Why did Paradox choose Provinces over a free/open map

[Sinbuster]

It's a long way off but I'm with you on no provinces. Capturing province x and y so I can encircle z is a simple and effective way to model blitzkreig but I always imagined a hexless map (or really small hexes) that was many, many, many times larger than the current generation computer could handle. Once you can properly model a huge landscape like eastern europe, complete with cities, towns, roads, rail, swamps, mountains, etc, the shackles will really come off the genre. Imagine utilizing actual roads and highways to form encirclements, using the terrain to your advantage, and watching swaths of troops move over 2000km of virtual landscape. Again, I say imagine because something like that is a few decades away yet. The AI for it, maybe a few decades more.

 

[unmerged(139259)]

Maybe hex map could be a solution, but not for a world map in HOI series. Even those small provinces in HOI3 (14000+) are quite pain to be managed properly by a micromanaging players (like me). Leading manually your forces is something I love to do and while you are invading Low Countries or Poland as GER, it can be done, but with Barbarossa, it starts to be really messy. Can't imagine this with hexes or even with an open map.

 

[Ricox]

I'm not really sure what you mean by a "free" map. All the games I've ever played have some sort of boundaries, to designate area of control to force units into combat. The only game I can think of that tried a truly open map was "SuperPower 1/2" and it just didn't work very well.

Gunnm is right, TW has provinces just like HOI they just represent them differently. Some games use hexs as "provinces" ie Civil War Generals 2, Gary G's War in the Pacific AE, and games like that.

Speaking of the TW series how is it doing? I haven't paid any attention to it since I got ripped off by ETW. They made a Shogun 2? If I remember correctly people had been crying for that for years, could it be that SEGA/CA is turning over a new leaf and actually listening to TW fans? If so, I might, might just have to force myself back. How is the AI or is it still as dumb as ever? Don't mention MP don't care for it. Sorry to derail.

And how does the free map in SP1 and SP2 not work out? It works out perfectly, just a bit annoying in multiplayer when someone decides to deploy 1 infantry units all over the province. It has provinces, thought.

 

[unmerged(139259)]

It's a long way off but I'm with you on no provinces. Capturing province x and y so I can encircle z is a simple and effective way to model blitzkreig but I always imagined a hexless map (or really small hexes) that was many, many, many times larger than the current generation computer could handle. Once you can properly model a huge landscape like eastern europe, complete with cities, towns, roads, rail, swamps, mountains, etc, the shackles will really come off the genre. Imagine utilizing actual roads and highways to form encirclements, using the terrain to your advantage, and watching swaths of troops move over 2000km of virtual landscape. Again, I say imagine because something like that is a few decades away yet. The AI for it, maybe a few decades more.

Honestly speaking, something like this is available already in the games. However, this level of detail belongs to tactical strategies, not to grand-strategy genre. In grand-strategy, you will always need some kind of abstraction - simply to orientate yourself in the game. I really don't want to be able to see every single tank on the whole Eastern Front (figuratively speaking).

 

[Modo]

I think PI is getting as low per province/hex/whatever as the hardware permits--see the progress between HoI2 and HoI3. Going further is a matter of processing power. It would be nice to see units move more precisely, with even smaller parts shown separately, but you'd need to have a more detailed map, creating a much bigger processor load. I'm sure fighting can be processing in parallel, but current hardware only gives you like 8 threads max, which is puny on a map of thousands times thousands of areas. (That's assuming that making fights work in parallel is trivial, which I'm not sure it is. And disregarding the headache of not having fixed provinces to place strategic buildings/targets.)

