What AMPU Can Borrow from Victoria 3

[OrangeP47]

3 minutes ago, MrPotatoTed said:

There's actually multiple rules in 3.0 for calculating EVs, in multiple sections.  Here's the first set:

 

• First check to see if you are in a census election. A census election is the first election after a decade. Thus, the 1860 election does not calculate new EVs because that is the decade year, but the 1862 midterm will calculate new EVs and new US Reps. 

• Use the historical EVs for an election modified by any events or actions that might have altered the historical number.  Therefore, if Ohio historically had 21 EVs in 1864, but your Ohio happens to have 19 EVs, then the historical number of 22 EVs for 1872, should be 20 in your game, because your Ohio is 2 EVs behind the historical value. 

• If a state joins before they did historically, then give that state 3 EVs and add it to the historical EVs for that election. 

• If a state has not yet joined when they did historically, then simply erase their EVs from the historical EVs for that election. 

• See the spreadsheet for ahistorical state EVs

There's also 3.0 rules on calculating number of Reps per state, but they also don't reference a cap.

Then there's a second set of rules later in 3.0.

 

The US Census will automatically be in play when the US Constitution is adopted, and the new census effects will be visible at the beginning of every Half-Term of a US Census Year. That is, the half-term in which an election will end with a 2, so 1792, 1802, 1812, 1822, 1832, and so forth, will show new or lost seats/EVs for each state for the upcoming election. The old census US House seats will be held by the incumbents until this election takes place, then the new seats are applied. 

 

• First, apply historical census changes. By adding the +/- that historically occurred to whatever the EVs are currently at.  

• Secondly, apply any adjustments to the EVs for states based on Scripted Events

  • Scripted Events regarding “Indian Removal” are the primary Scripted Events that will affect state growth. 

• Thirdly, apply an adjustment to the EVs for Industry shifts: 

  • If an industry is improved twice within 10 years, then there’s a 25% chance of +1 EV in that state at the next election under a new census. 

  • If a state becomes the new leader in an industry, then there’s a 25% chance of +1 EV in that state at the next election under a new census. 

  • If a state’s primary industry changes to a new industry, then there’s a 10% chance of +1 EV in that state at the next election under a new census. 

  • If a state industry decreases twice within 10 years, then there’s a 25% chance of -1 EV in that state at the next election under a new census. 

  • If a state is replaced as the state leader in an industry then there’s a 25% chance of -1 EV in that state at the next election under a new census.

• Fourthly, apply any adjustments to the EVs based on legis props:

  • The US Constitutional Convention might start off with some major adjustments because of the slavery question. 

• Fifth, add up the total # of EVs for every state, ensuring no state falls below 3 EVs. 

• Set the number of EVs based on the US House Cap total, if one exists, and the total number of EVs you’ve calculated above:

  • If no cap exists, then this process doesn’t need to be done. 

  • If a cap exists, then compare the EVs you’ve calculated above to the House Cap + the number of states x 2 to take into account the US Senators. (Do not count DC’s, the Moon’s, or Mars’s EVs). You’ll have to get your number in compliance by addition or subtraction. Randomly add/subtract EVs, ensuring that no state loses or gains two times as many EVs as any other state in this process. Also, no state should fall below 3 EVs when it is done. 

• You should now have EVs for every state. 
  
  It, too, doesn't reference there being a max or otherwise capping it unless there's a House Cap in place.
  
  Then there's a second set of rules for determining Reps again too.  Haha.
  
   

 

I've just been using the Census.

[OrangeP47]

I don't have it open but the other rules section might be the biases for seats above what the census doc has for defaults, for cases where the census has generated a new focus rep?

[MrPotatoTed]

1 hour ago, OrangeP47 said:

Now would probably be a good time to mention in Grad School my favorite professor said the easiest way in our field to get set for life was graduate in a year ending with 9 or 0 and join the Census Bureau.

Your professor lied.  ;c)  They lay off the majority of the census bureau workers after each census -- it's kind of seasonal work, and it doesn't pay enough to just not work in between censuses.  (Source: I'm a federal employee who hires other federal employees...I have a lot more candidates to choose from after each census, and they all just left the census bureau each time. haha.)

