Jump to content

Tank AI Help


Recommended Posts

What you can also do if you want to is to 'combine' the various orders types...

order 1. DASH to the first firering location

order 2. MAX ASSULT with no painted (yellow) order loaction but add a facing dot (alt-click to do (green dot)) in the direction of travel or the direction you like the tank to face.

order 3. The same as 2.

order 4. DASH to next firering location.

order 5. same as 2.

order 6. DASH to next firering location.

order 7. same as 2

order 8. same as 2

This way the AI tank will move at full speed to its first firering location and ones it gets there wait for TWO minutes (one for each of the added orders 2 and 3) with full (MAX ASSULT) fire readiness. after two minutes at this location it will move at full speed to its next firering locattion and wait for ONE minute (only one added order here...order 5.) at full fire readiness before moving on at full speed to its 3rd firering location and ones again wait at that location for TWO minutes (orders 7 and 8 ) at full readiness.

Adding extra orders like these without any yellow boxes painted you can add a delay (one minute/order) and also change the stance to anything you prefer or even add an areatarget order if you so wish...

 

 

Link to comment
Share on other sites

59 minutes ago, Glubokii Boy said:

the AI will get some 'stupid' 😉 idea of its own.

Oh yeah!!! If brains were dynamite, these drivers wouldn't have enough to blow their noses with 🤣!!

Ok @Glubokii Boy, these are all great points that help a lot. I've done just that now and also spread the tanks out a bit. I am using only a single square for each tank order, but per your suggestion, I'm putting the orders much closer together. I will try the dash next, but even on Advance it seems to be working very much better. I think we have the beginnings of something that's going to work. Many, many thanks!!

Link to comment
Share on other sites

35 minutes ago, Glubokii Boy said:

What you can also do if you want to is to 'combine' the various orders types...

order 1. DASH to the first firering location

order 2. MAX ASSULT with no painted (yellow) order loaction but add a facing dot (alt-click to do (green dot)) in the direction of travel or the direction you like the tank to face.

order 3. The same as 2.

order 4. DASH to next firering location.

order 5. same as 2.

order 6. DASH to next firering location.

order 7. same as 2

order 8. same as 2

This way the AI tank will move at full speed to its first firering location and ones it gets there wait for TWO minutes (one for each of the added orders 2 and 3) with full (MAX ASSULT) fire readiness. after two minutes at this location it will move at full speed to its next firering locattion and wait for ONE minute (only one added order here...order 5.) at full fire readiness before moving on at full speed to its 3rd firering location and ones again wait at that location for TWO minutes (orders 7 and 8 ) at full readiness.

Adding extra orders like these without any yellow boxes painted you can add a delay (one minute/order) and also change the stance to anything you prefer or even add an areatarget order if you so wish...

 

 

Oh man, this is REALLY helpful!! It won't work in this scenario because they have no idea there is any enemy out there (and neither does the other side). However, this will really help down the road. I had no idea you could add a facing dot until now, so that WILL be really helpful in this scenario as well as others. Very cool - thank you!!

Link to comment
Share on other sites

You can also take bits of inspiration from this post and related thread. https://community.battlefront.com/topic/121649-scenario-designing-with-ai/?do=findComment&comment=1648263

Main ideas (T setup or reversed version of it) is dealing with the random nature of individual units of an AI group between move zones. While it´s not perfect for anything it at least helps the AIP doing its random things in more optimized and also predictable ways.

Vehicle column movement can be initiated best when NOT deploying in column intially. All individual vehicles won´t care of other group members and get into each others way invariably. But one can at least to some degree control the move off timing of individuals. I.e 1st vehicle on the road pointing toward move zone, 2nd about 30m behind (security range for AI micro pathing), 3rd beside the road pointing bits away (~45°) from next movement zone, 4th (and any 5th or more) about 90° or more.

Using "dash" to encourage road movement, units will move off in column like this:

1st (lead vehicle) already on road

2nd also on road but 30m behind

3rd needs to turn on the spot thus spending some time before moving onto the road behind 2nd

4th needs even longer to turn on the spot and then starts moving behind 3rd and so on.

This needs bits of trial and error in scenario author mode, figuring out proper placement and turning rates.

It also helps if the intial setup (without painted setup zone to fix AI deployment) is in sort of a constricted area and the road is the bottleneck through it. A line of "heavy forest" tiles (with or without trees placed) will do i.e. This to tell the AI there´s no short cuts or other options.

