothetical election, period.
“The point is that either THAT election or the one the one that happened (or BOTH) elected the wrong winner. And you can mathematically PROVE it.”
No, you really can’t, but keep saying the same thing again & again…it’ll become true eventually…honest it will…not…ugh…
“Montroll was preferred to Kiss by a huge 54% to 46% majority.”
LOL…this is only “true” if you try & parse out (by making a whole range of invalid assumptions) how voters felt about one candidate as opposed to only one other candidate, which is almost impossible to do when the race was actually a five-way race…not a two-way race!!
BTW, I can keep this up as long as you’d like to continue trying Mr. Shentrup…you just let me know when you want to give it up…
March 26, 2009 at 11:49 pm
Clay Shentrup
Mister Guy,
The reason that purely hypothetical election matters is because it means that there’s at least a 50% chance that IRV elected the wrong candidate in the REAL election.
And you can indeed mathematically prove that. I’ll link you to the same simple mathematical proof I cited above, verified by a Princeton math Ph.D. who specializes in election science:
http://scorevoting.net/PuzzClay2Irv.html
In layman’s terms, if X goes from winner to loser by being ranked HIGHER, then either X wasn’t the right winner in the first place, or X still should be the winner (but isn’t).
As for Montroll’s huge majority over Kiss, I’m not making any assumptions. I’m going off the actual ballot data.
PLEASE read Warren Smith’s full analysis before continuing to repeatedly deny proven facts.
Lastly, I don’t doubt that you have the energy to continue this dialog as long as I can. But this is about verifiable facts, not who has more enthusiasm.
March 27, 2009 at 2:22 am
Clay Shentrup
Mister Guy,
Here’s the argument ender as far as I’m concerned:
http://scorevoting.net/BurlResponses.html
Good day sir.
March 28, 2009 at 8:58 pm
Mister Guy
“The reason that purely hypothetical election matters is because it means that there’s at least a 50% chance that IRV elected the wrong candidate in the REAL election.”
More utter nonsense, which can be only taken seriously if you start flipping the way people actually voted willy-nilly without taking into account WHY they likely voted the way that they did in the first place. BTW, you’ve already clearly stated that you simply don’t care about why they voted the way they did!
As for your buddy who’s supposedly “a Princeton math Ph.D. who specializes in election science”, he’s really just another obvious shill for range voting…just like you are!
“I’m going off the actual ballot data.”
Which simply cannot be used for the type of “analysis” that you are trying to do, period.
“PLEASE read Warren Smith’s full analysis before continuing to repeatedly deny proven facts.”
LOL…I’ve already commented *extensively* on his almost completely bogus analysis…that’s what this thread has been all about!
“But this is about verifiable facts, not who has more enthusiasm.”
I agree, but, if I were you, I’d be worried as being exposed for the range voting shill that you are that both has it in for IRV & who doesn’t give a damn why voters voted the way that they did. THAT’s the debate ender!!
————————————————————
“Those 4 officially ‘valid’ ballots each co-ranked two candidates top, one of them a write-in”
…which you simply can’t do in an IRV election, period.
“These were counted as ‘Wright>Kiss>(all others)’ which probably was not the intent of those voters. They probably intended ‘Wright>(all others)>Kiss.'”
It doesn’t matter either way, since those votes were given to Wright as first choice votes, and the rest of the rankings on those ballots were never checked, since Wright was one of the top two vote getters!
“Specifically, Wright’s supporters preferred Montroll over Kiss by over a 3:1 ratio. Because they voted for Wright, IRV ignored their stated Montroll>Kiss preferences”
…because they simply don’t matter under this IRV scenario, period.
“That is, the fact that the Kiss supporters preferred Montroll>Wright by over a 5:1 ratio was ignored by plain-plurality voting (and also by IRV)”
…again, because it simply doesn’t matter under this IRV scenario!
“That, rather arbitrarily, unfairly, and whimsically, gave some voters more say than others.”
Bullshit…this go-to whine line from IRV opponents has been tried again & again, and it simply falls flat on it’s face every, single time. IRV more fully allows voters to express their true feelings about who should occupy a particular political office.
“I mean, even Adolf Hitler and Bernie Madoff probably managed that.”
Wow, and this was stated in the “Personal attacks” part of this guy’s latest argument! LOL…what’s good for the goose isn’t good for the gander eh? Wow…
“The media also did not mention (that we noticed) the non-monotonicity of this election”
…because, as discussed at length before, it simply *does not matter*!
“Or the fact Montroll beat all rivals by majority preference-counts”
…because that’s simply not true when looking at the *five-way race* that this election was really about!
