The Future Of Complex Wargames Looks Bleak

[berto]

ORIGINAL: ComradeP

... although I guess WitP:AE covers the operational level fairly well due to there being far fewer divisions involved than in most Western/Eastern Front games, even though it's a strategic level game ...

[X(]

WITP:AE models every LCU (land combat unit, down to battalions and even companies), ship (as small as midget subs, cargo ships, PT boats, etc.), air squadron, and pilot that fought the Pacific War. For every combatant nation involved. Thousands of units and distinct combat entities, far more than any Western/Eastern Front games, WITE included.

As a Dutchman, the battles the Dutch fought in May 1940 are covered reasonably well as a sideshow in some wargames about Fall Gelb, but the battles that were fought in the Dutch East Indies are not.

I am extremely interested in PTO sideshows. A year or two ago, one of the WITP:AE devs released a monster DEI scenario, subsequently pulled. It's planned for a comeback. I can't wait!

[wodin]

Here is another novel idea..no AI..no sale. An extremely selfish view point in my opinion. Single players..oh s#d em. The whole reason I loved computer games from being a kid is because I didn't need to find anyone else to play against..looking at my collection of boardgames that where never played the computer was a godsend. Command Ops show sit can be done..but I think certain developers these days border on being Lazy..hence so many multiplayer and MMO games..

ORIGINAL: CarnageINC

ORIGINAL: Iain McNeil

On the issue of why are we not using PC performance to make bigger/detailed games - The PC has far more processing power than AI would actually need if programmed well. The performance really only comes in to play The issue is that creating a good AI is something that is specific to a game and its set of rules and a huge huge task that gets exponentially more complex as the game gets more complex so doing it for the complex games takes a lot of time and money. The AI has to be scripted to some extent in complex games as its impossible to really evaluate the rules and game situation. Even chess AI can beaten by a human and that game is extremely simple in terms of rules and scope. You cannot reuse the good AI from one game to the next so you start from scratch on the next game as it has different rules. There are sometimes elements you can reuse but it doesn't save much time. The other issue is the AI has to evolve as the game evolves. Lets say for example you add a small feature near the end of development - this could completely break the AI unless it properly accounts for it. Lets say you re-balance some stats - again this could break the AI.

Here is a novel idea, why not make PC war games without a AI. Do something similar to vassal but have the computer keep all the record keeping of what ever it needs, just like a normal pbem game. I know you kill off some of the market for solitary players but if they want to play they can do what regular board gamers have done for generations, play hot seat both sides.

This way you can make very complex games without worrying about how a AI will react. Has it been done? Shouldn't it be tried?

[Kuokkanen]

ORIGINAL: wodin

Iain..why aren't wargames then taking all that PC power and doing something with it. Yes tablets can do this and that but they are still no where nears as powerful as a PC.

Surely something innovative could come along and reshape how we see grog like wargames by using the power of the modern day PC. I'm not talking 3D RTS style or anything here but may really beautiful 3D maps or 3D Overlays..I'm not sure what can be done to be honest to use the power..but surely something can be..surely we haven't gone as far as we can do when it comes to the more Grog like wargames?

I like Settlers 7. But then again, I'm not grognard who'd enjoy WitP...

[warspite1]

ORIGINAL: Perturabo

ORIGINAL: Punk Reaper

ORIGINAL: warspite1

It may be "silly" to you, but frankly, I much prefer a game that involves the British as this gives me more interest in the subject matter. Yes, I have played games with no direct British involvement - WITE for example - but this is not my preferred game type.

I'll second that.

That's why we can't have good things.

warspite1

Eh?

[Mobeer]

A few scattered thoughts:

Any game I buy has to have an AI. Ideally the AI is competent if not clever. The AI should be free from making stupid mistakes, but need not be super clever. Have the game multi-threaded so the AI figures out it's moves while I work out mine.

I'm sure I did read somewhere about commanding 100 units in real life - maybe a Japanese river crossing in 1904 when the Russian general commanded all his battalions directly (and not very well)?



