Chops and Erosity - Great Lakes Style
       |           

Welcome, Guest. Please login or register.
Did you miss your activation email?
March 29, 2024, 04:05:49 AM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  General Politics
  Political Geography & Demographics (Moderators: muon2, 15 Down, 35 To Go)
  Chops and Erosity - Great Lakes Style
« previous next »
Pages: 1 ... 8 9 10 11 12 [13] 14
Author Topic: Chops and Erosity - Great Lakes Style  (Read 24711 times)
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #300 on: February 23, 2015, 02:10:21 PM »
« edited: February 23, 2015, 02:14:20 PM by Torie »

While everyone is dissing my Livingston chop map, understand that it is just an entry. If it stays on the Pareto frontier it goes on to a committee of real humans who decide things like whether Saginaw should stay with Midland or if Kalamazoo to Monroe is a bridge too far. They can determine if the South Lyon/Brighton/Howell urbanized area, which is separate from the Detroit urbanized area, means that Livingston can be split off as much as other CSA counties like Monroe and Washtenaw that happen to have an older center that qualified them for their own MSA.

It can stay on the Pareto frontier provided it survives after getting the penalty points it deserves.  Sure, it may be rather arbitrary how multi county UCC's are defined, and what's included where, if at all (e.g., the Bay and Saginaw Counties not being a multi county UCC), but you know what? The perfect is often the enemy of the good, and at least it provides some objective playing field, decided by "weirdos" in the census bureau, and Jimtex, with his urbanization and population cut off metrics, not in a political mode, so in this sense the rules are not gamed in advance - they are just there.
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #301 on: February 23, 2015, 02:21:22 PM »

The way to solve that issue to make everyone happy is to use your metric, that no more CD's are in the UCC than the nearest whole number rounded up, but add that 1)  the number of whole CD's entirely within the UCC must be the nearest whole number to the quota rounded down (thus Dayton, Columbus and Grand Rapids, must have one CD wholly within them, and Cincy must have two, Cleveland 2, etc), and 2) where there is an excess over the minimum using either metric (I have that excess in the Cleveland UCC, because the Akron CD butts in, while OH-14 butts out, so I have 4 CD's in the UCC, while anything over 3 CD's in the UCC, with a least 2 CD's entirely in the UCC, counts as an excess), the number of people involved in the excess that are equal to a macrochop and over, are penalized on a per macrochop incremental basis (my Akron incursion is below a macrochop, so no problem but the chop itself).

My preferred alternative would be similar, but simpler (and more in keeping with "a chop is a chop" mentality):

Take the quota for the UCC.  One chop point for each district that butts in when it doesn't need to; also one chop point for each district that fans out when it doesn't need to.
Logged
muon2
Moderator
Atlas Icon
*****
Posts: 16,788


Show only this user's posts in this thread
« Reply #302 on: February 23, 2015, 02:29:03 PM »

So is this what we can agree to?

The UCC size is defined as the population of the UCC divided by the quota and rounded up to the nearest whole number. The cover of the UCC is the number of districts that include any or all of the UCC. The pack of the UCC is the number of districts that are wholly contained by the UCC. Chop points are assessed for the difference between the cover and size of the UCC, and for the difference between one less than the size and the pack.

I assume that this is in addition to the normal county chop score. If so, I can live with this. I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible. It would certainly be a significant departure from the hashed out definition of a county chop two years ago.

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue

Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #303 on: February 23, 2015, 02:35:27 PM »
« Edited: February 23, 2015, 02:51:48 PM by Torie »

The way to solve that issue to make everyone happy is to use your metric, that no more CD's are in the UCC than the nearest whole number rounded up, but add that 1)  the number of whole CD's entirely within the UCC must be the nearest whole number to the quota rounded down (thus Dayton, Columbus and Grand Rapids, must have one CD wholly within them, and Cincy must have two, Cleveland 2, etc), and 2) where there is an excess over the minimum using either metric (I have that excess in the Cleveland UCC, because the Akron CD butts in, while OH-14 butts out, so I have 4 CD's in the UCC, while anything over 3 CD's in the UCC, with a least 2 CD's entirely in the UCC, counts as an excess), the number of people involved in the excess that are equal to a macrochop and over, are penalized on a per macrochop incremental basis (my Akron incursion is below a macrochop, so no problem but the chop itself).

My preferred alternative would be similar, but simpler (and more in keeping with "a chop is a chop" mentality):

Take the quota for the UCC.  One chop point for each district that butts in when it doesn't need to; also one chop point for each district that fans out when it doesn't need to.

You think a microchop is the same as a chop involving 300,000 people. I don't. Beyond the size of the chop issue, here it is a matter of policy, not complexity. The round down as well as up metric is simple.  It penalizes fan outs of up to 300,000 or so people without penalty. Without the round down concept, you can just bifurcate a UCC, as you did more or less with the GR UCC, because one CD was not wholly contained within it. The Pubs used to do that with Rochester, NY for decades, before the area got too Dem to make that safely work for them (and ditto with Columbus), and that could have been done back them, while generating really high scoring maps per the regime here from which I strongly dissent. Pretend that the Johnson-Wyandotte UCC in Kansas had 12,000 more people, plus a macrochop increment (it will probably be there in the next census), and thus was over one CD in population plus a macrochop increment. That UCC is getting way too Dem these days. Time to bisect it - for free! Just chop off Wyandotte as the Pubs do now. Tongue  Don't you want a system that just basically says no to that?
Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #304 on: February 23, 2015, 03:19:05 PM »
« Edited: February 23, 2015, 03:59:05 PM by Torie »

So is this what we can agree to?

