Announcement

**Mikael** · September 1, 1999, 13:24

You have some good points there.
However I contest one of them: why should a unit only be composed of 1000 soldiers? This could work well in the early stages of the game, but having for example the entire Wehrmacht made up of "1000 soldier" units would be impossible(Germany had some 3 million troops at their disposal on the Eastern Front in june 1941. If I remember my math lessons correctly this would be the equivalent of 3000 units...)
So why not have the player decide himself how many men there should be in a specific military unit?

**CormacMacArt** · September 1, 1999, 13:53

Hey! I did the best I could without any significant military knowledge! You have a good point. I picked 1000 more or less because of the Legion, but maybe having the player decide or possibly having the number increase over time would be better.

**Mark_Everson** · September 1, 1999, 20:16

Cormac:

Unit representation in a tactical system (which we Think we can do) is a nightmare with such a plan. I think that's the biggest problem with your general idea. That, and its one more interface to design, but we were thinking about a workshop for units anyway. The idea doesn't do much for me, but lets see what people think.

**CormacMacArt** · September 3, 1999, 09:06

I guess I don't understand what you mean by a "tactical" system. Sorry.

**Mark_Everson** · September 4, 1999, 00:37

In a 'tactical' system the battles can be a game-within-a-game. So instead of Red army meets Blue in square x and the next thing you hear is the result, a tactical system allows the player to actually fight out the battle with the armies broken down into smaller units. The smaller units can manuver to try and obtain advantage using their individual characteristics. This is more important in an army-based system like clash than in a dispersed unit-based system like civs. In Clash there should be fewer actual fights than in Civ so we can allow them to (at player option) be handled in more detail than civ.

Now this tactical system would still be fairly simple, something like 25 mini-squares with a relatively small number of phases to finish the whole engagement. Do a search under tactical if you want to see a lot more detailed discussion.

**Paul Crocker** · September 5, 1999, 16:05

It is my understanding that we are trying to imitate reality as closely as possible while maintaining a fun and playable game. If so, then CormacMacArt makes some good points here, and I think that they deserve a little more thought before dismissing them, and I think that Mark has unwittingly made the point for incorporating some of them into the strategic military model. Since I am most familiar with modern (1860 to present) militaries, I'll restrict most of my comments to that realm, but I'm pretty certain that most points I'll make can be applied to ancient and medieval military units as well.

My biggest beef with most (if not all) of the sims out there is the way in which they like to explicitly label units (eg tank, fighter, infantry, etc). In reality, units are not entirely one type or another, but rather a mixture of several types (except at very small scales, such as at the company level and below). I'll give an example from my personal experience:

I was stationed with a US heavy armored division in Germany in the early 90's. My brigade was designated as an armored brigade, but it had much more than just tanks. It was composed of 2 tank battalions (BN), 1 Mech infantry BN, 1 engineer BN, 1 support BN, 1 scout company, and some air and artillery assets that varied in size depending on whether they were assigned directly to the brigade or were attached at the divisional level. In addition, each combat BN had its own organic support, mortar, medical, and engineering units. In garrison, the units (at BN level) were usually "pure" (ie all tank, infantry, etc), but in training or battle, we were mixed together in teams. At our battalion level, that worked as follows: My tank BN swapped a tank company with the Mech infantry BN in exchange for an infantry company. In turn, the infantry company was broken down into several platoons, each of which was assigned to a tank company. This total integration is typical in modern armies, however the specific allocation of units varies greatly.

An armored cavalry regiment (ACR) has much more artillery and air support than a tank unit of comparable size. The higher generals simply felt that for an ACR to perform its duties (recon in force and the use of delay tactics), it was necessary to change the force structure. Currently, the army is contemplating eliminating some of the heavy divisional structures (such as my unit that I described) in favor of lighter, more mobile units that are more suited to fighting in urban/peace keeping operations.

The point that I'm trying to make is that by using a system proposed by CormacMacArt, a player can modify a unit's structure (when building it perhaps) to suit that player's needs. That would mean that all tank divisions in the game are not created equal: The player would learn that while his tank units could easily whip those of "Joe Schmoe", the tank units of the "Nasty Buggers" are too strong in artillery for him to tackle with the same sized attacking force, and he therefore requires a different strategy in facing the "Buggers" - maybe either a change in his own force structure, or the use of additional units.

