Loading ...
Sorry, an error occurred while loading the content.

Re: Roeck's Measure of Compactness / Minimum Bounding Circle

Expand Messages
  • jimrtex4192
    ... You misunderstood my question. BTW, who is you guys referring to? The State of Ohio is in the process of doing its legislative and congressional
    Message 1 of 19 , Sep 1, 2011
    • 0 Attachment
      --- In RangeVoting@yahoogroups.com, Jonathan Hall <edoguiguy@...> wrote:

      > This is the kind of plan that you guys should be working on, not some mathematical fantasy.   Think about the real world with people in it.

      You misunderstood my question. BTW, who is "you guys" referring to?

      The State of Ohio is in the process of doing its legislative and congressional redistricting for the next decade. They have solicited input from the public, and have provided mapping software to help the public generate plans. That software calculates Roeck's measure of compactness, though there is nothing in the Ohio Constitution other than a requirement that districts be "compact".

      There is also a group promoting a "contest" to produce the "best" plan. They are submitting the winning plans to the State of Ohio - mainly to promote the "contest" methodology, but also their particular scoring criteria. They also use Roeck's measure.

      I noticed that the two produced different values of the same measurement. Since my earlier note, the group doing the software used for the contest have agreed that it is an error in their calculation. They may or may not appreciate any assistance in an algorithm for the minimum spanning circle for a district (or its convex hull).
    • Jonathan Hall
      You misunderstood my question. BTW, who is you guys referring to? Yes I did.   So You Guys is withdrawn. The real point that should be made tirelessly is
      Message 2 of 19 , Sep 2, 2011
      • 0 Attachment
        You misunderstood my question. BTW, who is "you guys" referring to?

        Yes I did.   So "You Guys" is withdrawn.

        The real point that should be made tirelessly is that shapes can not matter.   Or more that adhearance to them would be harmful to democracy.  A shape based system would of necessity work against rural districts because the disruptions of natural boundaries/obstacles would be more likely to be introduced.   So right on the face of any merely shape based plan comes an attack on a specific constituency.   I would not hesitate to file a suit against such 'logic' if it was tried here in California.

        Jonathan Hall

        [Non-text portions of this message have been removed]
      • WarrenS
        ... --pathetic. But seems par for the course. The Roeck measure is a poor one anyhow, but even lamer if you compute it wrong. ... --correct. If you have a
        Message 3 of 19 , Sep 3, 2011
        • 0 Attachment
          --- In RangeVoting@yahoogroups.com, "jimrtex4192" <jimrtex@...> wrote:
          >
          > One of the criteria that the Ohio redistricting contest,
          >
          > http://drawthelinemidwest.org/ohio/
          >
          > uses for scoring the contest is compactness. In particular, they use Roeck's measure of compactness for measuring the compactness of a district, and then use mean of the district compactness scores to determine a compactness for an entire plan.
          >
          > The Ohio Secretary of State is also soliciting public plans for the Ohio Apportionment Board, which is charged with drawing legislative districts. Congressional districts are drawn by the legislature, but the SOS is also providing logistical support for there efforts.
          >
          > http://www.sos.state.oh.us/SOS/ReshapeOhio.aspx
          >
          > The software used by the SOS also calculate Roeck's measure of compactness, and the two sets of software **differ**

          --pathetic. But seems par for the course. The Roeck measure is a poor one anyhow,
          but even lamer if you compute it wrong.

          > Roeck's measure of compactness is defined as the ratio of the area of the district (polygon) divided by the area of a minimum spanning circle. A circular district would a Roeck's measure of 1.00 (100%), while a square would be 2/pi = 63.7%.
          >
          > The software for the redistricting contest is open source.
          >
          > https://github.com/PublicMapping/DistrictBuilder/blob/master/django/publicmapping/redistricting/calculators.py
          >
          > Am I correct that their error is that the minimum spanning circle is not (necessarily) centered on the centroid of the convex hull of the district polygon.

          --correct. If you have a set of points in the plane, one way to find the minimum spanning circle is
          (1) find their convex hull (say it is an N-gon)
          (2) for every 3 among those N extreme points, find the circle which
          passes thru all 3.
          (3) if the circle encloses (or touches) all the other N-3 points, then it is
          a "candidate" circle
          (4) for every 2 among the N extreme points, find circle with those 2 points
          as its diameter
          (5) if the circle encloses (or touches) all the other N-2 points, then it is
          a "candidate" circle
          (6) output the smallest candidate circle.

