Jump to content
Chris

Xenonauts-2 January 2018 Update!

Recommended Posts

I wasn't considering having pop-up events, no - but I do quite like the idea, it's just a question of whether there's a place they can slot into the design. We already have the somewhat similar "strategic operations" on the Geoscape which work similarly to the resistance events in XCOM2 where you assign a soldier to a non-playable mission for X days and get a reward at the end (potentially with your units suffering wounds etc), because I really liked the way they worked to make you rotate your squad a bit.

However because I don't want to change too many variables on the Geoscape at once, there's only one type of them so far - basically at the moment they're the only way you can raise relations with a region (shooting down UFOs prevents relations loss but does not raise relations). We'll see how that works out in the most basic form and then think about how we can expand / change it, particularly once we bring in the staff training system.

Share this post


Link to post
Share on other sites

@Drakon about the AI and tech in the game: the aliens need not make use of every element of tech available. If they did, then the human and alien factions would feel too similar; differences would seem superficial. Just because a particular game mechanic may be difficult to teach to the AI, it doesn't follow that you should deny it to the player. Sure, poor implementation might worsen your experience (e.g. AI doing stupid/unfair things) but players can usually forgive some element of ridiculousness if it is fun (e.g. jet packs and light sabres ;) ). So yeah, things like teleportation should be left out unless you can nail the way the AI uses it and I guess if elements like proximity mines are going to be implemented then other mechanics may need to follow to forestall the rage (e.g. running-soldiers can't spot them but static or walking-soldiers can).

In terms of the Geoscape and air combat: I'd advocate bundling the overall strategic decisions into some turn based format. All the resource management, R&D, base building etc doesn't need to be real time and staggering those events is to no benefit; they can all be a set of decisions made in one instance, a logistical phase of each turn. However, as air combat is pretty much the only exciting thing that happens in the geoscape layer, overtly reducing that to a decision tree of text boxes and percentages would be sad. Some level of animation, or watching an event unfold, is a must. There is some stuff to pick apart here and I'd say it is worth figuring out what makes air combat exciting and what level of input is fun-strategic.

I'd say air combat involves two main elements: chasing and then engagement. The chasing part is satisfying to watch as red and green blips approach one another but does it need to be plotted on the standard geoscape? Also, how many real-time decisions made at this juncture are meaningful? Does the question of a successful chase have to hinge upon you bringing pilots in from other bases? As for the engagement, I can appreciate that reaction based combat is not everyone's cup of tea but the dog fight should feel thrilling - quite the opposite of an autoresolve and deserving of some kind of mini-game. Maybe take the aspect of FTL that involves targeting parts of the enemy ship(s) but fold that into a turn based system. Say you give your pilots orders for distinct actions, 1 time-unit per turn: making a pass (firing) or other maneuvers (closing distance, evasion or retreat). Just like FTL, the mechanics could be as simple as %chance to hit (no messing around with 3d spaces) but as your engagement takes place over several turns your strategy would have room for complexity, though in actual fact it might be a semi-solved problem i.e. someone could calculate the heuristics of what orders would likely yield each outcome.

Share this post


Link to post
Share on other sites

Again, the aliens already will need better stats from base to create a somewhat challenging game, because tactically they will almost certainly be inferiour to 90% of all human players. Giving the human players then additional tactical options that the alien AI cannot use will automtically necessiate even more drastic steps to give the AI any chance of doing damage at all.

Throughout most of my Xenonauts 1 plays a majority of my soldiers didn't have advanced weaponry, because plasma weapons take a lot of time to manufacture and i skipped laser weapons and that wasn't a problem at all - so i don't see how adding alenium as a limiting factor would change much. I also could only afford advanced armor for a few, and since that offered significantly increased tankiness and partially flight, i cannot really see where the difference to Xenonauts 1 will be.

