In Development

[sPzAbt653]

Solitaire vs. Multi-Player
There was a poll here if you hadn't seen it:
tm.asp?m=4106483

In post #32 of that thread I give the results of another type of 'AI vs. Human' poll.

I saw a video a couple years ago by the developers of Civ where they stated that they believe that up to 95% of their games were played solitaire.

I am a strictly solitaire player but have no animosity towards mult-player folks. I think most of us are realistic in knowing that an 'AI' for any wargame isn't an economically sound venture, but it seems to me that not providing a decent Programmed Opponent would have a negative impact on sales.

[PipFromSlitherine]

The online population for >>most<< games is going to be in the 5-15% range. AIs are unlikely to be going anywhere, but as Iain says, every new rule requires new AI code to deal with it.

Cheers

Pip

[sPzAbt653]

So each new game generally requires AI from scratch.

I guess it is obvious that the problem of poor or no 'AI' in games is a result of too many games and too many developers. Look at how many games are listed on the main page here. How many can be honestly said to be 'completed'. How many developers would honestly say that if given the chance, there wouldn't be more work they would do on their game(s).

Those of us from the boardgame days know that one of the reasons we have fond memories of those times is because if there was a bug in game, it didn't take hours of coding to fix. If any of those are like me, they appreciate the efforts that developers and publishers put into to transferring this hobby into the convenience of the computer. I do wonder at times if it wouldn't be better for the industry to have fewer games with dedicated staffs that get an annual fee from players for support, rather than multitudes of come and gone developers and unfinished games that reap a one time purchase payment.

[sPzAbt653]

The online population for >>most<< games is going to be in the 5-15% range.

Thanks for that detail that I suspected but had no 'official' reference for. From there I think it would be fair to predict that most of the 'non-online' population are solitaire players. I don't mean ALL, but 'most' or a larger percentage.
Case in point, I don't know when being 'on-line' came about, but I played computer games from the time they appeared in the 80's and never went on-line concerning games until 2007, after having played TOAW for 5 years. The only reason I ended up here at that time is because I had a burning question, and it seemed easier to go to a website listed in the manual than to write a letter to the developer!
I am glad I did, I enjoy hearing input from others, even if it is only from the 5-15%. But I still would not play multi-player, it doesn't suit my lifestyle.

[berto]

Most players wish for a decent AI, but be careful what you wish for.

I am in the process of extending the helicopter AI for Campaign Series: Vietnam. I can dream up all sorts of neat ideas, things to consider, assessments to make, nuanced actions to take. But apart from the extreme difficulty in coding some of these neat ideas, I know that many of them are very processor intensive. They would slow the game way down.

When playing vs. the AI, I will typically spend 10-30 minutes or more per turn planning and making my moves. Would I, would anybody, sit still waiting for the AI turn taking as long?

No, impatient as we are, we expect the AI to run through its turn quickly, in a minute or two. Any AI turn longer than a few minutes, most players would find unacceptable.

A decent and fast AI? Like having your cake and eating it too. Ain't gonna happen!

[stormbringer3]

I only play vs. the AI. I have passed on more than one game here because of poor AI reviews. I'm looking for a challenge not eye candy. My RPG games give me plenty of that. I have bought games here where you have an option to add to the AI processing time.

[TheGreatRadish]

ORIGINAL: jack54

ORIGINAL: vonRocko


The whole reason I play pc games is not needing a human opponent or playing "hotseat". I have boardgames if I want that.

+1... exactly the way I feel.

As do I. I'm 34 now and I had a tiny group of friends at school who would play these things with me at one time but real life makes it increasingly difficult even with the first class MP systems that Matrix/Slitherine use. Games that don't have AI or barely make an effort are an instant no purchase from me even if the subject material is exactly what I want. I don't think that make me a whiner. It disturbs me greatly that people think I should just accept only being able to play against myself in hotseat - why would anyone ever want to do that?

[berto]

ORIGINAL: stormbringer3

I have bought games here where you have an option to add to the AI processing time.

Which will get you a marginally better AI, but likely still not decent. If you want a truly decent AI, give the computer as long as it takes. Which will be way more time than most players will have patience for.