 

[unmerged(219522)]

Perhaps a system whereby you can micro-manage an individual attack on a province in basic form? Default would be AI control of the attack (basically what we have now), but the option could be given to focus your tank spearhead on the right side of the province, and infantry support, while you send another division to attack on the left to tie up enemy defences. We're talking about quite a while in the future though because, obviously, the technical implications of this would be quite large

Would add a larger sense of control when fighting real important battles such as city sieges, or pivotal battles along your spearhead, hence why the default would be set to the standard current form where a "generic" attack is made. As a result, micromanagement wouldn't be an issue. We would however, need an AI that could deal with it, as you essentially add a 2nd layer to the AI; the strategic level which is currently existing, and the new general "tactical" (really it's still strategic but a bit more zoomed in) AI. Perhaps you could also achieve this idea by splitting individual provinces into further hexes/sections (that appear when you zoom in further onto a province) to perform these manoeuvres, with the movements done at brigade rather than divisional level. Not only would this add greater immersion, but it would also be much more fun seeing your brigades attempt to fight through the city sections of Stalingrad as opposed to divisions attempting to just attack the city.

This idea would enable a "free map", whilst not taking away from what HOI really is or adding any sort of extra micro burden on the player.

Personally I think this idea I just came up with it much more effective than my idea earlier. Would people be in favour of such a system? (for hoi 4 or 5)

Edit - Generic provinces like "plains" would each have the same tactical layout, yet important urban places could have their unique tactical views.

 

[North_Wolf]

For the econ then provinces make things alot easier to handle (TA series does this as well). I prefer the current system since it makes it easier to zoom out and see whats going on in general (infrastructure, weather, terrain).

I would suspect a border free game would increase the macro level to much, currently you lump 4-5 divisions into a provice and they do their own thing (their commanders handle the on ground details). With no borders then you would be wanting to direct divisions much more carefully and in turn increasing the workload for the player by a factor of 4x-5x times.

Hardware or not, I think it would increase complexity for the human for limited gain.

A border free concept might be better suited for an operational level game. Like the invasion of Poland etc.

Currently there is no way for example to simulate the later stages of stalingrad (two small pockets inside the city) although you could with a border free map. I think the added depth would be out of place in a game of such massive scale.

I love Paradox as a developer, maybe they could cook up an open map game of a better scale, I am getting visions of an Army level simulation where terrain, supply, weather all play a part in how a battle plan goes.

 

[Premu]

I think PI is getting as low per province/hex/whatever as the hardware permits--see the progress between HoI2 and HoI3. Going further is a matter of processing power. It would be nice to see units move more precisely, with even smaller parts shown separately, but you'd need to have a more detailed map, creating a much bigger processor load. I'm sure fighting can be processing in parallel, but current hardware only gives you like 8 threads max, which is puny on a map of thousands times thousands of areas. (That's assuming that making fights work in parallel is trivial, which I'm not sure it is. And disregarding the headache of not having fixed provinces to place strategic buildings/targets.)

Hardware limits are not the only reason - it's also the playability which is taken into account.

For example, in Vicky II the amount of provinces is similar to Vicky I or HOI II, as in this game smaller provinces wouldn't really add anything to the game play. It would make the game only more complicated.

 

[Modo]

Hardware limits are not the only reason - it's also the playability which is taken into account.

I hear you. However, I think another level of detail can be achieved while increasing the fun factor. I think units can be made to move "freely" (i.e. along a hex grid small enough to not look like a grid at first glance), and strategic targets can be turned into areas, not just points on the map. But I also think this will require a much more sophisticated AI, and serious use of parallel processing (including hardware availability), so it's more likely to come in HoI5 than in HoI4.

 

[unmerged(53322)]

It's a long way off but I'm with you on no provinces. Capturing province x and y so I can encircle z is a simple and effective way to model blitzkreig but I always imagined a hexless map (or really small hexes) that was many, many, many times larger than the current generation computer could handle. Once you can properly model a huge landscape like eastern europe, complete with cities, towns, roads, rail, swamps, mountains, etc, the shackles will really come off the genre. Imagine utilizing actual roads and highways to form encirclements, using the terrain to your advantage, and watching swaths of troops move over 2000km of virtual landscape. Again, I say imagine because something like that is a few decades away yet. The AI for it, maybe a few decades more.

you're basically thinking of War In The Pacific.

 

[Danmark Rising]

I personally love the current province system. But them I spend most of my time in Political map mode becuase I watching my forces turn a province to my color is part of the thrill for me.

 

[Heide]

Anyone remember the vaporware "Road to Moscow"?

you're basically thinking of War In The Pacific.