 

7 hours ago, Ninothree said:

Just because a particular game mechanic may be difficult to teach to the AI, it doesn't follow that you should deny it to the player. Sure, poor implementation might worsen your experience (e.g. AI doing stupid/unfair things) but players can usually forgive some element of ridiculousness if it is fun (e.g. jet packs and light sabres ;) ).

Some people - me and Bobby Gontarski would be examples - would see that differently. Firaxiscom offers hyperpowered super soldiers with "fun" extreme weapons (haven't played Firaxiscom 2 myself, but heard it's pretty much the same), with graphics and sound assets that Goldhawk simply for financial reasons cannot match, but third rate writing and an AI that is anything but impressive. Trying to compete with the power fantasy that Firaxiscom offers seems like a very poor choice for me: the voice acting of the soldiers that Firaxiscom has just makes that fantasy way more immersive. For some players the tactical gameplay is the major focus rather than a power fantasy, and hence those players would be readily willing to sacrifice "fun" items like light sabres if the tactical gameplay gets better for it, and hence would readily rather buy Xenonauts 2 rather than a Firaxis competetor, if that is what Goldhawk promises to deliver.

 

7 hours ago, Ninothree said:

about the AI and tech in the game: the aliens need not make use of every element of tech available. If they did, then the human and alien factions would feel too similar; differences would seem superficial.

I do not agree. In contraposition i would post that it is not a bad idea at all to have a number of technologies that are available to the aliens, that human players cannot obtain - psionics and teleporting would be examples from Xenonauts 1, even though neither was particularily well implemented and hence not necessarily well liked. If the humans get tech that the aliens do not get, i would suggest it to be something that the aliens wouldn't bother with, because they think they have better technology already - for instance no transportable bridges or ladders because the aliens have teleportation.

Share this post


Link to post
Share on other sites

I think before anyone assumes the AI "won't even challenge a moderately skilled human" we should see how it turns out.  If I remember correctly many of the things that were possible/planned to do with the AI weren't implemented in Xenonauts 1 because they ran out of time to do them.  I'm going to assume that Jjis-Jan (probably spelled wrong) is going to be able to put more "brains" into it this time around.  He had all the ideas and solutions in planning before, so this time it would seem he'd be able to implement more of them.

Edited by StellarRat

Share this post


Link to post
Share on other sites
1 hour ago, Kiril2704 said:

For me turn-based geoscape will ruin the game.

1 hour ago, Kiril2704 said:

 

So. I think that kiril thinks this forum is a poll. I have an idea. Let's make a poll: "What kiril thinks this forum is about?"

Edited by Rosen

Share this post


Link to post
Share on other sites

Well after a fair bit of research, I found this:

5 hours ago, Kiril2704 said:

For me turn-based geoscape will ruin the game.

According to the date log, they said it quite recently; I reckon that that is valid evidence but I do like poles...

Share this post


Link to post
Share on other sites
5 hours ago, Rosen said:

So. I think that kiril thinks this forum is a poll. I have an idea. Let's make a poll: "What kiril thinks this forum is about?"

18 minutes ago, Ninothree said:

 

18 minutes ago, Ninothree said:

Well after a fair bit of research, I found this:

"For me..."

According to the date log, they said it quite recently; I reckon that that is valid evidence but I do like poles...

5 hours ago, Kiril2704 said:

For me....

face-with-tears-of-joy_1f602.png

Edited by Rosen

Share this post


Link to post
Share on other sites

Okay folks, let's not pick on people who want to voice their opinions, yeah? Address the argument, not the person making it. 

Share this post


Link to post
Share on other sites
On 26.1.2018 at 8:15 PM, StellarRat said:

I think before anyone assumes the AI "won't even challenge a moderately skilled human" we should see how it turns out.  If I remember correctly many of the things that were possible/planned to do with the AI weren't implemented in Xenonauts 1 because they ran out of time to do them.  I'm going to assume that Jjis-Jan (probably spelled wrong) is going to be able to put more "brains" into it this time around.  He had all the ideas and solutions in planning before, so this time it would seem he'd be able to implement more of them.