“Burlington Vermont is a rather exceptionally left-wing place since 64% of its voters were registered as Democrats and it has a long recent history of winners from the Progressive Party, including Kiss himself. (Obama beat McCain by a thumping 17252-to-3086 margin in Burlington 2008.) This caused the Republican Wright arguably to be the ‘third-party’ candidate, while the Democrat Montroll arguably was the ‘centrist.'”
Well, this might be true, if VT ever held a lot of it’s local elections in the fall instead of holding them in the middle of winter. The FACT is that voter turnout on VT Town Meeting Day has been waaaay lower in recent years than in the November 2008 election. I for one hope that VT eventually dumps Town Meeting Day, but I’m not holding my breath at all…because I know VT pretty well at this point. There were only around 8400 voters in this recent Burington election, period.
“What they do is assign preferences to voters that the voters themselves did not indicate on their ballots.”
Exactly, this is because they apparently don’t care about WHY voters vote the way that they do. They’d rather focus most of their efforts on “alternate-realities” which will likely never come true in the real world. This is a *very* common tactic used by those that try to manipulate science & math in ways to confuse the average person.
“There were 2008 W voters who did express a preference between M and K in their vote; they preferred M over K by over a 3-to-1 ratio”
…which, again, doesn’t matter one bit in this IRV scenario.
“The problem is that there also were 1289 W voters who did not express any preference between M and K.”
There is NO “problem” here. Bullet-voting is a perfectly valid method of voting, which is usually used by people that really, really want their candidate to win over all other options.
“Now of course, it is inconceivable that all these 1289 voters really absolutely did not care by even the tiniest jot who won among M or K.”
Sure it is. In fact, it is more likely that these voters really, really didn’t want either Kiss or Montroll to win…so they didn’t bother ranking them at all. This is totally valid voting stragtegy, period. I’ve actually used it in the many IRV elections that I’ve participated in!
“The truth is those voters simply did not care enough to express a preference in their vote”
Wrong…Burlington isn’t Australia…you simply can’t apply a different set of rules after an election has been had.
“Those two hypotheses both would be defensible”
…again, only in some “alternate reality” where the actual will of the voters doesn’t matter at all.
“In short: if FairVote wishes to claim a no-show paradox did not occur then they are assuming that the ‘silent’ Wright voters must have preferred Kiss over Montroll by about a 4:1 (or greater) ratio.”
This is sheer nonsense, since I have yet to read of such a claim from the FairVote people. Wow, your buddy here really, really has it in for FairVote eh?? Even a dead man could notice that, and it speaks *volumes* about the speaker & his apparent motivations.
“You the (range)voter have to make that judgment on a case-by-case basis each time you vote”
…and so does the IRV voter.
“In that sense obeying MM (but not Condorcet) arguably is undesirable”
…in the narrow opinion of the writer of this “analysis”, who, of course, happens to be a range voting shill.
“Using the exact same rules and same votes except that some Wright voters switched their vote to Kiss”
…which, again, would likely never happen…so it simply doesn’t matter. Woulda, coulda, shoulda…that’s ALL you guys have at the end of every, single day…ugh…
“We also are a little worried about what would have happened inside his mind if the crux Wright↔Kiss switching voter who made the winner change the ‘wrong way’ had, in fact, been Rama Schneider.”
LOL…do you guys even KNOW who Rama is?? Obviously not, since he’s one of the most staunch & outspoken Progressive-types that I know of in the state of VT! Again, a LOT of your “analysis” is extremely flawed because you simply don’t understand why the voters in Burlington voted the way that they did. But, of course…once again, you don’t care about that in any event, which speaks *volumes* for your side. Hint: It’s not a winning strategy in the long run…
March 29, 2009 at 2:04 am
Clay Shentrup
To Mister Guy,
>> “The reason that purely hypothetical election matters is because it means that there’s at least a 50% chance that IRV elected the wrong candidate in the REAL election.”
> More utter nonsense, which can be only taken seriously if you start flipping the way people actually voted willy-nilly without taking into account WHY they likely voted the way that they did in the first place.
Well, I cited a straightforward mathematical proof. So it’s not “utter nonsense”, it’s just a proven fact. You support that position by failing to cite any problems with the proof.
And to be clear, I did not “start flipping the way people actually voted”. I cited the ACTUAL BALLOT DATA to contrast it with an ALTERNATE and contradictory IRV election result, to demonstrate (PROVE by example) that IRV produces the wrong winner in at least one of each two non-monotonicity halves.
“BTW, you’ve already clearly stated that you simply don’t care about why they voted the way they did!”