[berto]

ORIGINAL: TheGreatRadish

It disturbs me greatly that people think I should just accept only being able to play against myself in hotseat - why would anyone ever want to do that?

For the historical immersion? As a sort of living history book?

Back in the day, I played all of my board war games solo hotseat, even by choice. I had no difficulty bifurcating my mind, alternately thinking and playing one side then the other. I had a great time.

Would playing against a competent AI be better? Sure. A human opponent. Maybe. [8|] But if I like the subject matter enough, if I crave the historical immersion enough, I won't let the lack of a decent AI, or a compatible human opponent, deter me.

Competition ain't everything.

Some people like to spend hours and hours putting together puzzles (not my bag). Maybe playing a war game, even one without a decent AI, is like solving a puzzle?

Decent AI or no, war gaming is just a pleasurable pastime however you "play" it. For some folks.

Not saying my POV is right; it just suits me (and possibly a good many others). YMMV.

[loki100]

ORIGINAL: berto

...
Decent AI or no, war gaming is just a pleasurable pastime however you "play" it. For some folks.

...

Yep, completely agree.

And at different times/games I get my fun out of vs AI, vs human and hotseat. Like others in this thread I started playing paper/counter games and simple geographical issues meant that had to be left hand-right hand. So there are games I really like vs AI whether its because they have a wego order system (eg Flashpoint), one side is essentially passive (WiTW) or they are smaller puzzle based games (FoG or Campaign Series). Not saying that these are not also fun PBEM but for different reasons vs AI works well.

At the other end of the scale I find some of the more recent AGEOD games a bit mheh vs AI such as Civil War or Revolution under Siege but great fun left hand-right hand, with a detailed history book open besides the computer.

I do this for fun and I get that fun in lots of ways, but I work freelance and contract basis. So I sometimes have a lot of time (ie no income) and sometimes I am very busy (ie income) and the two extremes are not always PBEM friendly.

[wodin]

Thankfully I'm it seems as useless as a kid under 13! As I'm useless at games cos on the whole an AI !done well is a challenge for me. One thing I don't do is game the mechanics looking for the AI weaknesses.

[Mobeer]

ORIGINAL: berto
I am in the process of extending the helicopter AI for Campaign Series: Vietnam....

When playing vs. the AI, I will typically spend 10-30 minutes or more per turn planning and making my moves. Would I, would anybody, sit still waiting for the AI turn taking as long?

I'm not familiar with that particular title, but there are plenty of wargames like:
- Turn starts
- Red player gives orders
- Blue player gives orders
- Turn execution - both sets of orders run

The ordering of the Red and Blue order phases makes no difference. So if the human player is Red and takes 10-30 minutes to give their orders, then the AI should also have THE SAME 10-30 minutes to give its orders - it just works in parallel.

So for example:
- turn starts
- Red (human) player takes 20 minutes to give orders
- Blue (AI) player takes 15 minutes to give orders, then waits an extra 5 minutes for Red (human)
- turn execution

or else:
- turn starts
- Red (human) player takes 10 minutes to give orders, then waits an extra 5 minutes for Blue (AI)
- Blue (AI) player takes 15 minutes to give orders
- turn execution

War in the Pacific is the Matrix Game's title that comes to mind as amenable to working like this.

[MikeJ19]

Very interesting discussion. So far, I only play against the computer. Like many others, I started playing board wargames and finding opponents who would play was always hard. I played Squad leader and found a good opponent, but he was three games behind me and I would keep forgetting which rules were in play.

Over the last few years, I have been playing Flashpoint and now TOAW IV. The AI is not the best, but it is good and gives me a challenge. I can play when I have time and have fun.

In the fall, I'm planning to try for the first time to play against someone using the PBEM. I'm looking forward to the challenge, but concerned that I will not be fast enough with my turns. However, it will be fun to try.

Have a great evening,

[berto]

ORIGINAL: Mobeer

ORIGINAL: berto
I am in the process of extending the helicopter AI for Campaign Series: Vietnam....

When playing vs. the AI, I will typically spend 10-30 minutes or more per turn planning and making my moves. Would I, would anybody, sit still waiting for the AI turn taking as long?