          The algorithm I just outlined takes order N^4 time to run once the convex hull is found.
          There are faster algorithms, however. The best ones run in O(N) time. The first
          O(N)-time algorithm was found by Nimrod Megiddo, but it was rather complicated.
          Simpler algorithms were later found by Emo Welzl that employ randomization.
          There is now public source software, called "miniball" for a version of Welzl's algorithm
          (implemented by Bernd Gaertner), you can download it from here:
          http://www.inf.ethz.ch/personal/gaertner/miniball.html

          > Is it also true that the minimum spanning circle for the district, and the minimum spanning circle for the convex hull are the same?

          --yes.
        • WarrenS
          ... --it seems to me (which I think is disagreeing with you?) that GERRYMANDERING is extremely harmful to democracy. Currently re-election rates in congress
          Message 4 of 19 , Sep 3, 2011
          • 0 Attachment
            --- In RangeVoting@yahoogroups.com, Jonathan Hall <edoguiguy@...> wrote:
            >
            > You misunderstood my question. BTW, who is "you guys" referring to?
            >
            > Yes I did.   So "You Guys" is withdrawn.
            >
            > The real point that should be made tirelessly is that shapes can not matter.   Or more that adhearance to them would be harmful to democracy. 

            --it seems to me (which I think is disagreeing with you?) that GERRYMANDERING is
            extremely harmful to democracy. Currently re-election rates in congress are about 98% meaning voters only get to play a role in about 4% of elections. That really is not democracy at all.
            Gerrymandering usually simply cannot happen if the district shapes are demanded to be sensible (various mathematical definitions of "sensible" work here). Therefore, if we simply demand sensible district shapes, the gerrymandering problem is done with forever.
            What is the cost of that? Well, sometimes the mathematical shape-quality measures, disagree with certain intuitive notions of shape-quality. If you examine the actual district maps for all 50 states produced by the shortest split line algorithm or by Olson's computer program, and compare them with the official human-drawn distrct maps, then
            you can see for yourself which does a better job. It is my opinion that SSA does a better job than current legislators do 98% or more of the time, and SSA defintely does it for
            far less monetary cost far faster. Here are the links to go to to see the pictures:
            http://www.rangevoting.org/Olson2000list.html
            http://www.rangevoting.org/SplitLR.html

            > A shape based system would of necessity work against rural districts because the disruptions of natural boundaries/obstacles would be more likely to be introduced.   So right on the face of any merely shape based plan comes an attack on a specific constituency.   I would not hesitate to file a suit against such 'logic' if it was tried here in California.
            > Jonathan Hall

            --well... it seems to me that you are in favor of gerrymandering and against democracy.
            I have no doubt that lawsuits will be filed against virtually every anti-gerrymandering plan. For example that immediately happened in Florida which recently passed a (rather ineffectual-sounding) anti-gerrymandering referendum. The lawsuits failed to stop it
            (so far anyhow) but it remains to be seen whether this referendum will actually have a
            useful anti-gerrymandering effect.

            --re "work against rural districts" we actually found the opposite. That is,
            Olson's algorithm seems to (at least with present USA parties and demographics) ENHANCE the power of rural versus city voters. However, our conclusions on that question are not very convincing at present. But they are more convincing than your (utterly unsupported) claim... since we actually did examine some maps trying to carefully count...
          • WarrenS
            To see an example where (I think) Jonathan Hall has a good point, check Virginia, whose splitline districting map based on 2000 census data is here:
            Message 5 of 19 , Sep 3, 2011
            • 0 Attachment
              To see an example where (I think) Jonathan Hall has a good point, check Virginia,
              whose splitline districting map based on 2000 census data is here:

              http://www.rangevoting.org/Splitline2009/va.png

              and note that the green district in the NE "looks bad."
              You can also look at

              http://www.rangevoting.org/Olson2000/VA_ba.png

              which shows the official district map (left) and Olson's computer-generated
              map (right). Despite the problem in that district I still think overall both Olson's
              and SSA's districtings are clearly more democratic and less gerrymandered than the official maps.

              Also, note the Delmarva peninsula, which is disconnected from the rest of VA, is
              dealt with worse by both Olson & official, than by SSA, so J.Hall's kind
              of criticism backfires in this case (SSA is doing better, not worse). I picked VA since
              it is a state in which SSA seems to exhibit a flaw. VA is a atypically
              bad state for SSA, but I still think SSA is superior to the official gerrymander.