The UCC size is defined as the population of the UCC divided by the quota and rounded up to the nearest whole number. The cover of the UCC is the number of districts that include any or all of the UCC. The pack of the UCC is the number of districts that are wholly contained by the UCC. Chop points are assessed for the difference between the cover and size of the UCC, and for the difference between one less than the size and the pack.

I assume that this is in addition to the normal county chop score. If so, I can live with this. I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible. It would certainly be a significant departure from the hashed out definition of a county chop two years ago.

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue



Except for the below, I read that as an acceptance of my system, absent the macrochop increment penalty. That still treats fan outs the same, whether the population involved is 40,000 or 300,000. I am not giving up on this one ever, other than via being persuaded it causes map deterioration, or some other negative fallout, that makes such a sliding scale inappropriate, and/or that my scale is too punitive, and the penalty scale needs to be more relaxed. No such evidence of a tangible nature has so far been adduced. Folks just reiterate their bottom line opinions.

I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible.

Really (even with the up to a macrochop pad)? Why? Beyond that, that sounds like discrimination to me. Why should multi county UCC's be treated as sacred cows, beyond the aggregation concept?

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue

You obviously have ceased looking at my maps. Tongue  The problem is, is that this maps generates two erosity points in exchange for avoiding a fan out of GR. If one gets only one more chop point for the fan out, the penalty in this case is pretty toothless. It pays to fan out with but a one point penalty. Which gets back to, well you guess it, incremental penalty points! A fan out per our previous maps deserves more than one penalty point. In my system, it would get 2 or 3 penalty points (2 in Jimtex's map, and 3 in mine).



 
Logged
jimrtex
Atlas Icon
*****
Posts: 11,828
Marshall Islands


Show only this user's posts in this thread
« Reply #305 on: February 23, 2015, 03:20:49 PM »
« Edited: February 23, 2015, 08:27:50 PM by jimrtex »

I understand this, and I think it makes a good basis for an algorithm. However, what it can't do is evaluate a plan like the one I posted for MI. To do so, it has to reverse engineer regions where none are plainly evident. It is possible to propose regions consistent with the plan, but that choice may not be unique. We take some care in our maps, but what about citizens who propose maps that aren't so careful to minimize chops. I wouldn't want to see them discarded, but I would like to see them scored.

The second issue I have is that it presumes the sanctity of the UCC. I see no reason why excess UCC splits should be viewed as anything more special than excess county splits. In my posted plan, I actually did use regions to devise it. It has eight regions, which is the same number as your graph, but my regions don't respect the UCCs. Shouldn't I get scored based on my regions with perhaps a penalty for violating the UCCs? If you try to score it based on UCCs it would get a prohibitively large penalty for population shifts that you might as well say that UCCs are inviolable. That was never the intent of the UCCs as I understood their development.



Region 1 (CD 1) 1.0000
Region 2 (CD 2+4) 1.9950
Region 3 (CD 3) 1.0035
Region 4 (CD 6) 1.0025
Region 5 (CD 7) 0.9920
Region 6 (CD 8 ) 1.0042
Region 7 (CD 5+9+11) 3.0019
Region 8 (CD 10+12+13+14) 4.0010

Shifts 8->5 (0.10%), 7->5 (0.19%), 6->5 (0.42%), 4->5 (0.10%): 0.81%
Shifts 3->2 (0.35%), 4->2 (0.15%): 0.50%
Total shift 1.30% (adjusted for rounding). The red is the only shift that was required on the map.

Why shouldn't this be scored as such, with an additional penalty for the 3 UCC chops?

Alternatively I could claim CD 5 as a 9th region since it is within 5% of the quota. How should that be treated?
Your regions do not comply with the requirement that UCCs be contained within a region.  Your combination of CD 2 and 4 into a region hides that you need a chop in Muskegon County.  The purpose of my analysis is to identify the population that must be shifted across county lines (other than within UCC).

Sanctity is not a word choice that I would make.  Because the Detroit UCC has a population close to 6 districts, there is little flexibility.  It is quite reasonable to keep 6 CDs in the core area of the metropolitan area.  On the other hand there is all kinds of flexibility with the Grand Rapids UCC, and a reasonable amount with the Lansing UCC.

In 2020, Michigan will have 13 districts and the Detroit UCC will have a population of around 5.5 quotas.  We will be able to extend out into large counties such as Washtenaw or Genesee, or smaller counties.  We will have options of 5 whole, and another 1/2 in 1/2 out, or perhaps 4 whole, and 2 3/4 in, 1/4 out.

There are over 200,000 persons who cross the Wayne-Oakland border every day (which incidentally has the same relationship as DuPage-Cook and Orange-Los Angeles, slightly greater outflow than inflow).  If we consider persons living in the households of these cross-border workers, that is more than 1/2 a CD that crosses the border between your regions every day.

The concept of UCC originated in Texas, where they would have been defined in the Houston, DFW, and I-35 corridor between San Antonio and Waco.  There might also have been smaller ones in the McAllen-Brownsville and El Paso areas.  The goal would be keep districts tight to these areas, permitting the more rural districts to be created in between.