[OrangeP47]

1 minute ago, MrPotatoTed said:

Your professor lied.  ;c)  They lay off the majority of the census bureau workers after each census -- it's kind of seasonal work, and it doesn't pay enough to just not work in between censuses.  (Source: I'm a federal employee who hires other federal employees...I have a lot more candidates to choose from after each census, and they all just left the census bureau each time. haha.)

We're upper level material though 😉

[MrPotatoTed]

1 hour ago, OrangeP47 said:

@vcczar @MrPotatoTed

Okay, here's an in depth report on the 1850 and 1860 (estimated) census from our game. Please note, it's only 1856 in game, so the 1860 figures are sort of the numbers you'd get from wikipedia between cycles that are from the American Family Survey or whatever.

IOTL The US had a population of 23,191,876 in 1850. Our census came out to roughly 23,371,876, or 100.78% of the OTL's population.  This really isn't a concern for game balance as you can see, but to be fair the 1840s were a really bad decade economically in our sim.  Indeed, realistically, we probably should have come in below 100% of actual 1850 population if we were just to eyeball a number, but this is what the system produced. For calculation purposes, IOTL there were 234 Congressional Districts as a result of the 1850 census and our census produced 236.

For 1860, we're on course for an ESTIMATED increase of 11 more Congressional districts (on average) for a total of 13 additional districts since 1840 (the 11 are since 1850). This doesn't count historical increases.  Counting them, it would mean the historic US apportionment in 1860 gave 241 seats and we'd have 254.  These 11 extra seats represents an addition 1.32 million people in the country, which at first sounds like a lot, but the 50s so far has been a boom time and is probably realistic, plus it still isn't that great percentage wise. The real 1860 census enumerated 31,443,322 people in the US, and ours would enumerate 32,943,322.  This is a 104.77% increase over OTL, but also that figure had nearly a 1% head start from 1850.  It's not terribly a big deal.  I see you are concerned with the run away, exponential effect, which WOULD be true if this was actually simulating population growth, but it's not, it's simulating congressional districts, and how many people per congressional district has increased over time, so each census the effect is blunted.  Yeah, you're adding more and more people for the same number of congressional districts each time if you add, say 10 each time, but you have an even larger pool of *existing* congressional districts that are also ballooning which reduces the effect.

 

1 hour ago, vcczar said:

Ok, I won't fix it then. 

I'm not sure I understand how @OrangeP47 is calculating actual population.

For my game, we just started implementing the new census rules right before the 1790 census.  That saw a net of -1 EV from the historical number, so that's not too shabby at all.

However, I haven't even reached 1800 yet, and already we're looking at a net of +10 EVs.  That's on top of the real life growth of the country (+44 EVs) -- though notably I likely won't have as many states as real world 1800 had by then. 

Important to note that almost all of that +10 EV growth was from the Cotton Gin and Cotton Boom events -- which of course are historical and therefore already baked into the historical EV numbers. 

[vcczar]

9 minutes ago, MrPotatoTed said:

Your professor lied.  ;c)  They lay off the majority of the census bureau workers after each census -- it's kind of seasonal work, and it doesn't pay enough to just not work in between censuses.  (Source: I'm a federal employee who hires other federal employees...I have a lot more candidates to choose from after each census, and they all just left the census bureau each time. haha.)

 

7 minutes ago, OrangeP47 said:

We're upper level material though 😉

I worked for the 2010 US Census in NYC. It paid $18.75 an hour, but it lasted only about two or three months. 

[vcczar]

19 minutes ago, MrPotatoTed said:

There's actually multiple rules in 3.0 for calculating EVs, in multiple sections.  Here's the first set:

 

• First check to see if you are in a census election. A census election is the first election after a decade. Thus, the 1860 election does not calculate new EVs because that is the decade year, but the 1862 midterm will calculate new EVs and new US Reps. 

• Use the historical EVs for an election modified by any events or actions that might have altered the historical number.  Therefore, if Ohio historically had 21 EVs in 1864, but your Ohio happens to have 19 EVs, then the historical number of 22 EVs for 1872, should be 20 in your game, because your Ohio is 2 EVs behind the historical value. 

• If a state joins before they did historically, then give that state 3 EVs and add it to the historical EVs for that election. 

• If a state has not yet joined when they did historically, then simply erase their EVs from the historical EVs for that election. 