The relevant point isn’t that I don’t CARE about it. It’s that it doesn’t MATTER. Nowhere in that proof do we make any assumptions of voters’ reasoning for their rankings. We simply show that, GIVEN they have those preferences, IRV gives them the wrong result in at least one of the two scenarios.
> As for your buddy who’s supposedly “a Princeton math Ph.D. who specializes in election science”, he’s really just another obvious shill for range voting…just like you are!
No, we are not shills: http://dictionary.reference.com/browse/shill
And even if we were, that is a personal attack, not a critique of the evidence and/or arguments we’ve presented regarding this Burlington IRV election. You can call us names, but you can’t refute the facts we’ve presented.
>> “I’m going off the actual ballot data.”
> Which simply cannot be used for the type of “analysis” that you are trying to do, period.
Yes it can be. In fact, it was. If you can find any methodological errors, please cite them.
> LOL…I’ve already commented *extensively* on his almost completely bogus analysis…that’s what this thread has been all about!
You’ve just said things like “that’s utter nonsense”, which is no better than saying “my opponent is wrong”. You’ve proven unable to actually justify that by citing some logical or data errors.
> if I were you, I’d be worried as being exposed for the range voting shill that you are that both has it in for IRV & who doesn’t give a damn why voters voted the way that they did. THAT’s the debate ender!!
No, that’s a personal attack. In logical and rhetoric parlance it’s called an “ad hominem” logical fallacy. I’ve cited robust evidence proving my case that IRV exhibited non-monotonicity failure and other problems in Burlington. The accuracy of those facts does not in any way depend on my personal status as a “shill” or any other personal attribute.
I think it is telling that you do not exhibit any understanding of this, or even seem to know what a logical fallacy is.
>>“Those 4 officially ‘valid’ ballots each co-ranked two candidates top, one of them a write-in”
> …which you simply can’t do in an IRV election, period.
So you’re AGREEING with Warren Smith that these ballots were invalid. Good. (Unfortunately you do not even seem to understand that you are agreeing with him.)
>> “These were counted as ‘Wright>Kiss>(all others)’ which probably was not the intent of those voters. They probably intended ‘Wright>(all others)>Kiss.’”
> It doesn’t matter either way, since those votes were given to Wright as first choice votes, and the rest of the rankings on those ballots were never checked, since Wright was one of the top two vote getters!
Yes it DOES matter if we’re specifically talking about the ballot validity/spoilage rate, which is in fact precisely what Smith was talking about in that paragraph.
>> “Specifically, Wright’s supporters preferred Montroll over Kiss by over a 3:1 ratio. Because they voted for Wright, IRV ignored their stated Montroll>Kiss preferences”
> …because they simply don’t matter under this IRV scenario, period.
They most certainly DO matter, in that they are relevant to our points that
1. IRV does not always elect a majority winner, contrary to claims that it does.
2. IRV can give voters an incentive to be insincere (e.g. these voters had an incentive to top-rank Montroll instead of Wright).
>> “That is, the fact that the Kiss supporters preferred Montroll>Wright by over a 5:1 ratio was ignored by plain-plurality voting (and also by IRV)”
> …again, because it simply doesn’t matter under this IRV scenario!
Wrong. I just listed at least two ways that it DOES matter.
Once you get to the point that you’re cursing, I have to go ahead and bow out. There’s not much point in trying to have an objective fact-based discussion with you when you’ll counter a mathematical proof with a factually empty response like “utter nonsense”. To anyone reading this thread with an eye for the truth, the point has been made.
April 1, 2009 at 7:04 pm
Mister Guy
“You support that position by failing to cite any problems with the proof.”
Ugh…I’ve already told you several times what is wrong with this kind of “analysis”…it completely & totally ignores the way people actually voted and, much more importantly, WHY they likely voted the way that they did in the first place!
“I did not ‘start flipping the way people actually voted’.”
Yes, you absolutely did, and it’s there in the “analysis” for everyone to see.
“The relevant point isn’t that I don’t CARE about it. It’s that it doesn’t MATTER. Nowhere in that proof do we make any assumptions of voters’ reasoning for their rankings. We simply show that, GIVEN they have those preferences, IRV gives them the wrong result in at least one of the two scenarios.”
LOL…keep trying to change the subject here…it’s not working. At the end of every, single day, you’ll be left with the cold, hard truth that you simply DO NOT care about the will of the voters as was expressed in this most recent IRV election. In fact, you care MORE about some purely hypothetical scenario that would have little chance of EVER becoming true in the real world that is Burlington politics, period.
“No, we are not shills”
“And even if we were, that is a personal attack, not a critique of the evidence and/or arguments we’ve presented regarding this Burlington IRV election.”