              Check the maps for the 50 states and decide for yourself; this is an experimental question.
            • James Borda
              ... Jonathan Hall might want to check out Chen & Rodden s paper: *Tobler’s Law, Urbanization, and Electoral Bias: Why Compact, Contiguous Districts are Bad
              Message 6 of 19 , Sep 3, 2011
              • 0 Attachment
                >
                >
                > > A shape based system would of necessity work against rural districts
                > because the disruptions of natural boundaries/obstacles would be more likely
                > to be introduced. So right on the face of any merely shape based plan
                > comes an attack on a specific constituency. I would not hesitate to file a
                > suit against such 'logic' if it was tried here in California.
                > > Jonathan Hall
                >
                >>
                > --re "work against rural districts" we actually found the opposite. That
                > is,
                > Olson's algorithm seems to (at least with present USA parties and
                > demographics) ENHANCE the power of rural versus city voters. However, our
                > conclusions on that question are not very convincing at present. But they
                > are more convincing than your (utterly unsupported) claim... since we
                > actually did examine some maps trying to carefully count...
                >
                >
                Jonathan Hall might want to check out Chen & Rodden's paper: *Tobler�s Law,
                Urbanization, and Electoral Bias: Why Compact, Contiguous Districts are Bad
                for the Democrats*
                *
                *

                ABSTRACT: When one of the major parties in the United States wins a
                substantially larger share of the seats than its vote share would seem to
                warrant, the conventional explanation lies in manipulation of maps by the
                party that controls the redistricting process. Yet this paper uses a unique
                data set from Florida to demonstrate a common mechanism through which
                substantial partisan bias can emerge purely from residential patterns. When
                partisan preferences are spatially dependent and partisanship is highly
                correlated with population density, any districting scheme that generates
                relatively compact, contiguous districts will tend to produce bias against
                the urban party. In order to demonstrate this empirically, we apply
                automated districting algorithms driven solely by compactness and contiguity
                parameters, building winner-take-all districts out of the precinct-level
                results of the tied Florida presidential election of 2000. The simulation
                results demonstrate that with 50 percent of the votes statewide, the
                Republicans can expect to win around 59 percent of the seats without any
                �intentional� gerrymandering. This is because urban districts tend to be
                homogeneous and Democratic while suburban and rural districts tend to be
                moderately Republican. Thus in Florida and other states where Democrats are
                highly concentrated in cities, the seemingly apolitical practice of
                requiring compact, contiguous districts will produce systematic
                pro-Republican electoral bias.


                [Non-text portions of this message have been removed]
              • Jonathan Hall
                Mr. Borda,  The abstract seems to suggest the exact kinds of shinanigans that I am trying to prevent.  My system uses a human measure of compactness, so
                Message 7 of 19 , Sep 4, 2011
                • 0 Attachment
                  Mr. Borda,

                   The abstract seems to suggest the exact kinds of shinanigans that I am trying to prevent.  My system uses a human measure of compactness, so instead of indifferent mathemtical butchering, communities with centers are respected and scored by how they can internally communicate.   This is a neccesary criteria that is new to the discussion but still vital.  Large city centers will not get choped up with this method.   This new criteria may not favor democrats, but its unlikey to penalize them.   The whole idea of gerrymandering is to spread constituentcies out thus dilluting them.  But the compact human communities are I think will break up constituencies even less then some arbitrary mathmatical method.

                  Boundries will fall where there are sparce populations and follow along regions that lack viable roads.  At some point communities do have to get seperated because some will have more people then a district can have, but with this method there will still be the one or more district covering the mass and the slicings are pushed away from the vital centers.

                  If need be a simple modification is possible.   Rather then just adding drive times to the area centers add the squares of drive times.  This will definitely penalize tortured districts.   But I think with just the simple drive time sum, the maps will find the best way for all districts to have a core.

                  Jonathan Hall

                  [Non-text portions of this message have been removed]
                • jimrtex4192
                  ... In this case, they are not using shape as the driving force, but rather as part of their scoring criteria. The idea is that you can draw fair and
                  Message 8 of 19 , Sep 4, 2011
                  • 0 Attachment
                    --- In RangeVoting@yahoogroups.com, Jonathan Hall <edoguiguy@...> wrote:
                    >
                    > You misunderstood my question. BTW, who is "you guys" referring to?
                    >
                    > Yes I did.   So "You Guys" is withdrawn.
                    >
                    > The real point that should be made tirelessly is that shapes can not matter.   Or more that adherence to them would be harmful to democracy.  A shape based system would of necessity work against rural districts because the disruptions of natural boundaries/obstacles would be more likely to be introduced.   So right on the face of any merely shape based plan comes an attack on a specific constituency.   I would not hesitate to file a suit against such 'logic' if it was tried here in California.

                    In this case, they are not using shape as the driving force, but rather as part of their scoring criteria. The idea is that you can draw "fair" and "competitive" districts without use of odd shaped districts.

                    The Ohio constitution requires House districts to be composed of whole counties, and if that is not possible of whole counties and whole townships. Further, it requires that in counties entitled to more than one house district, that as many whole districts be created within the county, with any remnant created in a district shared with neighbors. Ohio counties are populous and squarish enough that the largest districts can only have 3 or 4 counties, and it is pretty hard to create districts that are truly non-compact.