And here you see the issue quite clearly. Most game companies do not have the option of postponing the launch of their game 3 times, like Blizzard did with Starcraft I. While that may give you a game that is lauded for decades, and basically cements your status as a leading game companie for a similar amount of time, it is usually just financially not doable. The unfortunate reality commonly faced instead is rather that around launch date, time and money runs out, and parts of the game are left either unimplemented or implemented shoddily. Setting yourself up in an even worse strategic situation, gving your AI programmers an even harder time from the get go seems like an obviously poor choice to me. If there's time left at the end, and everything else is already polished and fine, then sure, why not, add in a few more cool gadgets and try and have the AI use them too, and if there's ONE particular item that the AI cannot use then it's probably not such a big deal (or if it turns out to be, you can overcome it by just removing that one item), but adding a ton of stuff in the design phase that will later on only cause additional problems is as strategically unsound as you can get.

 

There is of course the question what you type of game you are trying to make: a solid challenging tactical simulator, or a shiny ride with nifty gadgets and flashy special effects. Firaxis made a decent amount of money with the latter, so you could argue why not try and jump the bandwagon, i just think that it makes more sense to actually offer a different game, potentially targeting a different audience than trying to make "Firaxiscom 2b" with less money and resources. The successes of games like "Faster than Light" or "Minecraft" and the immense recent interest in Roguelikes (or what is called such these days) indicates that there actually would be enough potential buyers for a more challenging, well thought out game, even if - for financial reasons - it lacks some nice assets like complete voice acting and additional visual special effects.

Share this post


Link to post
Share on other sites

The intention is that aliens will become tankier for the troopers without special Alenium-powered gear, but will be relatively less tough for units that do have them. The numbers can always be adjusted to ensure that is the reality of the situation, and in that case even if we do end up shipping a game with AI that will offer no challenge at all to anyone then it won't matter. And if we do have competent AI then we can lower the tankiness of all the aliens, as they'll all inherently be more dangerous.

I'm guessing the deeper point is more that you just don't like the idea of giving the player more fun tools that the AI doesn't get because this game is meant to be a realistic game, and you don't want the players to have more powerful tools because it sounds to you like the superpowers of the soldiers in XCOM. I'll leave aside the assumption that adding more tactical options to the game is dumbing it down and stopping it from being the "solid challenging tactical simulator", because I think the fundamental issue here comes in your definition of "challenging". From what you've written, it seems like you think a game can only be challenging if the AI and the player have similar limitations.

I think it's easy to lose sight of the fact that this is a game and the AI is expected to offer a fun experience rather than doing whatever is optimal for it to win the game. For instance, the optimal tactic for the AI across the entire game is to form all its soldiers into a blob every mission and charge your troops, ensuring they focus fire on and kill one or two high-ranking soldiers every single time you go into combat. Sure, all the aliens die - but there's effectively unlimited aliens and it means that when the final mission rolls around you have no veterans and the final mission will be much harder.

Now I'm not going to debate whether that tactic would necessarily work - the point here is just to show that if the AI took the optimal choice, the game wouldn't be very fun to play. The fact of the matter is that the player and the aliens have different goals on a mission, and irrespective of the results screen a mission can be a win for the aliens even if they kill only two or three Xenonauts and get fully wiped out. Also, the pacing of the mission is to a large extent defined by the AI - if the enemy acts aggressively and comes to you, you won't be doing any of the cautious exploring that makes up the classic X-Com experience, you'll just have a massive shootout two turns into every mission when every alien on the map comes running towards the sound of gunfire.

If you go too far down the "AI and player should have the same tools and objectives" argument, you end up in a situation where you're essentially advocating we should also remove that sort of sanity check from the AI ... and trust me, you wouldn't like the game that resulted from it.