THIS more accurately reflects reality (at the strategic level) than simply having a tank unit, a fighter unit, etc. The player could simply say, build "tank" units with the following composition: 40% tank, 20% infantry, 20% artillery, 10% engineer, 10% support. The unit size is arbitary, but this could also be a setting under the control of the player...they could build 10 battalions of 600 men if they wanted, or simply build a single brigade/ small division with 6000 men. We would have to set limits to the proportions in a unit and to its size.

At the tactical level, this system shouldn't be too hard to incorporate, since ANY system that we do at the tactical level will be a HUGE abstraction unless you really want to build another game inside Clash. You can't really imitate a portion of a battle within a mini-world of 25 cells with total disregard to what's going on in the cells surrounding those 25 cells. What if the enemy had a huge force just outside of the battle within those 25 cells? In reality, you would hardly commit to a battle without some consideration to the flanking units.

Nonetheless, a simple tactical system like Mark proposed would both work and be fun. Simply give the player a number of units in proportion to the force mix of the parent unit (30% tank, 50% infantry, 20% artillery becomes 3 tank units, 5 infantry units, 2 artillery units at the tactical level) and let them fight it out in the tactical battlefield. Not very realistic, but neither is the present alternative that's been proposed in earlier threads.

Just some thought here from someone with a little bit of actual military experience for anyone that's interested You can take it for what it's worth - I think that Clash will be great using either model (the existing model or this proposed model). It's just a question of accuracy. Any thoughts, comments, cheap shots?

Paul

**Paul Crocker** · September 5, 1999, 16:16

Bump (it didn't update the date on after my last posting).

**Mark_Everson** · September 5, 1999, 19:29

Paul, Cormac:

I agree that the "flexible unit structure" idea is both interesting, and more realistic. Other than the visual objections to that which I raised above, my main concerns with regard to flexible unit structure are in the area of AI. The AI's job is already tough enough, I feel, without having to worry about varying base unit composition too. For instance, has the enemy just switched the composition of its tank units in a way that will invalidate my calculations of their combat power? And I know a lot of people who would play games with the unit structure just to try and "break" the AI's thought processes.

Then there is the problem of run-away use of memory as each civ now needs to keep track of say, 30 unit types for every other civ it meets. I get 30 by assuming there are roughly 15 unit types in any historical period, and there may be an old and new type of each unit currently on the field. So instead of, say, 100 unit types spanning the game there could be more like 10,000.

All that said, it is a cool idea. I think it could work provided:
That we keep to a prescribed number of unit icons, and that people will just have to pick which one they want to represent their flexible unit.
That there is sufficient demand for it to make it worth all the extra AI tweaks and memory usage that would be required.

The proposal Would help to make technology better integrated into the military system also. That's IMO because a fixed number of units forces a certain lumpiness into the system. With the flexible unit design, one could distinguish, for example, the evolution from Spitfires to Mustangs in World War II fighters gracefully. I don't think many people would object too much to having those two types of fighters represented by the same fighter icon. The capabilities of the flexible fighter unit would just improve as the composition were switched from the one to the other

I think all that would be required to make sure people can't fool the AI, is to give each modified type of unit a unique identifier. So even if they were named the same, and used the same icon, the AI could tell the difference.

Well, I think you're bringing me around to the idea . This is exactly what happened with the tactical combat feature by the way. I was initially against it, but some strong arguments for it eventually convinced me that we should definitely have it.

Let's see what other people think. I don't know what Druid will say about it. Also, every additional bit of complexity just increases the lag time until we get Clash out... But I think for now I'm favorably disposed towards the idea.

**Paul Crocker** · September 6, 1999, 13:01

You bring up some good points, Mark, about the limitations of the "flexible" unit proposal, and I had similar reservations as well. Your solution of a single icon representing a unit type is exactly what I had in mind (real-life militaries do the same thing: armor symbols in the US Army are the same regardless of whether they represent M1A1's or T72's).

The problems confronting the AI are obviously more serious, but I think that there is a relatively simple solution to that. If the AI is comparing parameters to base its decisions on (eg attack, defense, firepower, maneuverability, etc), those parameters are modified by the force structure of the unit. The AI wouldn't need to compare 20 tank units for each player, but rather just compare the strengths and weakness against each other. FE Nasty Buggers' defense value in cell [][] is too high so don't attack (the AI doesn't need to consider WHAT is making the defense value so high, just that it is too high to attack it with the current units at its disposal.)