Stringing areas together makes it harder more expensive to campaign and represent, and make it less likely that the representative is representative of the area.

There are close ties between the counties within a UCC, such as the commuting ties - which after all was what placed the counties in the same metropolitan areas to begin with.

======================
Clinton County residents work in:
Ingham 47.6%
Clinton 31.2%
Eaton 10.2%
Shiawassee 2.4%
Ionia 1.7%
Gratiot 1.1%
89.4% in UCC
----------------------------------------
Eaton County residents work in:
Eaton 45.2%
Ingham 40.4%
Calhoun 3.3%
Clinton 3.0%
Kent 1.3%
88.6% in UCC
----------------------------------------
Ingham County residents work in:
Ingham 80.5%
Eaton 8,2%
Clinton 1.9%
Jackson 1.8%
Livingston 1.4%
Washtenaw 1.1%
Oakland 1.0%
90.6% in UCC
======================
Kent County residents work in:
Kent 89.4%
Ottawa 4.9%
94.3% in UCC
----------------------------------------
Ottawa County residents work in:
Ottawa 61.8%
Kent 25.6%
Allegan 6.1%
Muskegon 4.1%
87.4% in UCC
======================
Livingston County residents work in:
Livingston 42.9%
Oakland 19.8%
Washtenaw 13.7%
Wayne 12.6%
Genesee 3,8%
Ingham 2.6%
Macomb 1.3%
76.5% in UCC
----------------------------------------
Macomb County residents work in
Macomb 58.4%
Oakland 23.9%
Wayne 15.0%
98.3% in UCC
-------------------------------------------
Oakland County residents work in
Oakland 69.8%
Wayne 18.2%
Macomb 6.7%
Wahtenaw 1.3%
Genesee 1.1%
95.5% in UCC
-------------------------------------------
St. Clair County residents work in
St. Clair 63.7%
Macomb 27.7%
Oakland 5.2%
Wayne 4.6%
Lapeer 1.4%
96.2% in UCC
-----------------------------------------
Wayne County residents work in
Wayne 74.2%
Oakland 14.8%
Macomb 4.9%
Washtenaw 3/8%
94.2% in UCC

Other counties with 20% plus to another county:

To Grand Traverse: Benzie 39%, Kalkaskia 32%, Leelanau 41% (Traverse City is largest city in Northern Michigan)

To Kent: Barry 25%, Ionia 25%, Montcalm 26%, Newaygo 23% (rural counties to Grand Rapids)

To Iosco: Alcona 24% (Iosco is tiny population, Au Sable is just south of border)

To Ottawa: Allegan 20% (Ottawa is not a pure satellite of Grand Rapids, but gets placed in the Grand Rapids UCC because Grand Rapids is toward the western border of Kent).

To Saginaw: Bay 22% (Saginaw-Bay is close to becoming a multi-county UCC).

To Elkhart, IN: Cass 21%

To Houghton: Keweenaw 23%

To Marinette, WI: Menominee 36%

To Wexford: Missaukee 40.3% (Cadillac is near Missaukee border)

To Oakland: Lapeer 23%

To Muskegon: Oceana 21%

To Kalamazoo: Van Buren 26%

In-between counties may be kept out of metropolitan areas because jobs are in both direction.  For example commuting from Monroe to Lucas is almost as strong as from Monroe to Wayne,  In the past Monroe has been in the Toledo metropolitan area.  Shiawassee has 18% to Ingham vs. 12% to Genesee helping to keep it out of either metropolitan area.
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #306 on: February 23, 2015, 03:25:15 PM »

So is this what we can agree to?

The UCC size is defined as the population of the UCC divided by the quota and rounded up to the nearest whole number. The cover of the UCC is the number of districts that include any or all of the UCC. The pack of the UCC is the number of districts that are wholly contained by the UCC. Chop points are assessed for the difference between the cover and size of the UCC, and for the difference between one less than the size and the pack.

I assume that this is in addition to the normal county chop score. If so, I can live with this. I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible. It would certainly be a significant departure from the hashed out definition of a county chop two years ago.

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue

Looks beautiful.

Torie's most recent map does in fact avoid the Grand Rapids penalty.  That might be enough to make it strictly optimal here (modulo the Detroit lines, and in particular whether 47% BVAP is good enough or not).
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #307 on: February 23, 2015, 03:43:32 PM »

So is this what we can agree to?

The UCC size is defined as the population of the UCC divided by the quota and rounded up to the nearest whole number. The cover of the UCC is the number of districts that include any or all of the UCC. The pack of the UCC is the number of districts that are wholly contained by the UCC. Chop points are assessed for the difference between the cover and size of the UCC, and for the difference between one less than the size and the pack.

I assume that this is in addition to the normal county chop score. If so, I can live with this. I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible. It would certainly be a significant departure from the hashed out definition of a county chop two years ago.

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue



Except for the below, I read that as an acceptance of my system, absent the macrochop increment penalty. That still treats fan outs the same, whether the population involved is 40,000 or 300,000. I am not giving up on this one ever, other than via being persuaded it causes map deterioration, or some other negative fallout, that makes such a sliding scale inappropriate, and/or that my scale is too punitive, and the penalty scale needs to be more relaxed. No such evidence of a tangible nature has so far been adduced. Folks just reiterate their bottom line opinions.