Share this post


Link to post
Share on other sites

I can give 2 good examples of what Chris is talking about re. AI. Early on in X1's AI development, the first thing the AI did was to have each AI unit do a 360 spin so it knew what was around it. The thing was, because the AI enemies were spread out across the map and because all AI enemies shared map data, the AI would know where the player was not across a large section of the map. It could then deduce where the player was most likely to be and send speculative fire at the tiles it thought the player units were, which it did and the AI was accused of cheating.

Later on after this was stopped, the AI did exactly what Chris said it would do. Once the AI spotted you, this information was shared among all enemies.The AI was designed to clump together, so the enemies would form up into waves of enemies which charged the dropship. This would make thrilling and exciting battles around the dropship as wave after wave of grouped enemies charged in, but it made the maps pointless as you didn't have to go very far, you just had to wait for the AI-formed blobs to turn up.

Share this post


Link to post
Share on other sites
1 hour ago, Chris said:

I'm guessing the deeper point is more that you just don't like the idea of giving the player more fun tools that the AI doesn't get because this game is meant to be a realistic game, and you don't want the players to have more powerful tools because it sounds to you like the superpowers of the soldiers in XCOM.

While i do dislike tools for the player the aliens do not get, both because i think it does not fit thematically and because it logically doesn't make any sense lore wise, both these arguments fall by the wayside compared to whether adding those tools will make for a better game. I did bring clear arguments why that in all probability will not be the case - provided your focus is to make a tactical simulation game rather than computer barbie for men. And don't get me wrong - i'm not trying to say that computer barbie for men is awful; i think we all had our moments of gloating over "Ooooh ... all these shiny gun attachments! Which one should I put on tonight?", in that respect i plead guilty and have enjoyed my time playing computer barbie for men, i just tried to find a genre that actually describes the situation, whether you put priority on that or on the enjoyment of the actual tactical fight.

2 hours ago, Chris said:

I'll leave aside the assumption that adding more tactical options to the game is dumbing it down and stopping it from being the "solid challenging tactical simulator",

That partially depends very much on what those options are and how they are implemented. I do not think that anyone decided to equip plasma weaponry in Xenonauts 1 when they had mag weaponry available. In so far, adding that option did not actually broaden the decision space of the player, and paradoxically, putting in something overpowered can actually narrow down the number of valid tactical options. Think of "Magic - The Gathering" decks where EVERYONE runs the same cards, regardless of what deck they are building, because those cards are just that good that not using them would be an obvious strategic mistake. That's basically reducing the maximum number of cards in a deck by the number of overpowered cards, resulting in a lower possible variety of decks. Conversely, options that vary in quality depending on the situation will expand the decision space and make the game more interesting, but are usually a lot more work to implement and get right. Think of adding a ladder as possible equipment for human soldiers versus a jetpack. The ladder is obviously way more work to put in the game than the jetpack, but requires more player thought to use, and thus comparatively makes the game a better tactical game.

1 hour ago, Chris said:

the point here is just to show that if the AI took the optimal choice, the game wouldn't be very fun to play. The fact of the matter is that the player and the aliens have different goals on a mission, and irrespective of the results screen a mission can be a win for the aliens even if they kill only two or three Xenonauts and get fully wiped out. But the experience would feel cheap if an alien came charging out of heavy cover and sacrificed its life to take a point-blank shot at your closest soldier and kill him; it's the optimal move but it's not fun for the player.

If you go too far down the "AI and player should have the same tools and objectives" argument, you end up in a situation where you're essentially advocating we should also remove that sort of sanity check from the AI ... and trust me, you wouldn't like the game that resulted from it.

I'll have to tell you that you are quite literally wrong about this. When i switched to the Community Edition of Xenonauts the very exact scenario you described happened. I lost several soldiers to Sebillians hiding in cornerspots with shotguns before i learned to check my corners, and those Sebillians were mowed down right afterwards by my troops because they blasted away short range at full auto and didn't save any TUs to retreat. I was elated. Finally i actually had to look at the screen playing, and i quickly learned to check my corners. If a Caesan took a long range pot shot at me and then ran out of sight for cover, it was easy enough to determine the optimal hiding spot, and just flush it with a flashbang. The AI in Community Edition is still bad, but it is so much better than in the base game. Sure it'd be nice for them to actually have survival instinct and clever group tactics, but i'll gladly take hyper-campy aliens over the basic AI that makes it's units waste TU by running from one cover to the other and back again.