                    Within counties, townships are few enough and populous enough that you can't really arrange them in non-compact shapes and get the right population.

                    So the compactness criteria gives an incentive to vary the population of districts to increase the apparent compactness. So you might have a line of 3 townships and not add the 4th, even though it would make the population closer to the target.

                    In other instances, the district are non-compact because they follow city limits. You can have more compact districts by dividing the city among more districts.
                  • jimrtex4192
                    ... I m not sure why it was chosen. It may be that the SOS chose the measure for their reports. The competition software was modified to use Roeck s measure,
                    Message 9 of 19 , Sep 4, 2011
                    • 0 Attachment
                      --- In RangeVoting@yahoogroups.com, "WarrenS" <wds@...> wrote:
                      >
                      >
                      > --- In RangeVoting@yahoogroups.com, "jimrtex4192" <jimrtex@> wrote:
                      > >
                      > > One of the criteria that the Ohio redistricting contest,
                      > >
                      > > http://drawthelinemidwest.org/ohio/
                      > >
                      > > The software used by the SOS also calculate Roeck's measure of compactness, and the two sets of software **differ**
                      >
                      > --pathetic. But seems par for the course. The Roeck measure is a poor one anyhow,
                      > but even lamer if you compute it wrong.

                      I'm not sure why it was chosen. It may be that the SOS chose the measure for their reports. The competition software was modified to use Roeck's measure, which was where the error arose.

                      The software is open source, and its default measure uses a perimeter-based measure. A curiousity was a district that was based on a city with a lot of tentacle annexations. It was among the very best under Roeck's measurement because the city was somewhat round except for the tentacles which were fairly fat, but was among the worst under the perimeter based measures.

                      The previous contest in 2009 used a perimeter-based measurement, along with using the median district compactness to rate the plan. A district along the Ohio River would score poorly because of the use of the perimeter. But you could plaster over irregularities with some districts, which would then let the other districts be relatively smooth.

                      At least for legislative districts in Ohio, there really isn't a need of a compactness measure since it is extremely difficult to get non-compact districts because of the scale of the constitutionally-required building blocks: counties and townships. And it actually leads to more population deviation and bad policy choices.

                      > > The software for the redistricting contest is open source.
                      > >
                      > > https://github.com/PublicMapping/DistrictBuilder/blob/master/django/publicmapping/redistricting/calculators.py

                      The source has been updated.

                      > Simpler algorithms were later found by Emo Welzl that employ randomization.
                      > There is now public source software, called "miniball" for a version of Welzl's algorithm
                      > (implemented by Bernd Gaertner), you can download it from here:
                      > http://www.inf.ethz.ch/personal/gaertner/miniball.html

                      The comments of the new version mention both Welzl and include a subroutine called miniball, so I assume that it that algorithm. See the link above.
                    • jimrtex4192
                      ... You use a presentation that hides political boundaries and also discourages close examination in metropolitan areas. The software used in the contest lets
                      Message 10 of 19 , Sep 5, 2011
                      • 0 Attachment
                        --- In RangeVoting@yahoogroups.com, "WarrenS" <wds@...> wrote:

                        > Gerrymandering usually simply cannot happen if the district shapes are demanded to be sensible (various mathematical definitions of "sensible" work here). Therefore, if we simply demand sensible district shapes, the gerrymandering problem is done with forever.
                        > What is the cost of that? Well, sometimes the mathematical shape-quality measures, disagree with certain intuitive notions of shape-quality. If you examine the actual district maps for all 50 states produced by the shortest split line algorithm or by Olson's computer program, and compare them with the official human-drawn distrct maps, then
                        > you can see for yourself which does a better job. It is my opinion that SSA does a better job than current legislators do 98% or more of the time, and SSA defintely does it for
                        > far less monetary cost far faster. Here are the links to go to to see the pictures:
                        > http://www.rangevoting.org/Olson2000list.html
                        > http://www.rangevoting.org/SplitLR.html

                        You use a presentation that hides political boundaries and also discourages close examination in metropolitan areas.

                        The software used in the contest lets maps be overlaid on political and street maps. It is also open source. It also permits plans to be uploaded (CSV lists of census block and district numbers). Does Olson have maps based on the 2010 census data?

                        > --well... it seems to me that you are in favor of gerrymandering and against democracy.
                        > I have no doubt that lawsuits will be filed against virtually every anti-gerrymandering plan. For example that immediately happened in Florida which recently passed a (rather ineffectual-sounding) anti-gerrymandering referendum. The lawsuits failed to stop it
                        > (so far anyhow) but it remains to be seen whether this referendum will actually have a
                        > useful anti-gerrymandering effect.