I'm not familiar with that particular title, but there are plenty of wargames like:
- Turn starts
- Red player gives orders
- Blue player gives orders
- Turn execution - both sets of orders run

The ordering of the Red and Blue order phases makes no difference...

"Plenty of wargames." Really? And how do you know if the other-side AI processing doesn't primarily occur during the turn execution phase? We're not (I think) talking about human-vs-human PBEM WEGO here. How can you be sure when and in what sequence the AI does its thinking? Do you have access to the source code?

... So if the human player is Red and takes 10-30 minutes to give their orders, then the AI should also have THE SAME 10-30 minutes to give its orders - it just works in parallel.

Anyway, true. Multi-threaded code and parallel processing are a possible fix for this. But you have just raised the technical bar far higher than what I myself could jump over.

(Maintaining and extending older code as I do is a mixed blessing. No way I could reinvent these wheels, much less turbocharge them.)

This:

ORIGINAL: IainMcNeil

We'd love to publish more hardcore wargames but they take a long time to develop and there are not many people able to make them and the pool seems to be shrinking. There is a real shortage of wargamers who can code to a level needed for modern commercial games. Games have become more technically demanding and it limits who can perform the role. And programmers coming from a more mainstream background struggle to get to grips with concepts and ideas from wargames. There is a real shortage of personnel...

In addition to the AI, throw in also the requirement to code spiffy 3D graphics, fancy (and customizable!) UIs, robust multi-player, etc., etc. Could anything less than a team of specialty programmers pull it off?

Programmers who can code at such a high level will likely not be developing games for the tiny niche wargaming market. There are bigger paydays to be had elsewhere.

[berto]

More thoughts about parallel processing wargame AI.

As I understand it, in gaming, even some of the most successful attempts at parallel processing involve splitting the graphics and/or the networking from the core game functions. It is far, far harder to separate the core game engine into concurrently executing parallel segments, with the AI "player" reacting (adapting strategies) in real time to the human player's actions.

From Wikipedia:

In some cases parallelism is transparent to the programmer, ..., but explicitly parallel algorithms, particularly those that use concurrency, are more difficult to write than sequential ones, because concurrency introduces several new classes of potential software bugs, of which race conditions are the most common. Communication and synchronization between the different subtasks are typically some of the greatest obstacles to getting good parallel program performance.

Consider, too, the vast complexity of many wargames (so many units, huge maps, hundreds and hundreds of rules, etc.). Disentangling it all to run well and (near) bug-free in a multi-processing coding environment is darned near impossible. For mere mortals anyway.

[rico21]

To complete this thread, it would be interesting to talk about the evolution of players and more precisely their consumption. Personally, as I get older, I have more and more time to buy and play new games and paradoxically I buy less and less.
The main reasons are that the new games are very similar to the old ones to stimulate the buying impulse but also that I need more time than before to "finish" a game, that is to say exploit fully its potential .
Yet I am always looking for the Grail, the ultimate game, one that will make me forget all the others.
So gentlemen developers, with all the admiration I have for your art and your dedication,
Surprise me!

[Alan Sharif]

Must confess, I can live with poor graphics, and/or a poor interface, but no AI means no sale where I am concerned.

[Lobster]

ORIGINAL: MrsWargamer

I'll just say it, we ONLY have AIs in our wargames to placate whiny wargamers that refuse to actually do something to find a human adversary, or who simply refuse to accept playing the game vs themselves in hotseat mode

If this were true then Matrix might as well turn off the lights and lock the doors. They would go broke in no time. I would wager the vast majority of games are played by solo gamers and you've just insulted all of them by labeling them whiny. Also there are many games where the computer opponent does a very good job of giving the human side a go.

[rico21]

#whiny too [:D]

[MrsWargamer]

ORIGINAL: Lobster

ORIGINAL: MrsWargamer

I'll just say it, we ONLY have AIs in our wargames to placate whiny wargamers that refuse to actually do something to find a human adversary, or who simply refuse to accept playing the game vs themselves in hotseat mode

If this were true then Matrix might as well turn off the lights and lock the doors. They would go broke in no time. I would wager the vast majority of games are played by solo gamers and you've just insulted all of them by labeling them whiny. Also there are many games where the computer opponent does a very good job of giving the human side a go.

Your post insinuates I was unaware I had insulted them