Whatever your "optimal strategy" is, hand it to players for a couple of days and they'll shred it easily. All your troops are grouping and then moving against me as a blob? I just put everyone on overwatch with rocket launchers then, and given that my troops have cover and your's don't, there's little doubt who'll win that engagement. Now you stagger them so they don't all die from response rocket shots? Say hello to Mr. Redshirt, the new recruit i take on every mission, who is sitting in front of my troops as a canary. When he dies, i know your troops are there, and they get obliterated by my countercharge. I actually read an amusing entry in a discussion on the X-Com 2 reddit, where one of the gamers pointed out that the fact that the AI cheats and always knows where his soldiers are actually makes the game easier for him. He can anticipate the AIs pathing and just sets up his soldiers in overwatch and as the pods float in, they get taken down by a hail of fire. Boring as fig stew in my opinion, but certainly effective.

And i do not say this thinking i belong to an elite minority of epic gamers of awesomeness while the plebs won't be able to handle the hard mode for "real men". Quite the opposite: developers keep underestimating the gamers capacities rather drastically. Literally everyone i talked to about it loved the moment when for the first time in Metal Gear Solid 2, after you had taken out all the guards in an area, a little while later one of their walkie talkies would turn on: "Patrol 2, status! ... Patrol 2, respond! ... No response from patrol 2. Send in the backup team!" If your game is good, sure, players may gripe about it being too hard initially, but soon enough they'll develop strategies to deal with the problems and win anyhow. Ask around how many games people disliked and put away because they were too hard. There will be practically none. Games are dropped because players feel frustrated, and that doesn't come from the game being hard, that comes from the players feeling cheated or treated unfairly. That's exactly the case in FiraxisCom 2, where the players noticed the AI cheating, and that's exactly the reason why conversely the Dark Souls franchise is so very successful and highly praised.

2 hours ago, Chris said:

From what you've written, it seems like you think a game can only be challenging if the AI and the player have similar limitations.

No. I'd say that's probably literally impossible. I'd like to meet the guy who can write an AI that can come up with creative adaptive tactics like a human can. Either way, he'd probably be above your budget. So the AI will be in many aspects way more limited than a human player will be, which necessiates some workarounds: higher stats for AI units, for instance, or access to gear the human player won't get. Every tool you give to a human player that the AI doesn't have means you need more workarounds to give the AI a chance. Full vision over the entire battlefield from the get go, for instance? And that is where players start calling bullshit and get frustrated. Add to that the issue that the AI needs to be able to somewhat anticipate player actions. Every option you give to the humans makes the decision tree that much more complicated for the AI programmer, if he is to anticipate the possibility of it.

2 hours ago, Chris said:

The numbers can always be adjusted to ensure that is the reality of the situation, and in that case even if we do end up shipping a game with AI that will offer no challenge at all to anyone then it won't matter.

You are actually stating that you see no difference in quality between an AI that is challenging because of smart decisions and that you have to outthink versus an AI that is challenging because of high stats that you have to out-equip? I am baffled.

Share this post


Link to post
Share on other sites
11 minutes ago, Drakon said:

Whatever your "optimal strategy" is, hand it to players for a couple of days and they'll shred it easily.

You're completely missing the point. It doesn't matter if you can beat the optimal strategy or not - if the optimal strategy is forming a big blob of aliens, that's what they will and should do in literally every mission by your logic. And then you either get steamrollered because you don't know how to beat the AI, or you figure it out and win every single time by using exactly the same tactics.

Congratulations. You've won the game by bringing rocket launchers to every mission. This game where the AI tries its hardest to win sure is fun, isn't it?