FE a mech infantry unit (70% infantry, 30% tanks) would have a lower attack value and higher defense value than a tank unit (70% tanks, 30% infantry). It shouldn't be much different than what the system currently calls for (unless it has been really revamped since I last looked at it a month ago). The only difference is that the flexible unit proposal allows the players to tweak the capabilities of their units and give each country a unique flavor. Your description of smoother transitions in technology is also right on the money and more realistic.

Regarding players that will fool the AI - you're going to have players do that regardless of what you build into the program IMHO. Those players aren't as interested in how closely the sim reflects reality anyway, so I'm not sure if we really need to worry about them. It seems that the allure (sp?) of Clash will be the level of detail and historical accuracy, and so it should attract a lot of players that are more interested in playing "fairly" (or I might be a bit naive!) Maybe THOSE players should be the target audience, at least moreso than the "cheat" crowd. There are plenty of "cheats" available out there for other games, but I refuse to use them, since they take away from the fun and challenge of the game.

Any other opinions on this proposal out there? I realize that it would mean another interface (although we had already planned on a "workshop" module anyway) and may mean a little more programming, but by what factor would it really increase the work? It seems like this method is merely "different" and not necessarily more complicated (just appears that way at first glance). It would certainly put Clash in a class of its own, since I've NEVER seen a sim with this much flexibility.

**Mark_Everson** · September 6, 1999, 13:28

Paul:

Just one comment. In your piece above you're assuming that a player should Know what another player's unit composition is (this is the assumption behind just being able to calculate attack/defence of the unit). I was assuming the AI would have to Figure it Out. I think your approach is much better for the game, although less realistic, and does remove the power estimation problem trivially. Then there would be no way for the players to "screw with the AIs brains" in this regard anyway.

**Druid2** · September 7, 1999, 15:24

The flexible "unit" make-up idea is what got us started on the "Task Force" concept in the first place.

The currently proposed model allows the player to define his "basic unit" [which we call a "Task Force"] in any way he chooses.. with whatever components and types available to him. He "builds" the components [which we call "units"] individually, then combines them into TF's.

Thus, he could build 2 TANK unit, a MechInf unit, and a Engineer unit all in different places; ship them to an assembly point and create a Tank Army Group [land TF].

Further, each TF may have a different makeup. It may change dynamically, in the field, or in a base camp, by simply adding new components, or moving out old components.

So... this is a good idea, which we have "sorta" built in. Perhaps it was the difference in nomenclature that confused things. I profess absolutely NO special knowledge of the right names for things military.

**Paul Crocker** · September 7, 1999, 17:32

When you explain it that way...
I agree that using a TF is probably the best compromise that we'll get between accuracy and playability. Getting the nomenclature right is probably important to reflect that accuracy. FE I wouldn't go calling the individual "tank" units a division, corps or army. Battalion or "team" (something that the US Army experimented with during the cold war - it was usually company to battalion sized. That's 100-600 men for those interested.)

You might want to consider placing limits to the min and max size of TF's (if you haven't already) to reflect reality. You'd never see a TF (division) with 40,000 men in it! Or create penalties in unit efficiency for REALLY large TF's. Another idea is to say that a TF with x number of units is a division, y number of units is a corps, etc. That would solve all of the problems/misunderstandings with nomenclature (I realized that we're not at that point yet!)

**Mark_Everson** · September 7, 1999, 20:48

Paul:

Actually size limitations of TFs will come naturally from required unit dispersion, supply limitations, command and control capabilities of the age and other factors.

But I Was talking about the individual units that Make Up the TFs as being of flexible composition, its just a bit much to bite off at the start, but we Could try it at some point...

Thanks for pulling me back down to Earth Druid. I'm easily seduced by over-complicated models

[This message has been edited by Mark_Everson (edited September 07, 1999).]

**Illmuri** · September 18, 1999, 23:52

This all sounds good but will I have to read 5 chapters on how to do this stuff? How long will this take just to make all you TFs and add/remove components? When I have large large armies? It will take forever! And thats just the military part. Sounds like a lot of micromanagement to me.

At least make it so you can have some predefined TFs. For example: I make a TF w/ a bunch of tanks and a few supporting inf (and little bits of other stuff). Instead of making 10 of these over and over, I decide I might use this one in the future (as I dont want to come up w/ something new each time). Maybe have a place where I can save this paticular makeup (maybe a max of 5 different ones to save space etc). Then I can just say make TF1 or something. Your ieas are good though, I was just addressing the problem of time it took to play the game.

------------------
-=Illmuri=-

Announcement

Military Units

Military Units

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment