I think I will test this out tonight with varying numbers. 16, 24, 32, etc., to see the hit on performance and what Raven means by:
Us modders can't do it easily because it causes issues with units appearing in groups, but you guys at Stardock have the animation tools to fix that.
My initial thought that the tiles could be made bigger (by Stardock) and/or the units could be made smaller (by modeling/modding). I won't know how they are spaced out until I try it, though. But from my experience, Raven's assessment that this can be fixed by Stardock is very likely accurate, but it may not take animation modifications to do this. Simply better organization of troop formations may be enough to resolve it. I don't know, but it's definitely resolvable. The <PewterStandModel>, specifically the model within, may be the key.
As long as you're running on a decent system and not a Notebook performance shouldn't an issue unless you push it to INSANE levels, like 4 or 5 thousand, That you might take a hit to performance. According to Frogboy there's no engine limitations behind the choice for smaller unit counts.
The animation parts I was talking about would be if they kept the turn based system and put in more units, it would look best with interim animations between turns. Which they could do easily. That isn't what's stopping us from just changing the XML here. What I meant by it's harder for us is the unit spacing. When you go above..umm. can't remember if it's 20 or 25 men in a unit, they don't look right within the tile. They aren't spaced evenly and the formation looks odd. That's the part I don't know if we can fix easily when we go to re-balance it for the mods that will be using epic numbers. I'm not sure if unit spacing in e formation is handled in the XML or in Python that we don't have access to yet.
The DL Team and I have been going through the XML making notes of where things are, and I've already laid out the basis for a standard that multiplies across the board for keeping the math balanced at higher levels. When I get this more fleshed out it's going to go up for discussion in the modding forums as there are so many mods that want to use a Epic Scale. My goal there is to make a "Epic Standard Model" that All the mods that want epic scale can use as a basis and then tweak from there as needed for their mod's tastes. We also plan on releasing it as a stand alone mod by its-self just for those who want epic scale in the vanilla game. Mainly though it's not knowing how to get the spacing of the groups right with just XML that's kept us from just jumping up the numbers in the XML you've shown there as a start.
If we can get the spacing right with XML alone, then awesome, but I'm thinking it's going to be something we either need API's access to make it look right on the field, or access to Stardock's animation plugins so we can remake a until animation tile and set our own lines and formations, but that's more in the modeling department which I'm not knowledgeable on in the slightest.
To sum that up minus the confusion, we can add in larger numbers and change the math to make it balanced, but without either the knowledge to space out the formations in XML or whatever other code is needed to do the spacing, it would look rather dumb. We haven't played around with it much as there's a lot of other parts of the mod we're working on while we're waiting on the rest of the tools, plugins, and API's from Stardock. So for now it's a bridge we'll cross when we come to it and have the right tools for the job.
1) Exactly, how are you going to expand the map to accommodate additional units without causing game crashes?
2) Exactly, how are you going to add new larger 50-man squads (or whatever they're going to be) without causing the problems listed at * below?
3) Exactly, how are you going to re-balance the AI so that it copes with the newly scaled "epic" size and not become a "steam roll-over" instead of the "walkover" it is now?
4) Exactly, how are you going to redo the tactical battles, so that they don't degrade into 8-hr slugfest as we try to whittle 1000 hp "epic" units down one hp at a time?
5) Exactly, how are you going to pro-actively prevent system and memory crashes with the new "epic size"?
* Problems that I can already predict will happen with new 50-man superman squads:
{Would a 50-man squad have the attack strength and defense of a 10 man squad, but the hit points of a 50 man squad? If so, a typical mace group would be Attack: 52 Defense: 24 HP: 1000; Based on one of my auto generated unit types. If this were true, then the actual game effect of this would be to create very long tactical battles as these units would take a long time to kill, but at least they're not "superman" squads. Frankly, the only thing players would get out of this (in a turn based system) is tactical battles that take about 4x longer on manual, no problem if using auto resolve, which is probably what I'd be doing all the time if this unit were added because it would probably take 30 minutes to resolve one battle if there were 8 of these 50-man squads on each side.
The problem with "epic" battles is implementing it practically in a turn-based system in such a way that:
6) The game doesn't’t become more unbalanced than it currently is: i.e. the AI needs to adequately "cope" with the new units.
7) Avoiding the creation of 50-man "superman" squads that are basically game-breaking automatic-game winning "nukes" and which turn the game into nothing more than a tech race as to who can get this technology first and thus win the game.
8) Avoiding evolving tactical battles into long drawn-out boring affairs as we try to whittle units with 1000 hps down 1 hp at a time.