I'm well aware that players are going to beat pretty much any AI a coder can create, but the point is that they should enjoy the experience of doing so. Making the AI try to win at all costs means it is marginally more likely to win against an unprepared player but it doesn't actually make the game more fun for anyone.

Share this post


Link to post
Share on other sites

Well, firstoff, Max_Caine wrote just recently

30 minutes ago, Max_Caine said:

Once the AI spotted you, this information was shared among all enemies.The AI was designed to clump together, so the enemies would form up into waves of enemies which charged the dropship. This would make thrilling and exciting battles around the dropship as wave after wave of grouped enemies charged in

so based on his words, yes, that seemingly does make for more entertaining battles, and secondly, if you write your AI as "do the same thing every time with no random variation" then yes, you will encourage the players to do the very same thing every round. If "form big blobs and run at the enemy" is resulting in consecutive losses, then it isn't a very optimal strategy, and if you think anything i wrote logically concludes that they should do that literally every mission, then i'll have to wonder who taught you logic.

I think it is you who are missing my point.

I at no point wrote: "The AI should at every point in the game be a god." I stipulate that in a single player tactical game the quality of the AI determines A LOT of the quality of the game, because the fun lies in out-thinking / outwitting the enemy AI, and hence, anything that lowers the quality of the AI should be viewed with caution. And, as i already wrote above, that is no absolute statement from god that must never be bent, no, there can be times when it is the better choice for the game to put something in that will make the AI worse, but those cases are definitely in the minority.

If the point of a tactical game is not to outwit the enemy, then what is it, in your opinion? And if there's nothing to outwit, then where is that fun you are talking about?

Share this post


Link to post
Share on other sites
14 hours ago, Drakon said:

I think it is you who are missing my point.

I at no point wrote: "The AI should at every point in the game be a god." I stipulate that in a single player tactical game the quality of the AI determines A LOT of the quality of the game, because the fun lies in out-thinking / outwitting the enemy AI, and hence, anything that lowers the quality of the AI should be viewed with caution. And, as i already wrote above, that is no absolute statement from god that must never be bent, no, there can be times when it is the better choice for the game to put something in that will make the AI worse, but those cases are definitely in the minority.

If the point of a tactical game is not to outwit the enemy, then what is it, in your opinion? And if there's nothing to outwit, then where is that fun you are talking about?

Yeah, I probably am missing your point. Basically I was just saying that any game AI is set up with the intention to provide a challenging and engaging experience for the player, rather than to win at all costs - and whilst there's overlap there, they aren't always the same thing. I'm sure that's something you agree with, right?

The broader point I was making is that symmetry between the AI and the player isn't always necessary. Xenonauts 1 has plenty of asymmetric equipment - for instance, the aliens don't get access to medkits, nor combat shields, nor stun weapons, nor any explosives beyond grenades (which they are barely capable of using anyway). I don't think that means that the player should be denied those items because AI can't use them / has no reason to do so, though.

Most of the proposed new items are definitely not going to have a major effect on game balance. Making laser weapons more powerful but less widely available is really not going to break the AI, and if they are too powerful then that is exactly the sort of thing that can be fixed by tweaking the damage numbers. More exotic powers like being able to teleport a soldier anywhere in the battlefield once per mission do indeed have more potential to break the game, but they *can* still be balanced in ways that don't involve simply making the aliens tankier (e.g. you can add a TU cost to the teleport).

One of the things that was faintly disappointing in Xenonauts 1 was that the humans deciphered alien technology and all they did with it was make slightly more powerful / more protective variants of the basic gear they already had. It meant the gameplay stayed pretty much exactly the same all the way through the game. I want to mix it up a bit more this time around because it makes sense both thematically and from a gameplay point of view; I wouldn't be too hasty to try and shut that down simply because you think the AI might not be able to handle it.

Share this post


Link to post
Share on other sites
9 hours ago, Chris said:

Basically I was just saying that any game AI is set up with the intention to provide a challenging and engaging experience for the player, rather than to win at all costs - and whilst there's overlap there, they aren't always the same thing. I'm sure that's something you agree with, right?

