Terrorizing the Economy

[Local Yokel]

ORIGINAL: treespider

ORIGINAL: Gary Childress

Has anyone given some thought on the possibility of removing Japan's ability to control its economic output, putting Japan on par with the Allies? I assume it can be done by simply removing all factories from any bases in the editor and replacing them with monthly build rates. Just give Japan a static monthly output on engines, devices, and planes that match the historical output. It sounds like giving the Japanese the ability to manage their economy leads to more headaches and problems than it may be worth.

Ummm.... then what purpose is reserved for the merchant fleet?

Or strategic air attack upon Japanese cities? Notwithstanding the shortcomings attributed to the in-game Japanese economy, I still think it is an impressive achievement, and one that is essential to represent the economic warfare that took place in the Pacific.

[Local Yokel]

ORIGINAL: JWE

ORIGINAL: Local Yokel
If not, what is causing these gain/loss fluctuations in the armament stockpile?

Hi John,

Good stuff. Made me run some tests of my own. Very interesting results. Found some peculiar burps that I'm pushing upstairs. So can't comment just now except to say that I too see a great deal of (unexpected) variability in certain aspects of the algorithm.

What ever comes of this, I must say you will be the fellow that prompted whatever fix comes into being. Very nice work, John. Please keep it up.

Ciao. John

Not knowing the detailed design behind the code, all I can ever do is go where I think the evidence is taking me.

The impression I have formed of both armament and vehicle production is that the Japanese economy operates largely on a 'Just in Time' basis. Although pool points get accumulated from turn to turn, arriving reinforcements that have only modest equipment needs seem to get outfitted direct from production on the day of their arrival. Such a diversion of daily production into a small unit's needs apparently serves to reduce the amount of points that would otherwise be added to the pool.

The example I posted of the Type 92 Howitzers suggests to me that if the armament draw is big enough on any given day, there is an expenditure of pool points to satisfy it, and the armament devices representing that draw are treated as having come from the pool, even though they were withdrawn from it the moment they were manufactured. In the particular example posted, 4 howitzers were produced in excess of requirements, and consequently remained in the pool.

I've seen more extreme examples of this in vehicle production. In my WitP PBEM last year I was engaging in a thoroughly ahistoric roll-out of Type 3 Chi-Nu to my tank units in 1943. Each time a tank unit upgraded, the factories burped out enough Type 3's fill the unit's TOE to full establishment AND add another 59 Type 3's to the pool!

Have an idea that any change to this behaviour is going to require a code change. If that happens, my personal preference would be for the player to be given some tools for managing the way in which LCU's get fleshed out.

[JWE]

ORIGINAL: Local Yokel
Not knowing the detailed design behind the code, all I can ever do is go where I think the evidence is taking me.

Yep, yep, yep. I too have seen some anomalous behavior.

Built a single-screen test bed with everything set inactive. Put a limited number of tailored units on spaced-apart arrival schedules at a base with a (similarly tailored) command HQ and 200,000 supply. All devices were turned off except for 1 Squad, 1 Vehicle, and 1 Gun (armament) type, and all LCs were set to 10, all device content was set to a multiple of 10. Manpower, vehicle, and armament pools were the variable. All pools of everything else were set to 0 (no build, and no device pools).

Response variables were pool (virtual pool) decrement upon introduction of a unit at full TOE and at a reduced TOE and the time rate for nominal rebuild to full establishment. Also evaluated unit/item virtual pool decrements against nominal cost (in so far as they are understood).

Lots of cool stuff came out. Also some not-so-cool stuff. Am working with Andrew and Michael on the observed anomalys. Maybe they are, maybe they aren't, don't know; but they are there and occasionally ugly, so need to get a good grip.

Haven't seen the double dipping that you are getting with the Type 3 Chi-Nu tanks, but there is a possible explanation. Michael has been doing a lot of tweaking on the upgrade, and recombine, and bears, oh my, part of the game, particularly with regard to the device upgrades within a combination of LCUs at respectively different upgrade levels. It is "possible" that what you are seeing is the "virtual" pool, of vehicle points, producing your Type 3s, and your "previous" vehicles being returned to the "actual" pools and then being automatically upgraded within the pool space, according to code consultation of the particular Device Data upgrade schedule (another of those nasty glitches that bite what-if stuff in unexpected places).

Implications, implications, implications. Woof !! Damn you John, for making me do this. Thank you John, for making me do this.

Ciao. J

[Local Yokel]

In haste, as I'm about to leave for the airport.

I dug into my old records to check what happened on Chi-Nu upgrades. The last upgrade that occurred seems to have drawn from accumulated Chi-Nu stocks in the pool rather than taking up production-of-the-day. This was at turn 576, and I show the device pool record for both Type 3 Chi-Nu, and the Type 1, from which model the upgrade to Chi-Nu took place.

So far as I can tell, the Type 1's replaced have gone into the pool, and all that has been drawn down is the total of Type 3 that were already available in the pool.

Of course, this is WitP rather than AE, though I'm not aware of production routines having changed much betyween the two games.

[Pascal_slith]

For anyone who wants to test the economic/production system, I've created a Scenario 75 (from Scenario 1) with all LCU, Naval, and Air units DELAYED till Jan. 1, 1945. Even better, keep a few merchies on map for economic transport set to Dec. 6, 1941.

I did this using witploadae.exe . Quite easy, just set the delay date in the ship and group file for all to 450101. In the location file, be careful. Only change the delay date for actual LCUs and HQs, not bases, industry, or OOB structures.

Procedure is to first use the AE Editor. Load the scenario you want to modify (here I used Scenario 1). Save the scenario in slot 75 (in my case). Take all the files ending in '075' in the filename and put it in a separate folder where you should also have witploadae.exe .