Ok, I'll try to break this down as best I can for you my friend. Keep in mind though I'm VERY tired and wasn't planning on making any more posts tonight as I'm falling asleep at the keyboard. I might not cover every number but just boil it down to the short and simple version for now. Also, part of your understanding on this is going to be based on what you know of the development process up till now. If you don't know the details of what the devs have been talking about for the last two years some of this might seem odd as it goes against what we have in game now, but was still planned by the devs, anyway, here goes:
1. The Map, both Strategic and Tactical can be custom made in the XML very easily. Frogboy sat down with me over IM and taught the modding community how to make Extra Large Maps and I then replayed the conversation to everyone else. It's in one of the stickied lists on the mod forums if you'd like to go take a look.
2. Adding the unit sizes can be done in XML very easily, it's the spacing of the units that looks funky when you do it that way. As of current, we're not sure if getting access to the Python API's will be able to fix this, or if we'll be able to fix it with the animation plugins (animations might not be related here). Once we get the spacing issue solved the rest all falls into place quite nicely.
3. The Ai doesn't need to be changed at all luckily. The AI will fight the same and use the same mechanics as it does now. With the upcoming patch the mechanics here are already being adjusted so that if there are 10 men in a unit, each man gets his own attack swing instead of them being pooled like they are now. This will fix a lot of the minor issues with tactical combat now actually so that's a good thing. The AI will still know how to move the units and everything else it does now, it doesn't really care about unit numbers. Frogboy is the AI programmer for the game, so, the smarter he makes the AI as time passes the better the game will be for all of us. It doesn't matter if the AI is moving 20 men, or 200. A tile full of men is a tile full of men. Having more or less men only changes the calculations meaning it will change the "pace" of a battle. It's quantitative across the board.
4. Battles will still be just as balanced as they are now. A unit of 100 men, fighting another unit of 100 men, will have a battle that lasts just as long as an equal battle of 10 vs 10 would now. More men swinging means more damage done and this applies for both sides. The battles at the beginning of the game won't change so the pacing of the game's battles will stay the same as it is now until you get to Late Game when you have Large Cities that can support bigger armies. Without a large population (because Elemental makes it's units directly out of the cities population) then large armies aren't possible. The population of a level 5 city in game now is a couple thousand people I believe. Honestly I didn't pay super close attention but you can see the number when you hover your mouse over the red bar in your city window. I know a level 3 city has a population of 500 according to my city window. If a level 5 city holds 2,000 people, then in times of war when a draft is instituted I can see 1/4 of the population being drafted to defend the city, even if just the men are taken.
5. We won't be preventing memory crashed, that's on Stardock. They should be fixing the memory errors and crashes as those are the Number One bug fixes being done first to my knowledge. When the memory errors are fixed by them then memory should have no bearing on our unit counts. We won't push them so high as to bog down people's machines though, that Is being taken into account and before the DL Mod ever goes live it will be tested, re-tested, and then tested some more to make sure we don't put out a buggy release. As with any game of this nature, how many units your machine can display without causing lag is based on your Ram and Proc speed and graphics card. Our running requirements should be just the same as the normal game's running requirements unless of course the truth was stretched when were told just what the engine can and can't handle. As long as there's no engine mis-haps then at the very least the epic scale mods will only have "Slightly" higher running requirements, but it shouldn't be a drastic difference as long as it's a quality engine (which a lot of people are still doubting honestly, but, everyone needs to realize Stardock made a Brand New engine here and it's going to need time to mature).
6. As I said above, the Ai won't have a problem moving more units around. It's the same mechanics that's used now for the units. Nothing changes but the unit count at later parts of the game. It would be nice if the API's for AI programming are exposed by then so we can write custom AI's aside from just the minor tweaks we can make to them now in XML. The AI Can be tweaked with the XML to give it higher values on some things, which changes the way it "thinks" in game, and this could come in very handy making custom faction AI's for any mod. The units moving and fighting and everything else they do won't change though, it'll just take a minute or two longer to fight out the late game battles depending on how big a force you and your enemy have. As has been done in the base game though, if you don't feel like fighting any of the battles, you can hit "auto calculate" at any time you want. Making sure that was covered was brought up extensively by the carebears who don't like fighting tactical battles during beta. Frogboy made it a Specific Point that the player could end battle any time they wanted to and let the computer calculate the rest of the battle. Personally, I like to fight every battle when I'm playing the game. Even the no loose over-powering battles where you know for sure the AI will win on your behalf, because no matter how smart the AI is, I'm Always going to be smarter than it is, as is pretty much any actual human player. AI's can be great, but there hasn't been one made yet that can come close to making the random decisions a human player can and does make during the coarse of a game.