As for epic (computer) war games:

1 - Complexity needs careful pruning. I don't see why as commander of the entire Allied war effort, I need to tell my squadrons to fly at X feet against the Japanese. Let's either set a high level doctrine, or just assume one. Depth does not come from complexity alone and quality and complexity are trade-offs (along with cost).

2 - Graphics are important; they have to be functional and pretty is better. But graphics alone don't make a game. I like a chess game with an elegant 2D design better than an ugly 2D design, but also like a adequate 2D design more than pretty 3D. I don't think that expensive graphics are needed for a epic wargame, but this is not an excuse for horrible graphics.

3 - They should be easy to pick up or put aside so that they can fit into a busy lifestyle. Requiring constant attention is a pain, so having tools to keep track of what is happening helps.

4 - Following 3: They need to be rewarding in some sense within a short period. I would like to be able to complete a turn in 15 minutes, so if I have an hour, I can have 4 turns. Bombing the Reich puts me off completely by seeming to need an hour for 1 turn.

5 - Taking years in development is no good. There needs to be more frequent versions to get earlier revenue. And maybe downloadable content like editors, scenarios, add-on tools.

6 - Whatever happened to marketing?

[Jamm]

ORIGINAL: PipFromSlitherine

Super-high-quality graphics are very expensive to produce, and yet generally will not give you a huge boost to the sales of a hardcore wargame (there does, in fact, appear to be a % of hardcore gamers that consider good visuals to be a sign that the game can't be very good... [;)]).

Cheers

Pip


[/quote]

I myself think graphics are the easiest part of a wargame to produce and cannot understand why so many are uninspired.
The game has to be good, but if it's butt ugly, I'm not buying it.

[ComradeP]

WITP:AE models every LCU (land combat unit, down to battalions and even companies), ship (as small as midget subs, cargo ships, PT boats, etc.), air squadron, and pilot that fought the Pacific War. For every combatant nation involved. Thousands of units and distinct combat entities, far more than any Western/Eastern Front games, WITE included.

My point was probably too vague: what I meant was that even though WitP:AE has all the units, it doesn't have an operational level map scale and the land combat model that is also somewhat more abstract than you would find in the average operational level game even though ground elements and/or men are counted as casualties (instead of casualties being abstracted into either some sort of step system or losses of companies/battalions).

One of the more surprising features of WitP:AE, to me, is the more operational level than strategic level time scale per turn. WitP:AE is, however, at its core not an operational level land combat wargame, even though it has an effective land combat component.

[IainMcNeil]

On the point about having no AI its really about sales.

The vast majority of people want a single player experience. That's the reality and so we have to provide it to them.

You can design games around the idea of multiplayer, but they need to be different in design and structure and tend to be lighter. E.g. games like Hero Academy.

[pzgndr]

ORIGINAL: Arjuna

With no disrespect catwhoorg, scripting is never going to produce good AI. Scripting as its usually referred to is scenario specific. Even with conditional scripting once you have played it a few times you will know that it will do X or Y but never Z and so it can be easily beat. If you want good AI you have to move to a generic system where the AI controlled forces are situationally aware, can adapt to what they find happening and can develop a plan to suit.

Never? I disagree. The Strategic Command series has matured to the point where the event and AI scripting is very powerful. There is sufficient variability to provide uncertainty. Even when I play my Advanced Third Reich mod with the scripting I wrote, I am still surprised by what the AI does. There are still AI weaknesses to be sure, as if every human player is infallible (sure!), but the AI is still very challenging and not easily beat on higher difficulty levels.

I will say that having been through this exercise in developing my mod and scripting the AI over the past several years (yes, years), it does take considerable time to create numerous (hundreds) clever and nefarious scripted strategies, and verify them with many many AI vs AI games to see what works well and what doesn't. So, I fully understand how time consuming it all is just for text-based scripting, even without code programming, but it is possible in the long run.