                        The Florida amendments were a lawsuit on a stick change. The Florida constitution specifically requires redistricting to occur in years ending in 2. The legislature is holding hearings this summer, and those who favored the amendment are going around demanding that legislators "show their plans" even though it it is impossible to introduce them as legislation, but so they can build their legal case.

                        > --re "work against rural districts" we actually found the opposite. That is,
                        > Olson's algorithm seems to (at least with present USA parties and demographics) ENHANCE the power of rural versus city voters. However, our conclusions on that question are not very convincing at present. But they are more convincing than your (utterly unsupported) claim... since we actually did examine some maps trying to carefully count...

                        The contest software includes a political analysis.
                      • Bruce R.Gilson
                        ... One thing I dislike about your idea is that it assumes everyone drives a car. I haven t even had a driver s license since 1973 or so, and I actually never
                        Message 11 of 19 , Sep 5, 2011
                        • 0 Attachment
                          --- In RangeVoting@yahoogroups.com, Jonathan Hall <edoguiguy@...> wrote:
                          >
                          > Mr. Borda,
                          >
                          >  [...]
                          >
                          > If need be a simple modification is possible.   Rather then just adding drive times to the area centers add the squares of drive times.  This will definitely penalize tortured districts.   But I think with just the simple drive time sum, the maps will find the best way for all districts to have a core.

                          One thing I dislike about your idea is that it assumes everyone drives a car. I haven't even had a driver's license since 1973 or so, and I actually never had a driver's license and owned a car at the same time. I would get to my polling place, usually, on foot. Occasionally, I'd need to take a bus; on one occasion, I needed to take two buses because for some reason, they assigned me to a polling place that was not close enough to walk nor on the bus line that passed where I lived. If, because of, say, one-way streets, getting from A to B by car might take a long time because you need to follow a roundabout route, but walking from A to B is easy and a short distance, your method considers them as far apart. On the other hand, if the only way to get from A to B quickly is on an interstate highway, with no sidewalks to walk on and not used by a public bus line since there would not be places to stop at, your system considers them close together. I would never like to see such a rule used.
                        • WarrenS
                          So if I understand aright, Jonathan Hall s idea is, draw districts to minimize average travel time, or average squared travel time, from each person to her
                          Message 12 of 19 , Sep 5, 2011
                          • 0 Attachment
                            So if I understand aright, Jonathan Hall's idea is,
                            draw districts to minimize average travel time, or average squared travel time, from each person to her "district-center"?

                            Sounds great at first... but what exactly is "travel time" and how do you exactly measure it? Put another way, if Nixon draws district maps, and Johnson draws district maps, then how exactly does Jonathan Hall decide which map is the better map, and how does he get a consensus from everybody else that that his conclusion is correct?

                            One of the advantages of a pure mathematical definition based on geometry is, there is one clear, unique, precise, answer. A definition based on a muddy unclear hard-to-measure-exactly concept like "travel time" may be unjudicable in court, may require a comparatively enormous
                            expense to try to (approximately) measure it, and could be manipulated by, e.g. closing certain roads, changing speed limits, enforcing speed limits (or not) etc.

                            Suppose Nixon creates a road project that slows down traffic. Does that change the answer about which map is better?
                            If so, why is that a good thing?
                          • WarrenS
                            On the plus side, Hall s idea would cause a lot of effort to be devoted to measuring travel times... well, extra effort isn t a plus, but what is a plus, is
                            Message 13 of 19 , Sep 5, 2011
                            • 0 Attachment
                              On the plus side, Hall's idea would cause a lot of effort to be devoted to measuring travel times... well, extra effort isn't a plus, but what is a plus, is that then that information would be there -- which could be a useful resource for other purposes later.
                            • Jonathan Hall
                              ... draw districts to minimize average travel time, or average squared travel time, from each person to her district-center ? ... This question was answered
                              Message 14 of 19 , Sep 7, 2011
                              • 0 Attachment
                                > So if I understand aright, Jonathan Hall's idea is,
                                draw districts to minimize average travel time, or average squared travel time, from each person to her "district-center" ?

                                > Sounds great at first... but what exactly is "travel time" and how do you exactly measure it?

                                This question was answered already.


                                >  Put another way, if Nixon draws district maps, and Johnson draws district maps, then how exactly does Jonathan Hall decide which map is the better map, and how does he get a consensus from everybody else that that his conclusion is correct?

                                Also answered.  Jonathan Hall selects the plan with the overall lowest travel time.   These are not vague estimates.  These are individual tallies using individually challengable and verfiable datums.

                                One of the advantages of a pure mathematical definition based on geometry is, there is one clear, unique, precise, answer.