I will point you to my drawing of the Pittsburgh area upthread for why I would find the macrochop increment penalty to be too punitive.  Presumably my map would get three macrochop penalty points for District 12 taking Armstrong and Greene, despite the fact that Greene is trapped behind the Pittsburgh UCC, and Armstrong is part of the metro area as well.  Forcing one to make the swing district Greene-Fayette-whatever inevitably leads to uglier lines all around the western half of the state, for no real defensible reason.

It also fails at basic apples-to-apples accounting, if only this one type of chop can accrue oodles and oodles of penalty points.  And, if you want to say that all macrochops are subject to this sliding scale penalty, then yes the danger of degenerate outcomes, where extra tiny chops are seen as "better" than one big macrochop, will rear its head.
Logged
muon2
Moderator
Atlas Icon
*****
Posts: 16,788


Show only this user's posts in this thread
« Reply #308 on: February 23, 2015, 04:00:35 PM »

So is this what we can agree to?

The UCC size is defined as the population of the UCC divided by the quota and rounded up to the nearest whole number. The cover of the UCC is the number of districts that include any or all of the UCC. The pack of the UCC is the number of districts that are wholly contained by the UCC. Chop points are assessed for the difference between the cover and size of the UCC, and for the difference between one less than the size and the pack.

I assume that this is in addition to the normal county chop score. If so, I can live with this. I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible. It would certainly be a significant departure from the hashed out definition of a county chop two years ago.

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue



Except for the below, I read that as an acceptance of my system, absent the macrochop increment penalty. That still treats fan outs the same, whether the population involved is 40,000 or 300,000. I am not giving up on this one ever, other than via being persuaded it causes map deterioration, or some other negative fallout, that makes such a sliding scale inappropriate, and/or that my scale is too punitive, and the penalty scale needs to be more relaxed. No such evidence of a tangible nature has so far been adduced. Folks just reiterate their bottom line opinions.

I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible.

Really (even with the up to a macrochop pad)? Why? Beyond that, that sounds like discrimination to me. Why should multi county UCC's be treated as sacred cows, beyond the aggregation concept?

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue

You obviously have ceased looking at my maps. Tongue  The problem is, is that this maps generates two erosity points in exchange for avoiding a fan out of GR. If one gets only one more chop point for the fan out, the penalty in this case is pretty toothless. Which gets back to, well you guess it, incremental penalty points! A fan out per our previous maps deserves more than one penalty point. In my system,



 

My bad for missing the map, I obviously fell behind in the scoring of plans as I addressed proposed changes. Too many subthreads embedded in this one. Tongue

I guess I really need to see your sliding scale in a table, similar to the one I put together for inequality. Then I'd like clear examples that show the problem we are addressing. I start from the position that the basic definition of chops, specifically county chops, were settled two years ago. If we are reopening that here, the onus is on the party seeking change to demonstrate why the proposed change is needed. You put me through no less to justify macrochops to deal with urban erosity, and I came up with a series of Kent county maps to make my case.

I said I could live with the cover/pack system for UCCs. I don't find it ideal, because then I am back to thinking of them as overlays and not just some kind of super-county. However, if the chop out problem is that severe (and it only seems a mild problem to me), your fix as described by me is a simple enough solution.

I would also caution thinking about direct trades of chops for erosity as you might have above. They are on different scales at present, with a likelihood of 1 chop equal to about 4 erosity in MI, but I lack sufficient data to assert that.
Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #309 on: February 23, 2015, 04:08:00 PM »
« Edited: February 23, 2015, 05:08:43 PM by Torie »

So is this what we can agree to?

The UCC size is defined as the population of the UCC divided by the quota and rounded up to the nearest whole number. The cover of the UCC is the number of districts that include any or all of the UCC. The pack of the UCC is the number of districts that are wholly contained by the UCC. Chop points are assessed for the difference between the cover and size of the UCC, and for the difference between one less than the size and the pack.

I assume that this is in addition to the normal county chop score. If so, I can live with this. I'm not sure I would sign off on this for counties, because I'm not sold on the notion that large counties should prefer having as many whole districts in them as possible. It would certainly be a significant departure from the hashed out definition of a county chop two years ago.

If we go in this direction for UCCs I think we will all head back to our software, since I'm not aware of any of the maps that avoid a penalty for GR. Tongue



Except for the below, I read that as an acceptance of my system, absent the macrochop increment penalty. That still treats fan outs the same, whether the population involved is 40,000 or 300,000. I am not giving up on this one ever, other than via being persuaded it causes map deterioration, or some other negative fallout, that makes such a sliding scale inappropriate, and/or that my scale is too punitive, and the penalty scale needs to be more relaxed. No such evidence of a tangible nature has so far been adduced. Folks just reiterate their bottom line opinions.

I will point you to my drawing of the Pittsburgh area upthread for why I would find the macrochop increment penalty to be too punitive.  Presumably my map would get three macrochop penalty points for District 12 taking Armstrong and Greene, despite the fact that Greene is trapped behind the Pittsburgh UCC, and Armstrong is part of the metro area as well.  Forcing one to make the swing district Greene-Fayette-whatever inevitably leads to uglier lines all around the western half of the state, for no real defensible reason.