LOL…you’re not “a person who publicizes or praises” range voting “for reasons of self-interest, or friendship or loyalty”?? Please…OF COURSE you are, and it’s exceedingly obvious!
Also, you really, really must not be serious at all about “personal attacks”, since your buddy makes a HUGE number of them against his chosen opponets…like I clearly pointed out above…even when he is in the midst of complaining about their supposed personal attacks on him…lol…that is rich!!
“If you can find any methodological errors, please cite them.”
I already did…the race was a 5-way race…NOT a two-way race!
“You’ve proven unable to actually justify that by citing some logical or data errors.”
Deny, deny, deny…that’s the mode that you’ve been in for quite some time now…saying basically that “our analysis isn’t flawed at all because it’s a ‘mathematical’ analysis”, which, again, may work on someone that doesn’t know much about science or mathematics, but it’s not working on me my friend…give it up…
“No, that’s a personal attack.”
LOL…no no my friend…it’s called the TRUTH, which you simply can’t refute or run away from anymore…now deal with it…
“The accuracy of those facts does not in any way depend on my personal status as a ‘shill’ or any other personal attribute.”
LOL…yes, pay no attention to that man behind the curtain!
“So you’re AGREEING with Warren Smith that these ballots were invalid. Good. (Unfortunately you do not even seem to understand that you are agreeing with him.)”
Really?? I’m agreeing with him huh?? Then why, praytell, does he go on to say this about those very same ballots:
“and since they were valid, we have trouble seeing why they were not counted.”
LOL…did you even READ the “analysis” that you claim to champion so much?? Those ballots were NOT counted in this past election because there were invalid ballots, period.
Arguing over whether the non-spoilage rate in this recent election was 99.99% or 99.96% is something that only a desperate shill of the first order would do in order to further the cause that he lives by, period end of story.
“Yes it DOES matter if we’re specifically talking about the ballot validity/spoilage rate, which is in fact precisely what Smith was talking about in that paragraph.”
Except for the FACT that those ballots were NOT spoiled at all as related to the way that the votes were counted…ugh…
“They most certainly DO matter, in that they are relevant to our points that
1. IRV does not always elect a majority winner, contrary to claims that it does.
2. IRV can give voters an incentive to be insincere (e.g. these voters had an incentive to top-rank Montroll instead of Wright).”
There you go again…look, you don’t like IRV at all…that much is obvious, which is why you simply ignore the FACT that in this particular IRV election, the way that Wright first place ranking voters voted further down their ballot simply isn’t relevant to the way this election panned out in the real world.
Also, IRV DOES NOT lend itself to having voters be “insincere” at all…it lends itself, in fact, to voters being MORE honest about how they would prefer candidates without fearing a spoiler. What you ASSUME here about the Wright first place ranking voters is that they really, really wanted to get rid of Bob Kiss, which is likely not the case at all. Those voters were, in fact, much more likely to simply really, really want their first place candidate Kurt Wright to be elected over ALL other names that were on the ballot, period.
“I just listed at least two ways that it DOES matter.”
No, you really didn’t…you just regurgitated the mantra/propaganda of those that are opposed to IRV, period.
“Once you get to the point that you’re cursing, I have to go ahead and bow out.”
When exactly did I “curse” at you Mr. Shentrup?? Don’t start making up reasons for why you want to “bow out” of this discussion now…grow some guts & just leave if you want to…
“To anyone reading this thread with an eye for the truth, the point has been made.”
I agree ho-heartedly!
February 1, 2010 at 11:09 pm
robert bristow-johnson
just to let y’all know that in a month, Burlington voters will vote on repealing IRV. it’s sad that it came to this, because the old Plurality Rule is worse (we know that!) and a Ranked-Order Ballot is necessary to collect enough information from voters as to how they would vote if they could not get their first choice, but the sad thing is (just like the 2-party system), our choice next month will be between Dumb (IRV) and Dumber (Plurality).
how long will it be that Condorcet continues to be passed over?
February 2, 2010 at 12:06 am
CS
Robert Bristow-Johnson,
Score Voting (aka Range Voting) and its simplified variant called Approval Voting are much better systems that any Condorcet method, in light of their resistance to the effects of tactical voting, and with regard to simplicity and the associated costs and voter education needs.
http://scorevoting.net/CondorcetExec.html
There’s even an argument that these methods actually elect Condorcet winners MORE OFTEN than real Condorcet methods.
http://scorevoting.net/AppCW.html
June 30, 2019 at 7:06 pm
John Moser
Instant Runoff Voting can be constricted to Smith/IRV; I have constructed examples where Smith/IRV has a soft failure:
10: A>B>C>D
6: B>C>D
15: F>E>D
In that example, candidates {A,B,C,D,E,F} are the Smith set, and eliminating B gives {A,C,D,E,F}, then eliminating C gives {D} as the Smith set. Tideman’s Alternative elects D; IRV continues the runoff and eventually elects F.