                                "Unique precise clear".  All criteria that right now match both my system and Gerrymandered maps as well.  I also offer the concept of community which is how political decisions are made.  People attend meetings and discuss.   But before they can attend, they travel.  To ignore community is ludicrous.  To ignore it in a way that relatively degrades rural participation and power, will call for court remedy against its assault on democracy.

                                A definition based on a muddy unclear hard-to-measure- exactly concept like "travel time" may be unjudicable in court, may require a comparatively enormous
                                expense to try to (approximately) measure it, and could be manipulated by, e.g. closing certain roads, changing speed limits, enforcing speed limits (or not) etc.

                                The speed down roads is already known to traffic engineers.  These estimates withstand court challenge already.  For those who are mystified by "travel time" the concept works like this:  Take the distances of each segment and divide by the speed.  Speed is expressed as distance over time.   The result of that division will be time.  Then add the segment results together.  If there is an arguement about how fast a road is, then the court is already deciding that issue.   And sometimes the court orders a traffic engineer to go out and measure it.  Both division and addition are well defined methods free of muddyness.  The input data for any single voter is also able to be objectively compared.   There is no piece of data that can not be isolated and directly challenged.

                                As already mentioned the travel times are taken from one of the current maping data bases.  These are already drawn "blind" to political machinations.   They are likely to have errors.   And so these can be issolated and the challenges made.   The ability to find and make politically effective challenges is an economic reality, but poor people have more time as well.  And all the time that the politically motivated changes are updating and increasing the accuracy, so to are those who are only motivated by accuracy itself.  The entropy here is toward accuracy and thus away from any political bias.

                                Suppose Nixon creates a road project that slows down traffic. Does that change the answer about which map is better?
                                If so, why is that a good thing?

                                Then Nixon is likely wasting his time as temporary projects would not be considered.  And if Nixon concocts a road project to make a community more compact, then that becomes part of the political reality.  Or if Nixon disrupts a community, then his hijinks will catch flack from all the bussiness effected.  And that interest is much higher then the normal effort directed to politics.   This particular objection is pathetic.   Roads are planned years in advance.  They are drawn on general plans which streach out for decades.  Those plans are driven by the best interests of the community and are under its control.  Bussinesses, builders and other investers depend on the plan being followed and send in armies of laywers when they are not respected.   Nixon isn't going to have a chance with trying to overide that.  Local bussiness doesn't just trump politics, it burries it alive.   The most efficient and honest political units are the
                                smaller cities.

                                Now look at the problem from Nixons POV.  He wants to group A with B so he advances a road to join them.   But A and B are only needing a pinch of closeness to join each other in his particular plan.    There are many other plans of which he is not going to have any knowledge.  His road being a fact on the ground is going to be known to the other planners, and they can notice that it also draws C closer to A.  For that example and many others he just isn't going to know how the road will 'score' on the unknown plans.  And more generally how can Nixon know that this new road will make any difference when the whole state is being recalculated with another large set of plans.   I would expect that each iteration of the method would make for increasingly closer and closer contenders.  But this all means they are ever more sacrificing their politcal goals in order to win.  And just one well crafted honest plan unburdoned by the second
                                objective is going to win.

                                Gerrymander lines are cheap to draw but roads are really expensive.   Special interests are in the habit of getting high returns on their contributions.   For forty thousand in contributions, they get four million in results.   This politicaly driven road building is going to offer miniscule returns.  And it poses a paradox, because the people that are going to have their voting rights diluted are going to see the economic benefit of the new roads.   Typically crooked politicians like to take with both hands.

                                BTW the engineering studies making speed estimates is already in the record, thus your vague charges of "muddy" and "hard-to-measure" should have at least tried to contend with that.   I would appreciate a better reading of the proposal before you waste the lists' time with an unresponsive 'response'.

                                Jonathan

                                [Non-text portions of this message have been removed]
                              • WarrenS
                                ... --actually, travel time at 3am is different from travel time at 2pm, is different from travel time on election day. I claim no traffic engineer in the
                                Message 15 of 19 , Sep 8, 2011
                                • 0 Attachment
                                  > >  Put another way, if Nixon draws district maps, and Johnson draws district maps, then how exactly does Jonathan Hall decide which map is the better map, and how does he get a consensus from everybody else that that his conclusion is correct?
                                  >
                                  > Also answered.  Jonathan Hall selects the plan with the overall lowest travel time.   These are not vague estimates.  These are individual tallies using individually challengable and verfiable datums.
                                  >
                                  > One of the advantages of a pure mathematical definition based on geometry is, there is one clear, unique, precise, answer.
                                  >
                                  > "Unique precise clear".  All criteria that right now match both my system and Gerrymandered maps as well.  I also offer the concept of community which is how political decisions are made.  People attend meetings and discuss.   But before they can attend, they travel.  To ignore community is ludicrous.  To ignore it in a way that relatively degrades rural participation and power, will call for court remedy against its assault on democracy.
                                  >
                                  > A definition based on a muddy unclear hard-to-measure- exactly concept like "travel time" may be unjudicable in court, may require a comparatively enormous
                                  > expense to try to (approximately) measure it, and could be manipulated by, e.g. closing certain roads, changing speed limits, enforcing speed limits (or not) etc.
                                  >
                                  > The speed down roads is already known to traffic engineers.  These estimates withstand court challenge already.  For those who are mystified by "travel time" the concept works like this:  Take the distances of each segment and divide by the speed.  Speed is expressed as distance over time.   The result of that division will be time.  Then add the segment results together.  If there is an arguement about how fast a road is, then the court is already deciding that issue.   And sometimes the court orders a traffic engineer to go out and measure it.  Both division and addition are well defined methods free of muddyness.  The input data for any single voter is also able to be objectively compared.   There is no piece of data that can not be isolated and directly challenged.