It also fails at basic apples-to-apples accounting, if only this one type of chop can accrue oodles and oodles of penalty points.  And, if you want to say that all macrochops are subject to this sliding scale penalty, then yes the danger of degenerate outcomes, where extra tiny chops are seen as "better" than one big macrochop, will rear its head.

Bear in mind, that an ugly map earns its own penalty points (erosity points, but also potentially chop points elsewhere), and if a county is trapped, then everyone incurs fanout points one way or the other, so it is irrelevant. It is only penalty points that can be avoided that are relevant. The minimum chop size into the GR UCC is 172K, and more penalty points only happen if you go past an additional macrochop, or about 208K.  With a one point penalty only, it always pays to chop off Ottawa - at least vis a vis erosity points. So we need to come up with maps, that show that with this scale, the maps with the highest scoring will just suck overall. I tend to doubt that will ever be proved. Greene doesn't have many people in it anyway, so it is not going to be a particularly big deal. Ottawa does, so it is a big deal. So is Livingston.
Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #310 on: February 23, 2015, 04:20:07 PM »
« Edited: February 23, 2015, 05:00:37 PM by Torie »

Mike above I posted that you come up with the number of erosity points versus chops that obtains for a state, and that gives you your ratio. You are not reading my posts, and unlike yours sometimes, I think mine are comprehensible. Smiley

This is not a litigation, about who came up with what when, and how long something has been in place. Nobody really understood your system anyway - until now. It was basically incomprehensible, or the implications just were not understood. Just pretend your system is being attacked in the public square. Any new system has the burden of proof.

Anyway, it turns out both the Jimtex and Torie maps earn 3 penalty points for the GR fanout. It costs two erosity points to avoid the fan out in my map. That seems like about right. Yes, if a chop point is deemed worth about 3 erosity points in Michigan, maybe that is excessive. It all depends on the ratio. Come to think about it, the ratio is really about the chop points in excess of the minimum, versus the erosity points in excess of the minimum, for the state as a whole, and comparing the ratio of the excesses. I have no real reason to believe that for good map making, the ratio should not be other than 1 to 1, but I have an open mind on that. The excess is the thing, and one should never forget that principle. It's darn important. For Michigan for example, it might be useful to draw a map with the absolutely minimum number of chops, using your system as modified, and mine, where we still disagree, and see how many erosity points it has versus say my map, and see what that ratio is. I suspect one of your maps above, effects that absolute minimum. More to the point in a strictly pareto optimality regime, one would look to see how many more chops are generated elsewhere from your "perfect" map from a chop standpoint, by avoiding the GR fanout, and see if adding 3 points for it (if you don't chop Kent at all, but just sever Ottawa, you can get it down to two points), generates enough chop points elsewhere, that it seems the GR fanout issue has been given undue weight. I don't think so.  Avoiding the GR fanout generates more erosity points than is typical I would think, because it's in a "bad" spot. Often the additional erosity points will be but one. So the issue really is about how much "worse" maps are made elsewhere from a chop standpoint, by virtue of levying x versus y chop points for excessive fanouts.



Logged
muon2
Moderator
Atlas Icon
*****
Posts: 16,788


Show only this user's posts in this thread
« Reply #311 on: February 23, 2015, 05:08:15 PM »

Mike above I posted that you come up with the number of erosity points versus chops that obtains for a state, and that gives you your ratio. You are not reading my posts, and unlike yours sometimes, I think mine are comprehensible. Smiley

This is not a litigation, about who came up with what when, and how long something has been in place. Nobody really understood your system anyway - until now. It was basically incomprehensible, or the implications just were not understood. Just pretend your system is being attacked in the public square. Any new system has the burden of proof.

I did read your post, and I don't think that your erosity versus chops will be so simple when actual data is analyzed, and I've said why in replies to others on this thread. I'd love to be surprised and find that it is that simple.

Actually I fervently disagree about the latter. I am no mood to redebate the issue from two years ago. No one had any misunderstanding about how to count county chops at that time. You were even so kind as to mark up a first draft of diagrams that I had tried to draw to illustrate the different ways to envision chops. After so many maps and so many posts back then, now you want to change.

I will happily take my views to the real public square, and I am confident that a view of chops based on how many pieces of a district are in a county regardless of nesting will prevail. I believe it already does in some states such as NJ that counts towns that are served by more than one rep in their reports. The splits reports for districts from the CA commission conforms to my definition of chop pieces for both cities and counties.

I regret that I clouded the original chop definition by trying to incorporate UCCs into the same framework. They are different and will seem so to the public as is apparent from the discussion in this thread. But if my definition of county chops is unacceptable, I will leave it to someone else to score plans.
Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #312 on: February 23, 2015, 05:26:42 PM »
« Edited: February 23, 2015, 05:41:14 PM by Torie »

I am not sure what the relevance of "nesting" is to this discussion (I think of rep districts being nested in senate districts, but if by nesting you mean aggregation of numbers, and nobody does that, that certainly does not surprise me). Anyway, if you dump the UCC concept, which is where this contretemps is centered, I am not sure what you are saying. Your system has been accepted for NJ and CA that micros are free, and macros are something special, or what?  Without the UCC concept, or the macrochop concept (you don't need to aggregate without it), and given you accepted that micros should have some penalty, the only thing we disagreed about was the scaling. I think your erosity system is great (I am sure that has been incorporated absolutely nowhere but it should), subject to my zoom suggestion modification. We were not discussing scoring for UCC's two years ago were we?  Anyway, I would rather have no UCC system at all, rather than one that can be gamed. More to the point, this is not about satisfying me or Train, or Jimtex, or anyone else, or playing the game of who has the highest scoring maps, but coming up with the "best" system. If you are confident enough about where you are going, that you don't think it worthwhile to spend time exploring new concepts, or explaining why, having thought of that idea ages ago, you decided to reject them, unless somebody else does all the work, I certainly understand. It's a cost-benefit analysis.