I raised the issue in another thread about the status of old games and how we can encourage developers to continue to improve and enhance games well after release, for additional features and better AI. Seems like the model is for a developer to create a game on his own dime, then make some sales, and then... move on. And that's life. But we can still go out and buy a '64 Mustang and refurbish it and extend its life quite a bit. How can we do that with old games? If there's a hobbyist with an interest in updating an old code and enough determination to improve it for the community, then it would be nice to find a way to allow that to happen. Because in reality it probably will be a hobby effort, a labor of love, cuz there's not a lot of opportunity for financial gain beyond a certain point.

[IainMcNeil]

I completely agree on the scripting - I would be surprised if there is any complex game with a competent AI that is not at least partially scripted.

[Missouri_Rebel]

Ian, do you guys own the rights to Vassal?

[Curtis Lemay]

My feeling is that AI is a gigantic black hole, down which an almost unlimited amount of coding time can be poured without any discernable benefit once you get past a rudimentary level. Usually that coding time can be better spent on actual game features for more benefit.

You can't just think about this as "What do I want, regardless of price?" (Would you buy a car on that basis?). There has to be a cost/benefit consideration. You have to ask yourself "What can I afford?". So, you might really like solid-gold hub caps on that car. But if they cost more than the rest of the car you'll probably pass.

As for the future of complex wargames, I don't see how it can go in any other direction. The simple, unrealistic, ones have already been made. It's just like for flight simulators - why build less realistic flight simulators? They're already available. The only reason for a new flight simulator is to push the envelope further towards realism. And when you get that ultimate flight sim, do you really want it to make anyone that uses it feel like they're Richard Bong, even if they haven't bothered to grasp any aspect of flight or those instrument thingys?

Wargames are war simulators. The more realistic the better.

[wodin]

I always see people bemoaning AI's..yet many games I've had a hard time against the AI..maybe I'm fortunate to be a useless player. Rarely the AI is awful..but as I said alot of games I find it easy good enough.

[wodin]

ORIGINAL: Missouri_Rebel

Ian, do you guys own the rights to Vassal?

I very much doubt that.

[catwhoorg]

Commander - the Great War is a good example of a challenging AI in a recent game

Its not as inventive as a human, but I certainly provides an interesting enough Single player game

[IainMcNeil]

We have no connection with Vassal.

[Missouri_Rebel]

ORIGINAL: Iain McNeil

We have no connection with Vassal.

O'K.. I thought for a time there was a forum here for Vassal pre Slitherine days. I must have been mistaken.

Thanks.

[Arjuna]

ORIGINAL: Iain McNeil

I completely agree on the scripting - I would be surprised if there is any complex game with a competent AI that is not at least partially scripted.

Well you do not what you are talking about I'm afraid to say. Our Command ops is not scripted and I think most would agree that it fills the buill as being a complex game.

[wodin]

Command Ops is a shining example of a great AI..far more so than anything scripted.

ORIGINAL: Arjuna

ORIGINAL: Iain McNeil

I completely agree on the scripting - I would be surprised if there is any complex game with a competent AI that is not at least partially scripted.

Well you do not what you are talking about I'm afraid to say. Our Command ops is not scripted and I think most would agree that it fills the buill as being a complex game.

[Lieste]

I suppose it depends on exactly what is meant with scripting ~ in that the AI is seeded with 'objectives' and a timetable, CmdOps has a very basic script to follow ~ but it does this in a much more organic way than is typical, and importantly except for the very highest level of 'grandscheme' planning it follows practically identical methods and rules for player and AI activities.

(It is similar to the debate of probabilistic and hitpoint methods for tactical game vulnerabilities ~ ultimately both are just points on a spectrum (100% chance of some fractional value of total hitpoints, versus some fractional chance of a complete damage to a single "hit-point")... that there is a possibility of gaining something from cautious blending of the two approaches is something sometimes vehemently opposed by the staunch defenders of each...)