Or perhaps they are envisioning this:

http://pc.ign.com/articles/131/131163p1.html

note the date on this article - talk about an ambitious game that would have been incredibly ahead of it's time.

Beta review here: http://groups.google.com/group/comp.sys.ibm.pc.games.war-historical/tree/browse_frm/thread/cd033b1474aecb58/d69e80eb8224f9b2?rnum=1&q=%22road+to+moscow%22+review&_done=/group/comp.sys.ibm.pc.games.war-historical/browse_frm/thread/cd033b1474aecb58/a14950c0f4f4578d?lnk%3Dgst%26q%3D%2522road%2Bto%2Bmoscow%2522%2Breview%26

 

[DanSez]

A 1999 review of a great idea on a game that never finished development. Someday when AI really means something more than writing scripts to trigger from a set of conditions - adaptive not reactive processing.

I haven't played the games mentioned with open maps. I would like HoI3/4 to do away with the provence system (inherited from HoI2/Vicki/EU) and create a set blocks of similar sized terrain. Weither that is 100km blocks or 20km blocks - might depend on what the processor and graphics engine can handle. Terrain Blocks can be organized into larger Provences and/or Regions for other game play purposes, but I would prefer a uniform land system - until this open system could be perfected.

Maybe I'm misunderstanding the word open. I wouldn't choose that format if open meant units can pop up anywhere. To move through forest takes more time and energy so the preferred path (unless specifically ordered otherwise) should be to follow the road or rail network. Each of those road milepost has some reference in the map - a specific location. The miles/kms traveled relate to movement speed and That Factor makes it relative to all other units in a Time/Space relationship. The map just doesn't have a grid overlay to force position. That form of open has appeal but think of the tracking and processing to know where every unit is all the time (as each is traveling down or occupying somewhere along the road). Not only to calculate the load capacities of highways when forced to handle the traffic of 2 armors (one at 75% strength, one at 90% strength) 3 Infantry (all different strength), a HQ unit and the supply chain coming from the latest distribution point.

 

[claystone]

I agree, I would love to zoom in and see roads, lands, cities with my men fighting in.

I guess they can try to incorporate the same system "Risk" used where when it was time to fight you were zoomed into the battlefield and was able to control your troops. You where able to see roads, houses, trees, etc. Hide in Buildings with your men.

 

[dermeister]

Rome Total War uses a coordinate system based on pixels just like HOI3. It even has provinces, using the exact same system. The difference is in the number of pixels, and the positioning of divisions/units. In HOI3 they are assigned to provinces, and in RTW they have a position on the coordinate-system.

If you look at RTW data files, you'll see the similarities.

 

[unmerged(219522)]

The problem with the free map is the frontline issue where it would be extremely difficult to simulate an actual frontline, particularly so for the AI without some zone of control system for divisions.

I made a post earlier about an idea, but no one answered so I'll put it again but in more clear terms

How about a system where you keep the province system, but in each battle the province you attack can be broken down into smaller sections when you zoom in further, and the battles can then be fought at brigade level if you wish, adding a greater element of depth to the important battles. For standard battles you would still just right click onto a province and the AI would handle the battle (basically no different from now), but for important spearhead or city battles you could play the battle out by yourself in realtime with the rest of the game, keeping the same battle system as per now, but just "zoomed in" at brigade counter level rather than divisional. Due to the idea that most battles you could fight just as now (right clicking on the large provinces) with only major fights that you would need/want to get involved in, micromanaging hell wouldn't be created.

Basically I don't think it would be too difficult to do, as really it's virtually the same system as of now, just really a 2nd level of ai would need to be created...1st level to move divisions to the province (same as now), then the 2nd level of ai where to position brigades in each sub-province.

This would require 2x as much processing power for this improvement I think (more or less), yet if we're talking about HOI4 computers should be more than 2x as powerful anyway due to Moores Law

Do people think that this step would be a good thing for HOI4? The step from HOI2-3 was basically more provinces. I doubt more provinces in HOI4 would be feasible really, so sub-provinces like this seems a logical solution to continue the improvement of the game.