Anyway, sorry this has become so contentious (I certainly bear at least part of the blame, and apologize for that). Have a good evening.
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #313 on: February 23, 2015, 05:45:21 PM »

Come to think about it, the ratio is really about the chop points in excess of the minimum, versus the erosity points in excess of the minimum, for the state as a whole, and comparing the ratio of the excesses. I have no real reason to believe that for good map making, the ratio should not be other than 1 to 1, but I have an open mind on that. The excess is the thing, and one should never forget that principle. It's darn important.

Given that the CHOP scores as currently calculated by Mike are in the mid-teens, and the EROSITY scores are in the 90-110 range... that strikes me as pretty clear evidence that a chop point is worth several erosity points, at least.  Given the way the scores are currently calculated, they are pretty far from 1 to 1, and even if your sliding scale were to be implemented across the board I'm quite confident they would remain so.  It's just not apples to apples at all, not without a serious ratio adjustment.
Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #314 on: February 23, 2015, 06:03:15 PM »
« Edited: February 23, 2015, 06:05:56 PM by Torie »

Come to think about it, the ratio is really about the chop points in excess of the minimum, versus the erosity points in excess of the minimum, for the state as a whole, and comparing the ratio of the excesses. I have no real reason to believe that for good map making, the ratio should not be other than 1 to 1, but I have an open mind on that. The excess is the thing, and one should never forget that principle. It's darn important.

Given that the CHOP scores as currently calculated by Mike are in the mid-teens, and the EROSITY scores are in the 90-110 range... that strikes me as pretty clear evidence that a chop point is worth several erosity points, at least.  Given the way the scores are currently calculated, they are pretty far from 1 to 1, and even if your sliding scale were to be implemented across the board I'm quite confident they would remain so.  It's just not apples to apples at all, not without a serious ratio adjustment.

Yes, but the minimum chops appear to be about 15 or something, and the minimum erosity score is probably near 90. So it becomes an issue of how maps change, and what we think about them, in order to reduce erosity points, towards the minimum, versus changing chop points towards  the minimum. It's all about the delta function. Maybe that is 4 to 1 as Mike suggests (tentatively), maybe it is 2-1, maybe it is 1-1 (affected of course by how one scores the respective size of the chops). I have no idea. It may be more than 4 to 1, if Mike does not calculate zoom points based on cuts of subunit highways above the minimum, given the size of the chop that triggered it, but counts each and every one, assuming zooms have to be triggered (perhaps giving undue weight to avoiding triggering them). And maybe the whole exercise is a fool's errand, and we just go with a hardline Pareto optimality regime (still leaving open how to weight UCC fanouts, if one wants to incorporate the concept, within a chop point regime of course).
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #315 on: February 23, 2015, 06:07:21 PM »
« Edited: March 06, 2015, 10:15:07 PM by muon2 »

For Michigan for example, it might be useful to draw a map with the absolutely minimum number of chops, using your system as modified, and mine, where we still disagree, and see how many erosity points it has versus say my map, and see what that ratio is.

I feel pretty confident that this map is as chop-minimizing as you can get, vis-a-vis the Torie UCC rules:

MI train 2015D


Bay, Jackson, and Lapeer are all I-chops.

The Detroit districts are, as last time, drawn to the 47% BVAP standard; another chop of Oakland would be necessary if you wanted to break 50%.

No, it's not a serious suggestion.  Tongue
Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #316 on: February 23, 2015, 06:41:42 PM »
« Edited: February 23, 2015, 06:47:42 PM by Torie »

Excellent Train. That maps heads to the finals.  I doubt it wins the highway cut test however. Hopefully it doesn't! The next task is to count the highways cut. You can lose a highway cut in Kent by the way by a slight modification. Then you compare the chop count of this map, using my formula for starters, versus the highway cut number over 90 or whatever it is. And then the chop count of the perfect map, that does not care about fanouts, ala Mike's map, or something inbetween, versus the highway cut total.