If the votes are instead:
10: A>B>C
6: B>C>D
15: F>E>D
Then the Smith Set is all candidates, and runoff eventually gives you a Smith Set of F. Further runoff gets you down to just F. Both of these vote sets imitate a linear single ideological axis from A to F; IRV elects the extreme F in both examples; and Tideman’s Alternative elects D with sufficient data.
IRV satisfies the Condorcet Loser criterion: if there is a Condorcet Loser, IRV does not elect it. By adding a candidate who loses to the Condorcet Loser, it becomes possible to elect the previous Condorcet Loser under IRV, so IRV can elect any candidate (as with Plurality).
The difference between Tideman’s Alternative and Smith/IRV is as such:
1. Eliminate all non-Smith candidates;
2. If only one candidate remains, go to 6;
3. Eliminate the remaining candidate with the fewest votes;
4. If Tideman’s Alternative, go to 1;
5. If Instant Runoff Voting, go to 2;
6. Elect the single remaining candidate.
Regular IRV enters at 2 instead of 1. I use a variant of Tideman’s Alternative that checks the Schwartz set for Condorcet, and otherwise eliminates all non-Smith candidates.
I showed that the above Condorcet failure extends even to Tideman’s Alternative. It’s an interesting failure: F is the Plurality winner.
Mathematically, you can demonstrate that Condorcet is Plurality when voters truncate their exposed preferences to 1; and you can show that this occurs even when preferences are deeper, such that the Plurality winner is the Condorcet winner. It demonstrably occurs in A23, a historical real-world election with 92 voters and 29 candidates.
Because of all this, primary elections are important; yet party primary can be mathematically demonstrated to produce non-representative results.
I’ve suggested a nonpartisan blanket primary by Single Transferable Vote because STV sort of generalizes this failure: for 1 winner, voters ranking no fewer than N candidates, and C total candidates, Condorcet performs as if voters are ranking all candidates; for W winners, voters ranking no fewer than N candidates, and C*W total candidates, STV performs as if voters are ranking all candidates.
Because you can get a representative breakdown of proportions, you can reliably select candidates representing the voters in such a way that the general election outcome approximates the outcome of a Condorcet social choice function run against the PRIMARY election with all voters accurately ranking all candidates.
I chose ranked methods for practical concerns: just as voters can’t reliably know about a great number of candidates (marginal utility of research, constraints on individual time), voters are bad at rating things in cardinal terms. Rating is a scientific process, and humans are universally bad at it—so much so that online ratings of as much as 1.5 stars difference out of 5 are hardly any better than a coin toss at predicting whether one product is superior to another. By contrast, when people DIRECTLY compare two alternatives, they are VERY GOOD at determining which alternative they prefer. Ordinal rankings carry greater accuracy than cardinal ratings because cardinal ratings are terminally imprecise.
Because of the lack of precision and lack of standardization between cardinal ratings, it’s impossible to determine any kind of social welfare from the data. That is: an SCF exists that determines social welfare effectively from cardinal ratings; but cardinal ratings of sufficient quality to produce a valid result do not exist, as humans are bad at creating those ratings in practice.
Besides this, score systems and the simplified Approval system are heavily-vulnerable to and even DEMANDING OF tactical voting, since any vote directly helps an opponent and must be carried out in consideration of whether your candidate can win and whether a less-preferred candidate will defeat both your first and second choices if you don’t rate them accordingly (and sacrifice your first choice in the process).
Under maximum strategy, Approval selects the Condorcet candidate WHEN voters all have perfect knowledge about all other voters—a condition in which you can simply announce that you’re skipping the election and dictating the winner and NOBODY WILL COMPLAIN because they all know you’re dictating the winner who would have been elected. In real-world large elections, approval causes voter anxiety in close elections.
Ranked votes cannot be used to test approval voting because it’s safe to rank an anti-approved candidate simply to show that an even less-favored candidate is worse, such as ranking Ted Cruz and not ranking Donald Trump when your ranks are Bernie>Hillary>Rubio>Cruz. Your approval votes might be Bernie and Hillary—IF you believe there are more democrats than republicans out, or that Rubio republicans will approve Hillary. If you think the opposite, you have to approve Rubio or you risk getting stuck with Trump.