 

[Slyguy3129]

I can't count the ways in which both SP1 and 2 are/were broken whenever the last time I played them was, which to be fair, I don't even remember.

Both games had the very annoying habit of making it practically impossible to make a cohesive front since iirc there was no "area of control" in which it would force combat between two opposing units. Unless the two units are literally right on top of each other then they will just continue to glide past the lines. Atleast for me that was all that happened to me 100% of the time that I played. Which again, is why I can't remember the last time I played those games, or even had it installed on my PC.

Don't get me wrong I loved the concept, obviously enough to buy both the games. But the "War" part of the game just lacked to much to please me in any way, as I am more of an Armchair General, rather than a Armchair Secretary of Treasury.

 

[PirateJack]

This sounds somewhat similar to the Sins of a Solar Empire games (though they're obviously sci-fi rather than historical strategy), though they do have provinces. The appeal of the game is that you can zoom straight from the star system map right down to the individual ships within orbit around each planet and have strategic/tactical play all at once.

Still, it would be absolutely awesome to have a HoI-type game in that style.

 

[Gregorovitch]

I think the main advantage of using a province system is that it makes sophisticated AI programming tractable. It does so by limiting the number of factors, such as the relative positions of forces, the AI needs to consider enabling it to make more intelligent decisions.

Chess programs are so good precisely because of this: the limit of 64 "provinces" on a chess board and strict movement rules and win conditions make alpha-beta AI algorithms work well partnered with exhaustive possible move tree searches and position scoring systems.

HoI cannot match that because you have many more "provinces", many more "pieces" that can in principle move anywhere, and many more "opponents" each with different objectives. It's a scale issue. But it does a pretty good job considering.

But take away the province-based system for a free-for-all and the AI would be much less able to identify realistic objectives, calculate reasonable plans of execution and make reasonably accurate tactical moves, certainly within an acceptable time frame. You need the core province object to hang the algorithms off.

As computing power increases you can increase the number of provinces, i.e. make the game more fine grained, which is exactly what has happened between HoI2 and HoI3. In principle, with enough computing power, you could put a global hex grid underneath the province/region/nation system which would in theory enable tactical battles in the style of say Operational Art of War (and make the game play a bit more like the Total War series, but retaining HoI's superior AI). I don't think current PC's have anything like the power necessary for this in a game on the scale of HoI3.

With sufficient computing power and further developments in the orders/AI assistance system started in HoI3 it concievable that this idea would work gameplay-wise, but there is a fundamental problem. Good tactical/operational level hex based war games like Operational Art of War are unable to beat human players of any sort of skill with equal forces, and this is dispite the fact that on any given map/scenario the designer can script certain AI moves based on overall battle plan and conditions occuring to give the human a run for their money at least.

In HoI the whole point of the game is that (more or less) anything can happen so no tactical scripting like this can be done in advance, so an average human player would run rings round the AI in tactical battle. The system HoI actually adopts, base on limited front size and province boudaries, makes the AI, all things being equal, a better tactician than most humans because it can accurately calculate the optimum attack sequences etc better than a human can (because the limited number of variables make this tractable). That means in order to beat the AI you have to know what you are doing strategy-wise or you will get trounced - i.e. HoI is a challenging game (which is why we love it).

Take away the province system, stacking limitations etc that give the AI this tactical edge at the front and the game will become easier in that you would be able to beat superior forces through tactical nouce and trickery the AI is unable to match. Basically a bit like Total War games.

 

[DanSez]

... But the "War" part of the game just lacked to much to please me in any way, as I am more of an Armchair General, rather than a Armchair Secretary of Treasury.

A kindred spirit - send that man a beer.
(or whiskey if you please)

I'm reading an old book I found at 1/2Price Books - The Politics of the Prussian Army by Gordon A. Craig. A bit dry in tone but a lot of good information about the development of the General Staff and how narrow minded (in political scope) most of them were. Bismark being a noted exception but not the only one who realized the War is a tool of diplomacy, and yes in many aspects is driven by economics, but when I play a war game, I would like there to be more mayhem than management.