More salient I guess, is how much is minimizing the chop total due to the number of chop points versus the GR fanout, versus other considerations, I wonder? What would a map look like if the chop of GR got the typical one  2 penalty points (one for the I chop, and one for the macrochop), versus my 4 points, so severing off Ottawa were a freebie? How much of that grotesque purple and blue districts is due to the GR penalty in other words?
Logged
muon2
Moderator
Atlas Icon
*****
Posts: 16,788


Show only this user's posts in this thread
« Reply #317 on: February 23, 2015, 07:21:27 PM »

Torie, my comments were only about what is a chop and how it is scored. Not microchops, not macrochops, just chops. Within the last two days I have said that I can work with microchops counting as chops at the county level and that macrochops need be nothing more than the threshold for using subunits for erosity. This is not about whether UCCs are in or out. I would like concurrence on plain old, vanilla flavored county chops that I thought were agreed to over two years ago when we were drawing plans for CA and other states.
Logged
jimrtex
Atlas Icon
*****
Posts: 11,828
Marshall Islands


Show only this user's posts in this thread
« Reply #318 on: February 23, 2015, 09:14:15 PM »

They can determine if the South Lyon/Brighton/Howell urbanized area, which is separate from the Detroit urbanized area, means that Livingston can be split off as much as other CSA counties like Monroe and Washtenaw that happen to have an older center that qualified them for their own MSA.
That is an interesting point.  I'd rather go in the opposite direction, and keep Monroe/Washtenaw in the greater Detroit area in keeping with my respect for CSAs; barring the adoption of that I've stuck to the next best thing and kept them with each other at least.  
This is simply a result of the Census Bureau grandfathering in some previous definitions when trying to determine where to split urban areas.   Because urban areas can hop and skip along highways they can extend forever.  So the Census Bureau decided to chop them at or near county lines.

Livingston used to be in the Ann Arbor metropolitan area (along with Lenawee).  So when defining urban areas based on continuous (plus hop and skip) development it got its own urbanized area.   Since the division line is near the county line (at an isthmus in the dense development peninsula), part of Livingston's urbanized area (it got its own because it is physically separate from Ann Arbor's US) it extends into Oakland County, and is named after a community in Oakland County (South Lyon).  Meanwhile the Detroit UA now extends into the northern part of the county.

The South Lyon-Howell UA is really a fiction.  Livingston is a bedroom county.

While St. Clair County has Port Huron it is relatively small, and on the eastern edge of the county.  It is easy to get into Macomb and also into Oakland and Wayne - and 1/3 of employees resident in the county do so.

Ann Arbor has a distinct identity and economy.  While there is some commuting into Wayne County, 77% of the employees stay in the county.   Monroe County has a split orientation.  There is almost as much commuting to Lucas, OH as to Wayne.  At one time Monroe was in the Toledo MSA.   The pull of Detroit has pulled it away from Toledo, but Detroit has not captured it.

If we did have to go outside Detroit UCC (as we will in 2020), then Washtenaw and Monroe would be natural targets.
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #319 on: February 23, 2015, 10:02:23 PM »

Excellent Train. That maps heads to the finals.  I doubt it wins the highway cut test however. Hopefully it doesn't! The next task is to count the highways cut. You can lose a highway cut in Kent by the way by a slight modification. Then you compare the chop count of this map, using my formula for starters, versus the highway cut number over 90 or whatever it is. And then the chop count of the perfect map, that does not care about fanouts, ala Mike's map, or something inbetween, versus the highway cut total.

More salient I guess, is how much is minimizing the chop total due to the number of chop points versus the GR fanout, versus other considerations, I wonder? What would a map look like if the chop of GR got the typical one  2 penalty points (one for the I chop, and one for the macrochop), versus my 4 points, so severing off Ottawa were a freebie? How much of that grotesque purple and blue districts is due to the GR penalty in other words?

Well, I haven't found a configuration where you can just sever Kent/Ottawa without forcing another chop elsewhere, so the Kent/Ottawa "freebie" is possibly in practice not worth it anyway in this case.  What you can do is, probably, save some erosity/sanity by still chopping Kent and fanning out, like my old map did.  Though the macrochop of Kent gives you enough of an erosity penalty that, perhaps, it's not that much savings.  Hm.

I think my Cincy, Columbus, and Pittsburgh lines are all better test cases here.  BRB with an attempt to redraw Pennsylvania so as to maximize pack in Pittsburgh (with a minimum of unpleasant ripple effects). 
Logged
muon2
Moderator
Atlas Icon
*****
Posts: 16,788


Show only this user's posts in this thread
« Reply #320 on: February 23, 2015, 11:27:38 PM »

For Michigan for example, it might be useful to draw a map with the absolutely minimum number of chops, using your system as modified, and mine, where we still disagree, and see how many erosity points it has versus say my map, and see what that ratio is.

I feel pretty confident that this map is as chop-minimizing as you can get, vis-a-vis the Torie UCC rules:



Bay, Jackson, and Lapeer are all I-chops.

The Detroit districts are, as last time, drawn to the 47% BVAP standard; another chop of Oakland would be necessary if you wanted to break 50%.

No, it's not a serious suggestion.  Tongue

EROSITY: 113 (includes Lansing UCC not acting as a supercounty for erosity)
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #321 on: February 23, 2015, 11:43:12 PM »

This is what I originally had in Pennsylvania, on muon's advisement that UCC fans were no big thang:



(The 5/9 line, and those within Washington/Allegheny, can be tweaked.  Presumably erosity could still be improved.  In particular I probably would like to change the 5/9 line to what I used below, assuming that traveling chops are not a worry.)

This is basically what Western PA would have to look like if the UCC fan was a sliding scale:



The 12/18 line goes from I-chop to macrochop; you need another (I-)chop between 12 and 5 as well.  Oh, and inequality has risen since Allegheny/Butler is close enough but underpopulated.