• See the spreadsheet for ahistorical state EVs

There's also 3.0 rules on calculating number of Reps per state, but they also don't reference a cap.

Then there's a second set of rules later in 3.0.

 

The US Census will automatically be in play when the US Constitution is adopted, and the new census effects will be visible at the beginning of every Half-Term of a US Census Year. That is, the half-term in which an election will end with a 2, so 1792, 1802, 1812, 1822, 1832, and so forth, will show new or lost seats/EVs for each state for the upcoming election. The old census US House seats will be held by the incumbents until this election takes place, then the new seats are applied. 

 

• First, apply historical census changes. By adding the +/- that historically occurred to whatever the EVs are currently at.  

• Secondly, apply any adjustments to the EVs for states based on Scripted Events

  • Scripted Events regarding “Indian Removal” are the primary Scripted Events that will affect state growth. 

• Thirdly, apply an adjustment to the EVs for Industry shifts: 

  • If an industry is improved twice within 10 years, then there’s a 25% chance of +1 EV in that state at the next election under a new census. 

  • If a state becomes the new leader in an industry, then there’s a 25% chance of +1 EV in that state at the next election under a new census. 

  • If a state’s primary industry changes to a new industry, then there’s a 10% chance of +1 EV in that state at the next election under a new census. 

  • If a state industry decreases twice within 10 years, then there’s a 25% chance of -1 EV in that state at the next election under a new census. 

  • If a state is replaced as the state leader in an industry then there’s a 25% chance of -1 EV in that state at the next election under a new census.

• Fourthly, apply any adjustments to the EVs based on legis props:

  • The US Constitutional Convention might start off with some major adjustments because of the slavery question. 

• Fifth, add up the total # of EVs for every state, ensuring no state falls below 3 EVs. 

• Set the number of EVs based on the US House Cap total, if one exists, and the total number of EVs you’ve calculated above:

  • If no cap exists, then this process doesn’t need to be done. 

  • If a cap exists, then compare the EVs you’ve calculated above to the House Cap + the number of states x 2 to take into account the US Senators. (Do not count DC’s, the Moon’s, or Mars’s EVs). You’ll have to get your number in compliance by addition or subtraction. Randomly add/subtract EVs, ensuring that no state loses or gains two times as many EVs as any other state in this process. Also, no state should fall below 3 EVs when it is done. 

• You should now have EVs for every state. 
  
  It, too, doesn't reference there being a max or otherwise capping it unless there's a House Cap in place.
  
  Then there's a second set of rules for determining Reps again too.  Haha.
  
   

 

Thanks for sharing this. I'll need a lot more undistracted time to untangle this, synthesize it, and rewrite it. Hopefully, Sunday I can do that. It might not be until Monday, but that depends if I can get all my school planning done on Sunday. 

[MrPotatoTed]

For funsies, by 1798:

DE: +1 EV more than their historical total.
GA: +1 EV
MD: +1 EV
NH: +1 EV
NY: +3 EV
PA: +2 EV
RI: +1 EV
 

[OrangeP47]

I should clarify. This was refering to data analyst positions not actually boots on the ground positions.

How I'm getting pop figures:  there's a chart here https://en.wikipedia.org/wiki/United_States_congressional_apportionment That has how many people were per each CD each census.  With that, it's just some quick math to figure everything up.  It's not an actual game mechanic, but a real life logic argument.

I don't see an additional 10 EV in 1800 as a problem.  I can do the math for you later if you'd like but I'm sure it's not alarming.  Stepping out ATM to get fried chicken.

[OrangeP47]

Also Ted, before I head out the door, I had a thought:  Are you aware each EV change only has a 25% chance to actually happen?

[vcczar]

6 minutes ago, OrangeP47 said:

I should clarify. This was refering to data analyst positions not actually boots on the ground positions.

How I'm getting pop figures:  there's a chart here https://en.wikipedia.org/wiki/United_States_congressional_apportionment That has how many people were per each CD each census.  With that, it's just some quick math to figure everything up.  It's not an actual game mechanic, but a real life logic argument.

I don't see an additional 10 EV in 1800 as a problem.  I can do the math for you later if you'd like but I'm sure it's not alarming.  Stepping out ATM to get fried chicken.