Immediately.

 

9 hours ago, Chris said:

The broader point I was making is that symmetry between the AI and the player isn't always necessary. Xenonauts 1 has plenty of asymmetric equipment - for instance, the aliens don't get access to medkits, nor combat shields, nor stun weapons, nor any explosives beyond grenades (which they are barely capable of using anyway).

As a sidenote: i learned to hate the alien grenades when the Xenonauts Community Edition AI was using them. It is the one thing that consistently does damage against my troops, combat shields or no, and they are using it well enough - actually usually preferring it over their main weaponry if they are in range.

But yeah, symmetry between the AI and the player would for instance actually mean the AI getting the same number of units as the player with the same stats, and in that case humans would stomp the AI easily, no challenge involved. I never advocated that. I warned against making the job for the AI team even harder than it already is.

 

9 hours ago, Chris said:

More exotic powers like being able to teleport a soldier anywhere in the battlefield once per mission do indeed have more potential to break the game, but they *can* still be balanced in ways that don't involve simply making the aliens tankier (e.g. you can add a TU cost to the teleport).

One of the things that was faintly disappointing in Xenonauts 1 was that the humans deciphered alien technology and all they did with it was make slightly more powerful / more protective variants of the basic gear they already had. It meant the gameplay stayed pretty much exactly the same all the way through the game. I want to mix it up a bit more this time around because it makes sense both thematically and from a gameplay point of view; I wouldn't be too hasty to try and shut that down simply because you think the AI might not be able to handle it.

The issue i see is that i think the balancing of such special powers is way harder than you make it out to be. If any of those powers, or in a much more likely scenario any combination of some of those powers, break the AI they basically become a supreme "I win"-strategy, which will actually not diversify but break the gameplay. And i'm not even talking about the AI completely hanging up and not responding at all any more, but just a situation where the AI uselessly shuffles back and forth between two positions and can be picked off by the player with ease. Once a player figures something like that out, they can use it every scenario and get a free win - compare for instance the player who abused the AI cheating and knowing of his position in Firaxiscom 2 that i mentioned above. Avoiding that becomes exponentially more difficult with each additional option you give the player, because the AI programmer needs to figure out what might be done with any combination of the existing items. My basic assumption was that if the AI programmer can figure out a way how to make the AI use the option the player has, then he can probably also figure out a way for his AI to operate around the players actions, and the game balance won't be further skewed in favour of the player.

I simply think that adding things in the design phase that will create an exponential additional workload later is as a rule a very bad idea, and should only be done very sparingly. If you show these things to your AI team and they say that their AI should be able to handle them no problem, then sure, why not, diversifying gameplay throughout the experience is certainly a good idea. But if those things get put in too early and without reflection, there will be massive adverse effects. If any players hear that they might get lightsabers, and lateron figure out they actually won't get lightsabers because those would make the game worse, there will be a lot of griping. But there are also developers who end up falling so much in love with an idea, or fall prey to the sunk cost fallacy, and then try and explain away the actual problems with the items or push the problem on the AI team. That's exactly how you end up with a crappy AI: by not planning ahead and not giving the AI team enough time or resources to do their job well. And as can easily be seen on the Firaxis message boards, that does make a huge difference in whether the players enjoy the game and consider it a good game or not.

(Sidenote: figuring out a weakness of the AI and capitalizing on it is actually part of the fun of the tactical gameplay. But that weakness should not have been an accident but instead designed by the AI team, so that rather than there being that one supreme strategy that just wins every mission, the AI can over time change it's behaviour and also announce those behaviour changes to the player, making them adapt an diversify their gameplay over time. Imagine if in X1 the player had intercepted a transmission from the aliens "No humans spotted in sectors 1, 2 and 4. Bombard sector 3!" There would have been no complaint about the AI cheating, and the player would have immediately known why he was being shot at from off screen and what he should do about it.)

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×