7. This one is touchy, here's why. In ANY GAME, there is ALWAYS going to be a combination of units, or an end game unit, that's always going to be the best of the best that a faction can make. This is true of ANY war strategy game. The key here is to make all the forces of any faction in the game have a balanced force that it "can" make that can counter anything their enemy can throw at them. It's just a matter of if the player is good enough to keep up with the AI in tech research and resource gathering so that when the time comes the player (or AI) can make a unit of the best fighting force it can make. Whether that's a 10 man unit, a 50 man unit, or even a hundred man unit, as long as it's the strongest that can be made (best weapon, best armor etc etc) there's Always going to be a "Super Stack" of "Something" that can wipe out anything else. This is true n the base game now and in every game of this type I've ever played. The best way to avoid it is to be prepared for it when you play. When you play a game now you know that eventually, if you let the AI live long enough, it's going to be cranking out units with War Hammers wearing Magical Plate Mail, so you need to be ready to make the same thing, or have a couple dragons, or whatever your winning strategy is when going up against powerful units. Increasing the unit counts won't have anything at all to do with the "stack of doom" problem that Every game faces today. In any game there's Always going to be a unit, or type of creature, or whatever, that's at the end of the tech tree and is the top of the line unit you can make. That's part of the strategy of the game.
8. Battles won't be any longer or shorter than they are now depending on what strategy you use during the game. 1,000 men vs 1,000 men would be equal in time to ten times as long as ten vs ten. You get some pretty long battles in game now when you have to maxed out end game armies with the best units and most powerful monsters you have in game. At the most, even if we maxed it out with MASSIVE two thousand man armies on each side, it would be Substantially Shorter than say, a game of Total War, or a game of Starcraft 2. I'd guesstimate (note thats a guess until balance is refined more) that an End Game Epic battle wouldn't or shouldn't take any longer then 5 to 10 minutes tops. Also though, as is possible now, the player can hit auto calculate Any Time they wish. You have to understand though that at least "part" of the motivation behind this is there are a lot of people who do want longer battles. When you have larger armies battles are going to take a little bit longer, but, that's all part of the epic feel. If I played a massive game that took me a week in real life to get done, I'd probably feel a little ripped off if the massive last battle of the game between me and my most hated enemy only took 3 minutes. When it's the Final Battle, it Should Feel Climactic and Epic, like the final battle in LotR, or the battle for Minas Tirith. These were great epic battles that should be meant to take a little bit longer than a standard run of the mill every day battle. You have to be careful here not to over-exaggerate the fear of how long this would extend the length of a battle as part of the desire here is to do exactly that.
When it comes to balancing the other numbers involved there are a lot of other numbers to change to make this happen, but, believe it or not that's a relatively Easy Part. The numbers and math we have in game now is ridiculously simple. I mean, a farm or other food resource normally gives a bonus of 1 to 4, or like other resources only adds 25% of something. These numbers are pre-school small and very simple to increase. I don't see why the numbers weren't made bigger here to being with, nor do many others. It doesn't make a lot of sense to me that a town of 500 people (lvl 3) would have structure that only added 1 or 2 of something and 25% to the rate of something else. To me, a much better system would have been one that works in the hundreds. If a town has 500 people in it, then there needs to be enough food in the town to feed the population, so why not have the food structure make 500 food and have each person in town consume 1 food? To me that makes a lot more sense then the boiled down child's play mechanics we have now.
The people in the community who are upset about the small numbers Aren't just talking about the numbers of soldiers. Many of the people are upset over the fact that Every system in game, from unit counts, to the economics, to the damages of weapons and defense ratings of armors...all of these systems have been given small base numbers to work with and a lot of the other systems that Should have been here from the start to add depth and complexity, like more variation on damage types, resistances, skills and abilities that progress as a character levels up, city walls to lay siege too in tactical combat, none of these have been added yet and all we have are these base systems where you can count the math out with your fingers if you wanted. That's one of the things so many people were pissed about at launch. We were told it was ready and by being told that everyone assumed that the other systems that were talked about, like all the ideas for spells and damage types and magical items, everything really..we assumed it was all in game and working because it was "ready", but then all of that was missing. The Beta testers saw the small numbers in the early builds, and when we brought them up and tried to talk about them and asked what the final balance would be, we were usually told the numbers in the beta were mostly "placeholders" and Did Not represent the numbers that would be in the final game at launch...and yet when launch came we were staring at the same low numbers that were supposed to have been temporary. I'm going to cut this paragraph off here as it's starting to stray off of your original questions, but, hopefully this will give you a little insight into the mechanics we have in place now and why they are so different from what everyone expected.
So, to cut that short, re-balancing the numbers to make larger armies ad more complex systems work at Late Game stages really is child's play. Any of us with experience playing strategy games can sit down with a piece of paper and a pencil and work out a complex and balanced system for resources and food and economics in one afternoon. It would actually take longer to track down the variables in the XML and replace them than it would to come up with a more complex system. I hope that explains a bit what you were looking for. I know some of that may sound complicated, but once you work it out on paper and have the tools you need to make the changes it really isn't all that hard or complicated.