I'd like to see the math for sure. 

[MrPotatoTed]

Just now, OrangeP47 said:

Also Ted, before I head out the door, I had a thought:  Are you aware each EV change only has a 25% chance to actually happen?

Yep!  The only difference is I roll the 25% chance in the moment and notate the outcome, rather than have to go back ten years and figure out how much each industry grew after the fact.  Haha.  

The Cotton Gin and Cotton Boom (I forget what it's actually called) scripted events both are huge boons to the plantation industry in every slave state.  And in my game, 13 of our 14 states are slave states currently.  Plus there were events that grew the finance industries too, which is why NY and PA are especially flush with EVs right now.  I did roll on the 25% chance, it just hit a lot.

It would actually be more than +10 EV, but it was offset by several states getting -1 EV last time, and as the rules specify that these EVs bonuses/penalties are cumulative for all future censuses, it evened out to +10.

[MrPotatoTed]

10 minutes ago, OrangeP47 said:

I should clarify. This was refering to data analyst positions not actually boots on the ground positions.

How I'm getting pop figures:  there's a chart here https://en.wikipedia.org/wiki/United_States_congressional_apportionment That has how many people were per each CD each census.  With that, it's just some quick math to figure everything up.  It's not an actual game mechanic, but a real life logic argument.

I don't see an additional 10 EV in 1800 as a problem.  

In 1800 alone, maybe not.  Just remember that it's cumulative.  Right now, I'm averaging +9 EV in every census.  So...

1800: +9 EV more than the historical EV count
1810: +18 EV
1820: +27 EV
1830: +36 EV
1840: +45 EV
1850: +54 EV.
1900: +99 EV.
1950: +144 EV.
2000: +189 EV.
2020: +207 EV.  So...if I don't institute a cap...I'm currently on pace to have 743 EV by 2020 in my current game.

 

[vcczar]

1 minute ago, MrPotatoTed said:

In 1800 alone, maybe not.  Just remember that it's cumulative.  Right now, I'm averaging +9 EV in every census.  So...

1800: +9 EV more than the historical EV count
1810: +18 EV
1820: +27 EV
1830: +36 EV
1840: +45 EV
1850: +54 EV.
1900: +99 EV.
1950: +144 EV.
2000: +189 EV.
2020: +207 EV.  So...if I don't institute a cap...I'm currently on pace to have 743 EV by 2020 in my current game.

 

Historically, the cap came in 1929, I believe. If it does go on this projection it would definitely mean I need to create rules for equalizing history/alternative EVs. I think doing away with the historical EVs would lead to some really strange results. 

[MrPotatoTed]

Maybe there needs to be more ways for a state to lose EV?  Right now, only being surpassed as industry leader (which only applies if a state actually becomes an industry leader) or losing 2 levels of industry in a given decade (extremely unlikely as the governor has to choose to do so and there's rarely an incentive for it) can cost you EV.  So 99% of the time, a state grows or stays stagnant -- there's almost no opportunity to get smaller.  This naturally leads to run-away EV growth nationally as I outlined above, over time.

So I think the current system is fine, provided that there's roughly equal opportunities for individual states to both gain and lose EV.  The issue is it's weighted too strongly in favor of gaining EV I think.

[OrangeP47]

6 minutes ago, MrPotatoTed said:

Maybe there needs to be more ways for a state to lose EV?  Right now, only being surpassed as industry leader (which only applies if a state actually becomes an industry leader) or losing 2 levels of industry in a given decade (extremely unlikely as the governor has to choose to do so and there's rarely an incentive for it) can cost you EV.  So 99% of the time, a state grows or stays stagnant -- there's almost no opportunity to get smaller.  This naturally leads to run-away EV growth nationally as I outlined above, over time.

So I think the current system is fine, provided that there's roughly equal opportunities for individual states to both gain and lose EV.  The issue is it's weighted too strongly in favor of gaining EV I think.

Back now.  Some events cause industries to decrease too.  Our 1840s almost got wiped out with a lot of EV losses because of that but it didn't happen.  I'll do math when I'm done eating and respond to all this stuff.

[OrangeP47]

Had a thought, though, and really it's the best kind of thought because it means I don't have to do math (it seems your objection is not really with the pop changes anyway).  What if we made it so having bad economic meters gave a % chance each linger phase to delete some EVs next census?

