<div dir="auto">Then Plurality is better than Black & Score; & Benham & Woodall are no better than IRV; & IRV is 4 times better than the best Condorcet methods? </div><div dir="auto"><br></div><div dir="auto">Of course JGA’s results lead to similar questions.</div><div dir="auto"><br></div><div dir="auto">Doesn’t this bring into question the meaningfulness & usefulness of manipulability as a measure of merit?</div><div dir="auto"><br></div><div dir="auto">IRV has an incomparably worse strategy problem than the best Condorcet methods.</div><div dir="auto"><br></div><div dir="auto">Plurality is incomparably strategically worse than Approval.</div><div dir="auto"> </div><div dir="auto">Benham & Woodall, as everyone would agree, significantly improve on IRV.<br></div><div dir="auto"><br></div><div dir="auto"><br></div><div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Sat, Apr 27, 2024 at 04:38 Kristofer Munsterhjelm <<a href="mailto:km_elmet@t-online.de">km_elmet@t-online.de</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Here are voter manipulability stats for some of the poll methods, using <br>
James Green-Armytage's spatial model with 4 dimensions, 4 candidates and <br>
99 voters. Each method is tested on 500k elections, with 32k attempts to <br>
strategize per election.<br>
<br>
The manipulability value is the fraction of elections in this model <br>
where the method elected a unique winner, and voters who preferred <br>
somebody else to the current winner could get that somebody elected by <br>
changing their ballots. Note that it does *not* check strategic nomination.<br>
<br>
I've prefixed entries that aren't actually part of the poll with an <br>
asterisk. I'll explain later why I've included them. Entries prefixed <br>
with a number sign are from JGA as my simulator doesn't support them.[1]<br>
<br>
The simulator uses full ballots, so Smith//DAC is the same as <br>
Smith//DSC. If truncation would make the method more resistant, that's <br>
not reflected here.<br>
<br>
0.698 *Borda<br>
0.668 #Approval (from JGA)<br>
0.545 Condorcet//Borda (Black)<br>
0.480 Copeland//Borda (Ranked Robin)<br>
<br>
0.417 Plurality<br>
0.417 Smith//DAC<br>
0.412 *BTR-IRV<br>
<br>
0.350 Baldwin<br>
0.333 Raynaud (Gross Loser Elimination)<br>
0.333 Schulze(wv)<br>
0.332 Minmax(wv)<br>
0.321 Ranked Pairs(wv)<br>
<br>
0.075 Woodall, Schwartz-Woodall<br>
0.074 RCIPE<br>
0.074 IRV<br>
0.074 Benham<br>
<br>
I've included Borda to show that my results are similar to James Green <br>
Armytage's. (Compare also the results minmax results.) In addition, I've <br>
included BTR-IRV to see how well it would do. Too bad it didn't do <br>
better, though...<br>
<br>
My simulator show higher manipulability for IRV and the Condorcet-IRV <br>
hybrids than JGA's simulator did. I think this comes down to that my <br>
simulator is more thorough and thus is able to uncover more <br>
nonmonotonicity- and pushover-related strategies.<br>
<br>
Finally, I'm working on the cardinal methods, but the devil's in the <br>
details so it's taking a lot longer than expected. More about that when <br>
I've solved it. But my preliminary tests put Smith-Range around 0.5 and <br>
STAR as worse than this. Unnormalized (fixed scale) Range, though not a <br>
poll method, even does worse than Borda. And if the preliminary tests <br>
give some indication, all the Range/Score-based methods do worse than <br>
Plurality.<br>
<br>
-km<br>
<br>
[1] Green-Armytage, James (2011). "Four Condorcet-Hare hybrid methods <br>
for single-winner elections". Voting matters (29): p. 7; <br>
<a href="https://www.votingmatters.org.uk/ISSUE29/I29P1.pdf" rel="noreferrer" target="_blank">https://www.votingmatters.org.uk/ISSUE29/I29P1.pdf</a><br>
----<br>
Election-Methods mailing list - see <a href="https://electorama.com/em" rel="noreferrer" target="_blank">https://electorama.com/em</a> for list info<br>
</blockquote></div></div>