You extract the scenario files to .CSV files by running witploadae (in a DOS command prompt window pointing to the proper sub-directory). You can use the line 'witploadae /?' to see all the command line options. '/e' extracts the files.

Modify the three .CSV files that concern ships, air groups, and locations accordingly. I just changed the entry in the 'Delay' column to '450101' and copied it all the way down for the ships and air groups files. In the locations file, ONLY change the HQs and LCUs delay figure.

Regenerate the scenario files using witploadae.exe . Be careful, witploadae ONLY sends these files to the subdirectory 'C:\Matrix Games\War in the Pacific Admiral's Edition\SCEN'. Make sure it exists (my game is located elsewhere).

[GaryChildress]

ORIGINAL: Local Yokel

ORIGINAL: treespider

ORIGINAL: Gary Childress

Has anyone given some thought on the possibility of removing Japan's ability to control its economic output, putting Japan on par with the Allies? I assume it can be done by simply removing all factories from any bases in the editor and replacing them with monthly build rates. Just give Japan a static monthly output on engines, devices, and planes that match the historical output. It sounds like giving the Japanese the ability to manage their economy leads to more headaches and problems than it may be worth.

Ummm.... then what purpose is reserved for the merchant fleet?

Or strategic air attack upon Japanese cities? Notwithstanding the shortcomings attributed to the in-game Japanese economy, I still think it is an impressive achievement, and one that is essential to represent the economic warfare that took place in the Pacific.

Same purpose as the Allied merchant fleet. Same purpose as bombing Allied cities.

[JWE]

ORIGINAL: Andrew Brown
Last time I checked, a lot of that "free fuel" had reappeared in the scenario files, including da Babes (I think). If you like JWE, I can grab the latest da Babes scenario, pull all of that fuel out again, and send you a copy? From memory it was something like 600,000 fuel points.

Andrew

Yes indeed. It's like weeds. Checked and it's not bad for Japan; only a couple score ships and a couple of the late arrival class additions (CHa, CH, etc..). Fixed all those. Allies were a pita. Fixed everything except Aden and Cape Town arrivals.

Everything arriving at its construction port arrives empty. Everything arriving at Eastern US and UK arrives empty. Everything arriving at an off-map base (Panama, Aden, Cape Town, etc..) arrives half-full (or half-empty [;)]). Almost done. Will send you a copy so you can check and make sure it's ok.

Thanks. Ciao. J

[Mac Linehan]

ORIGINAL: JWE

Something Maj Faustini said a while ago just sunk into my thick skull. "Why do I have to keep all the factories running, when I get full-up divisional sets for free?"

Seems many people are turning off factories as unnecessary and using the excess HI points to do more planes and engines, or shipyards, or some such. We worked a lot with Andrew Brown to make all new ship construction arrive with empty fuel tanks, and ships appearing at off-map bases arrive half empty, to keep people from getting a "free fuel" bonus. AndyMac has set many later war arrivals at a reduced establishment, but there's still a gazillion Divisions, Brigades, Rgts and bears, oh my, that appear fully equipped and ready to boogie at absolutely no cost to the pools. So ...

Gonna look at arrival dates for LCUs and tweak the arrival 'device content'. TOEs stay the same, but units will need to suck sustenance from the pools in order to get operational and get to TOE levels. That will force retention of vehicle and armaments factories and reduce the potential for excess HI points for use with more sexy stuff.

Everything, and I do mean everything, gets cut. 1942 arrivals get cut to 50%. 1943 arrivals get cut to 33%. 1944 arrivals get cut to 25%. 1945 arrivals get cut to 20%. Understand that some late-date units (like the tanks) were formed from previously disbanded units, so will have to be careful with the withdraw 1 or 2 thing. If one withdraws a tank unit and dumps the tanks into the pools, the later arriving (reformed/renamed/whatever) tank unit arrives totally empty and must rely on the pool content and intervening production to get fleshed out.

Not sure exactly what effect this will have on the long term econometric model, but do remember Andrew Brown taking this sort of freebie device pool stuff into account when he developed the thing - just like the freebie fuel thing - so don't think this will break anything except gamey JFBs hearts.

What ya'll think?

JWE -

Was reading the above very late last night, and it just came to me:

Babes Lite

The Big Bad Babes

The Killer Babes (most recent), and -

The Rabid Babes (TM The Babes Team)

Just my take <grin>

Mac

[JWE]

Well, Mr Palmer, can't speak to the Chi-Nu thing yet.

But yepperino, every time a Japanese unit enters the game, after opening day, it decrements the pools to the extent of its editor devices. I haven't determined just which pools are accessed in what order, but if there are no specific devices in the pools then the newly arriving unit will "attempt" to establish itself, according to the cost rules, from manpower, armament, vehicle points. If there ain't enuf of any one of these on arrival day (all other pools being 0), the unit does not arrive till there are.

Couple more things to look at; the relative draw down of 'device' pools vs 'econ' pools when a new boy hits town; the relative effect of different device upgrade status; and the time-rate of build-to-toe given 'device' or 'econ' pool numbers.

Woof !! Ciao. John

Actually John, you are quite correct, I was working under a false assumption. Since everybody pays the pool piper to arrive in-game, no matter how well fleshed out, it doesn't make sense to tweak/decrement the unit content vis a vis the TOE. A full toe unit will pay full toe costs, a reduced toe unit will pay reduced toe costs but have a build-time penalty and incur pay-as-you-go costs. This is beginning to make sense to me.

God help us all, though, if Kristian inserts his 'toe' into the equation. [:D]

[DuckofTindalos]

Aw, why'd you have to go there, Homie?