[vcczar]

14 minutes ago, MrPotatoTed said:

Maybe there needs to be more ways for a state to lose EV?  Right now, only being surpassed as industry leader (which only applies if a state actually becomes an industry leader) or losing 2 levels of industry in a given decade (extremely unlikely as the governor has to choose to do so and there's rarely an incentive for it) can cost you EV.  So 99% of the time, a state grows or stays stagnant -- there's almost no opportunity to get smaller.  This naturally leads to run-away EV growth nationally as I outlined above, over time.

So I think the current system is fine, provided that there's roughly equal opportunities for individual states to both gain and lose EV.  The issue is it's weighted too strongly in favor of gaining EV I think.

Ok, I'll look into that. I'll create some events that are based on historical state EV losses. We gotta get this sorted out one way another soon. I still think equalizing the historical/alternate EVs might be necessary. I wish there was a way to make these changes after Anthony codes the game so I can simulate a ton of things to figure out a better fix, but I'm pretty sure that will be a huge hassle on his end. 

[MrPotatoTed]

11 minutes ago, OrangeP47 said:

Back now.  Some events cause industries to decrease too.  Our 1840s almost got wiped out with a lot of EV losses because of that but it didn't happen.  I'll do math when I'm done eating and respond to all this stuff.

Interesting, do you know what event(s) it was?

[MrPotatoTed]

8 minutes ago, OrangeP47 said:

Had a thought, though, and really it's the best kind of thought because it means I don't have to do math (it seems your objection is not really with the pop changes anyway).  What if we made it so having bad economic meters gave a % chance each linger phase to delete some EVs next census?

This assumes that people flee the US when the economy is bad, but I'm not sure that's historically accurate.

I've always wished most of the meters were at the state level, with the national meters being an average of the meters in each state. Especially economy -- some states are doing great, some are doing terribly, etc.  Governors would naturally be focused on fixing it.   But that's probably AMPU 2.

Edited October 28, 2022 by MrPotatoTed

[OrangeP47]

Just now, MrPotatoTed said:

Interesting, do you know what event(s) it was?

I'll look in while.  The problem with it was it did a -1 to ag and a -1 to manufacuter, but if it did a -2 to just one thing instead it'd cream'd us.  That's what I was refering to when I said eyeballing it in 1850 we should have lost population.

@vcczar The problem with equalizing is ahistorical states.  I don't think you're going to be able to do it if we have Canada or Mexico or who knows what else, it's too complicated because we don't have historical data.

[OrangeP47]

Indeed, this is a general thought I've been having, but we really should have tested having Canada because I'm concerned about this very issue.

[MrPotatoTed]

Just now, OrangeP47 said:

Indeed, this is a general thought I've been having, but we really should have tested having Canada because I'm concerned about this very issue.

Good point.  Also the possibility that you have way more or way fewer states than historical censuses had.  So for "Historical total" it should be the historical total of whatever states are currently active.  If TN isn't a state in my game, then TN's historical EVs don't count for the historical census.

If Canada is part of the US, then you'd include Canada's (fictional) historical EV for that decade.  I believe @vcczar has already populated that.  

So the cap goes up and down depending on what states are in play (if we do a cap at all)

[OrangeP47]

Just now, MrPotatoTed said:

Good point.  Also the possibility that you have way more or way fewer states than historical censuses had.  So for "Historical total" it should be the historical total of whatever states are currently active.  If TN isn't a state in my game, then TN's historical EVs don't count for the historical census.

If Canada is part of the US, then you'd include Canada's (fictional) historical EV for that decade.  I believe @vcczar has already populated that.  

So the cap goes up and down depending on what states are in play (if we do a cap at all)

It's, in general, why I support no-cap.  It's a lot easier to handle ahistorical states/arrangements, etc, what have you.  That said, I was going to bring up, the cap law in Congress *definitely* needs to be a scale instead of just simply 435.

[OrangeP47]

Okay, here's the quick math for Ted's 1800 census.

Your 1800 population is 5,658,483, that is 350,000 higher than IOTL, for an increase to 106.59% from historical baseline.  I wouldn't call that too concerning.

Edited October 28, 2022 by OrangeP47