I have not calculated erosity for this map as opposed to the earlier one: it's entirely possible that the vagaries of Allegheny County's roughly 50,000 boroughs means that this could even score better!  The proliferation of teensy-tiny towns around Pittsburgh, and the iron necessity that a district line navigate through them, makes Pittsburgh possibly the worst place in America to apply our current erosity measurement.  It certainly fails the eye test, I think, with that ugly duckling 18 (and 12 isn't much better).  Though I guess 3 is actually nicer. Tongue

If fans are just given a flat one point hit, then my earlier PA (which– to be clear– might be improvable in Allegheny) would still be competitive.  But, since its 12 and 4 both span the UCC boundary, it gets some number of hit points.  At least two, since the old 12 has Greene (38K) and Armstrong (68K) outside of the UCC, which I guess is fortuitously just under the threshold for three hit points.  (4 is more thoroughly split between UCC and non-UCC; Pittsburgh is roughly 3 and a half districts after all.)

In any case, it makes my old map have a worse chop count, and I don't think it really deserves that.
Logged
traininthedistance
YaBB God
*****
Posts: 4,547


Show only this user's posts in this thread
« Reply #322 on: February 24, 2015, 01:46:32 AM »

To throw one last map on the fire (I probably need to take a break for a couple days after this, at least with the drawing if not the typing), here's  Ohio, minimally modified from my last map to remove UCC fans:







The Cincy area gets a macrochop of Clermont; Columbus is able to get away with an I-chop of Licking.  Erosity presumably increases some.

You will note that my District 8, especially in Clermont, is not the same as Torie's, and you're probably asking why.  His lines there look pretty sweet from a distance, no?  Why not just take Goshen, instead of Milford and Stonelick?  Well... if you zoom close enough, you'll notice that the Milford/Miami Twp. border has one of those lovely discontinuous exclave-y lines.  So that just won't do; the districts would be noncontiguous.  My lines in Licking County were, similarly, constrained by some exclave-y bits in and around Newark.

I imagine that lots of work could possibly improve this map.  It doesn't feel optimal.  I've done enough for now, though.
Logged
jimrtex
Atlas Icon
*****
Posts: 11,828
Marshall Islands


Show only this user's posts in this thread
« Reply #323 on: February 24, 2015, 05:49:54 AM »

We can further sever the graph into areas where the regions collectively are very close in population to an integer multiple of the quota.  When that is the case we can eliminate the shifts between these multi-region areas, since they don't improve the overall equality much.



To qualify as a separate area, the absolute deviation of the area from an integer multiple of the quota must less than the 0.5% x quota x sqrt(number_districts).  This is to prevent placing several underpopulated regions into an area and spreading deviation in such a way that all the districts are barely within the 0.5% limit.

For the example plan, the area of magnitude 8 (pink-lime-gray), must be within
0.5% x sqrt(8) of 8.00 quotas.  That is between 7.986 and 8.014 quotas.  Since 8.004 is within these limits it may be isolated.



As with the original 7-vertex graph, we can calculate the shifts within each area. The overall effect is small.  While reducing the number of shifts from 7 to 5, the total shifted is only reduced from 8.90% of a quota to 8.80% of a quota.

Logged
Torie
Moderators
Atlas Legend
*****
Posts: 46,057
Ukraine


Political Matrix
E: -3.48, S: -4.70

Show only this user's posts in this thread
« Reply #324 on: February 24, 2015, 07:51:34 AM »
« Edited: February 24, 2015, 08:06:14 AM by Torie »

Torie, my comments were only about what is a chop and how it is scored. Not microchops, not macrochops, just chops. Within the last two days I have said that I can work with microchops counting as chops at the county level and that macrochops need be nothing more than the threshold for using subunits for erosity. This is not about whether UCCs are in or out. I would like concurrence on plain old, vanilla flavored county chops that I thought were agreed to over two years ago when we were drawing plans for CA and other states.

You can score using whatever system you want Mike. I still like my scaling system for chops, because I think micros are between an ordinary chop and no chop for scoring purposes (I've always thought that, but yes, I would prefer microchops be penalized rather than no penalty at all, definitely), and that avoiding a county microchop is better than avoiding a subunit chop, unless the subunit chop is bigger than a microchop, and then I think they are about the same. But maybe the public square would disagree on the last point, in which event, as long as it is not a macrochop, subunit chops would score the same as county microchops (or you just limit to one subunit chop, and penalize only if a macrochop). Or as you suggest, just use macrochops for erosity purposes (hopefully adding erosity points only if the ensuing cuts are over the minimum for the size of the chop). Whichever scoring system you use on this point, it is pretty easy to make adjustments when it comes to scoring, and compare how the scores come out using different systems.

The fan out issue takes more work admittedly when it comes to making adjustments, and certainly for using my incremental approach. I just have this feeling that if the incremental approach is not used, probably the concept of UCC' should be abandoned. Mid size UCC's can just be chopped in half (albeit with a ton of highway cuts if you don't have a minimum threshold, which would the tend to have maps all over the lot survive for the Pareto optimality regime if strictly applied), so what is the point? It then really just obtains for mega UCC's.

For me this whole exercise is putting different systems to the test with real maps, so that we can see what the end product looks like.
Logged
Pages: 1 ... 8 9 10 11 12 [13] 14  
« previous next »
Jump to:  


Login with username, password and session length

Terms of Service - DMCA Agent and Policy - Privacy Policy and Cookies

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Page created in 0.086 seconds with 12 queries.