                                  --actually, travel time at 3am is different from travel time at 2pm, is different from travel time on election day.
                                  I claim no traffic engineer in the world knows what all those differences are, for all roads.

                                  > As already mentioned the travel times are taken from one of the current maping data bases.  These are already drawn "blind" to political machinations.   They are likely to have errors.   And so these can be issolated and the challenges made.

                                  --costly. New bureacracy needed for no real reason.

                                  > Suppose Nixon creates a road project that slows down traffic. Does that change the answer about which map is better?
                                  > If so, why is that a good thing?
                                  >
                                  > Then Nixon is likely wasting his time as temporary projects would not be considered.  And if Nixon concocts a road project to make a community more compact, then that becomes part of the political reality.  Or if Nixon disrupts a community, then his hijinks will catch flack from all the bussiness effected.  And that interest is much higher then the normal effort directed to politics. 

                                  --well I disagree. Here's the reality: politicians are willing to subvert the entire economy of an entire country for decades, destroy entire countries, kill 100s of thousands, just to get more votes from specific swing demographics or to gain other perks. For example, the corn-ethanol subsidy program. The Iraq war. The way G.W.Bush made his money, which was to get his pol-pals to declare an eminent domain land grab and then use the land to build a Texas Rangers baseball stadium, then give it to the Texas Rangers (main owner: Bush), causing him to make a fortune. The Tweed courthouse. The Tweed ring indeed diverted approximately 1/3 of the entire finances of NY city into their own pockets during the time they were in power. Enormous gerrymandering for 100s of years despite
                                  its immense unpopularity. Redefinition of "budget deficit" subverting
                                  all of economics and the entire country just so Bush could make false claims (ditto Clinton) about the USA's budget.
                                  You somehow think that the task of building (or closing) a road, is going to be in comparison to that, too enormous to be subject to political manipulation. Oh.

                                  I, in contrast, think that if a politician thinks it'd suit his purposes, he'd be willing to destroy all the roads in a country, or
                                  build useless unused roads. In fact, latter recently happened
                                  (or tried) with recent high-speed rail project for Florida.

                                  >  This particular objection is pathetic.   Roads are planned years in advance.  They are drawn on general plans which streach out for decades.  Those plans are driven by the best interests of the community and are under its control. 

                                  --no, your ignorance of history and childlike innocence, combined with
                                  arrogance, is pathetic. I repeat. Politicians are perfectly happy
                                  to distort and even destroy, entire countries, for decades, if it
                                  gets them political points. If you make our roads be a pawn in
                                  a political game, then you could be making a huge mistake, because roads actually are important. Similarly in the recent Florida
                                  antigerrymandering referendum case, I worry they've just made
                                  judges be a pawn in this game (since it likely will now all be about lawsuits) which would be bad since judges actually are important
                                  and this could cause immensely more incentive to appoint corrupt judges.

                                  Here's what politicians cannot manipulate/change: actual geography, the overall geometry of the planet, and where people live. (Actually they could change where people live and force huge migrations,
                                  Stalin and Mao did, but that'd almost always be highly unpopular
                                  in the view of almost everybody involved, and certainly was in their cases.)

                                  > Now look at the problem from Nixons POV.  He wants to group A with B so he advances a road to join them.   But A and B are only needing a pinch of closeness to join each other in his particular plan.    There are many other plans of which he is not going to have any knowledge.  His road being a fact on the ground is going to be known to the other planners, and they can notice that it also draws C closer to A.  For that example and many others he just isn't going to know how the road will 'score' on the unknown plans.  And more generally how can Nixon know that this new road will make any difference when the whole state is being recalculated with another large set of plans.   I would expect that each iteration of the method would make for increasingly closer and closer contenders.  But this all means they are ever more sacrificing their politcal goals in order to win.  And just one well crafted honest plan unburdoned by the second
                                  > objective is going to win.

                                  --huh? The politicians will have computer models of how every little change in roads will affect their votes. They'll spend hundreds of millions of dollars to get advice from those models. They'll micromanage road building from here on out if your ideas were adopted.

                                  Think I'm making it up? Wrong: they already spend that amount of money on gerrymandering consultants. Including even their OWN money,
                                  and that's saying something.

                                  > Gerrymander lines are cheap to draw but roads are really expensive.  

                                  --good reason to not base it on roads. Thank you for making
                                  my argument for me.

                                  Politicians don't care how much it costs, it is not their money. That is the problem, not the solution. For example,
                                  my state senator just sent me, at taxpayer (i.e. my) expense,
                                  a pamphlet containing no actual information about what's going on in Albany, but lots of expensive pictures of the world trade center and lots of hifalutin baloney about how we needed to remember 9/11 (upcoming anniversary) and how great America and God is. Did the expense of this move play the slightest role in his consideration of whether to do it?
                                  It probably cost about $500,000 in taxpayer money for him to do that one "informational" printing+mailing and it accomplished absolutely nothing objectively useful. It worked against him in the sense I believe he's an anus now, but I already knew he was.

                                  >Special interests are in the habit of getting high returns on their contributions.   For forty thousand in contributions, they get four million in results.   This politicaly driven road building is going to offer miniscule returns. 

                                  --you don't get it. The special interests will not pay for it. YOU will pay for it. The returns for the special interests will be: huge,
                                  they get everything they want; the costs for the special interests will be: zero.

                                  > BTW the engineering studies making speed estimates is already in the record, thus your vague charges of "muddy" and "hard-to-measure" should have at least tried to contend with that.   I would appreciate a better reading of the proposal before you waste the lists' time with an unresponsive 'response'.

                                  --there are a zillion variables affecting my travel time from A to B.
                                  Including my personal whims that day. It's not a precisely predictable quantity. 100s of predictions could be made, and the one will be chosen that suits politicians' purposes. In fact, if you look
                                  at virtually any engineering project being debated, there will be
                                  many estimates made which are wildly conflicting, and politicians choosing among them and ignoring others for their personal convenience.

                                  And if you think the government will solve this problem by declaring there be one, set in stone, uncorruptible standard way to measure "travel time" and road manipulation will be illegal...
                                  well, I have a bridge to sell you. Literally. In things like
                                  this, it is all about the details, and chnages in the details are made in smoke filled rooms at midnight and the media never covers them.
                                • jimrtex4192
                                  ... How do we do that when we are electing someone? (Gary?) Johnson presents his platform, and (Jay?) Nixon presents his platform and each voter determines
                                  Message 16 of 19 , Sep 8, 2011
                                  • 0 Attachment
                                    --- In RangeVoting@yahoogroups.com, "WarrenS" <wds@...> wrote:

                                    > Sounds great at first... but what exactly is "travel time" and how do you exactly measure it? Put another way, if Nixon draws district maps, and Johnson draws district maps, then how exactly does Jonathan Hall decide which map is the better map, and how does he get a consensus from everybody else that that his conclusion is correct?

                                    How do we do that when we are electing someone?

                                    (Gary?) Johnson presents his platform, and (Jay?) Nixon presents his platform and each voter determines which has more utility for them.

                                    So have a set of candidate district plans. Determine the intersection of the plans. This produces smaller areas that have a common set of proposed districts that the area may be placed in. So your house might be in District A1, District B1, District C27, etc. But the set of districts all contain your residence, and some of your neighbors (the area that shares the set of districts need not be contiguous, but you are at least close enough to some of these other persons that someone thought you should be in the same district).

                                    Rank the districts. The ranking of your districts determines the ranking of the plans. It is none of your business what another district on the other side of the city or state looks like, any more than it is who those people choose to represent them.

                                    Determine the winning plan(s) using Condorcet. We want a consensus of people who generally like a plan.

                                    Take the winning plan(s) and determine whether the people in each of its districts generally like the plan. A plan might be liked most of the areas of the state, but loathed by certain areas. We don't need to force these areas into a consensus like we might if we were picking a single winning candidate.

                                    Choose the plan which has the most districts whose voters are content with. These districts are locked into the final plan.

                                    Repeat the process for the remaining areas.
                                    >
                                    > One of the advantages of a pure mathematical definition based on geometry is, there is one clear, unique, precise, answer. A definition based on a muddy unclear hard-to-measure-exactly concept like "travel time" may be unjudicable in court, may require a comparatively enormous

                                    Any mathematical definition is based on a subjective measure of utility, including that a mathematical definition has utility.
                                  Your message has been successfully submitted and would be delivered to recipients shortly.