This AI group´s (assuming a tank group of 4-5 vehicles) first order should be "dash" and move zone shape a "line", maybe 1-2 action spots deep and 80-100m wide. This zone can be a couple kilometres away (theoretically). The road column once properly formed up near setup zone will use road move until shortly before reaching the painted move zone and then shakes out to line formation. No need to place intermediate orders/move zones. The less, the better.

Interesting thing if the AI group contains some tank rider or passengers, "dash" will lead to about half speed movement, also keeping road column tighter.

 

Link to comment
Share on other sites

Oh wow, looks like I have some more reading and testing to do 😳! Thanks very much, this IS what I'm trying to accomplish. I've at least not got them running into each other now so I am making progress.

I have removed the roads completely and constricting terrain has few options in this context (this is a desert scenario) but everything else looks like it should work. This will have to be a job for tomorrow at this point, but I'm definitely encouraged. I have got the link you provided open in a new tab so will read through it as well. Many thanks for this!!

Link to comment
Share on other sites

Reference the timings for the AI - they are 'exit before' and 'exit after' - you do not have to use both but, for instance, if you want a vehicle to move at the 10 second mark then set your 'exit after' for 9 seconds and leave the other setting blank or set it for 11 seconds.  You may have to tinker around with these combinations a lot to get them to synch or you might have to set both timings the same but this at least I'm pretty sure is achievable in the editor.  I wish you luck with it, personally if I was having this much difficulty with executing a concept I would change the concept.  There's a reason my AI plans are simple 😏.

Link to comment
Share on other sites

I seem to recall that the manual actually has two different explenations for these timings. In one location it mentions something like this:

- Exit before is the first time the AI group will be allowed to start that order

and 

- Exit after is the time by witch the AI group will try to have completed that particular order (reached its objective location).

and in another section it mentions something like.

- Exit before is the first time the AI group will be allowed to start that order

and

- Exit after is the time the AI group WILL start that particular order even if the movement has not been triggered by other means (terrain trigger or AI-order trigger).

 

In my experience the later explenation seems to be more correct and the way it actually works.

And...Sorry ! It's not my goal to try and correct you Combatintman but i belive that the timing options are limited to half minute increments...(maybe you ment to write minutes and not seconds ? 😇)...Or is it indeed possible to set these timings at the desired second level ? If so...How ? 😎

 

 

 

Link to comment
Share on other sites

4 hours ago, Combatintman said:

personally if I was having this much difficulty with executing a concept I would change the concept

<LOL> That's because you're sane - something nobody has ever accused me of... ever :D :D ! No, being serious for a moment, that would be the sensible thing to do. However in my case, some of my problems are due to me being new to the system. I'm not sure what are my own shortcomings and lack of familiarity and how much is simply something the game just can't do as it stands. I've always been a problem solver and that coupled with my somewhat obsessive desire to learn as much and as fast as I can, means things like this are simply a challenge to be overcome to me. Truth is, while I don't know if I'll be able to make this particular concept ever actually work (see down), it is fun trying and who knows, it might end up revealing something or coming up with a concept that nobody has thought of before. Unlikely given the level of talent and experience here, but hey, sometimes it's the "new kid on the block" that doesn't have the preconceived ideas from years of experience that hits upon something new. You never know. In any event, I've learned a tremendous amount of all kinds of things related to how to work the AI in doing this, so far, and which will have applications in pretty much all the scenarios I come up with in the future, two of which I think are about ready for "prime time". So even if this concept doesn't pan out (and the issues I've had up until now are only Step 1 of 2 in this scenario), it will pay dividends down the road in all sorts of ways. I'm starting to look into ways I can get the scenarios I have got done "out there", so hopefully these hard learned lessons will pay some dividends for a good number of people here :) .

You are right, as I found out, about timing, although it goes in 30s intervals, which is a bit long for what I need. However, that's definitely an option. In this case though, @RockinHarry hit the nail on the head for the timing thing (by turning tanks further down the line at an angle to start so the start time is delayed somewhat while they conduct their turns). I did a quick run last night using that technique and it worked perfectly. I still have some experimenting to do but the hardest part of this is over, and that will be very useful in all kinds of situations I can think of going forward. The one small "fly in the ointment" is that to get them to maintain a straight line (more or less, and it's pretty good in fact), you do have to use the "Dash" command, which has been suggested by @ASL Veteran (and @Glubokii Boy). In this scenario (which I guess I'm going to have to fully reveal because Step 2 in this is another conundrum completely as I'm finding out, but it will spoil the surprise for anyone reading this, sadly), they need to stop when they encounter the enemy. In my tests late last night, they did shoot back at the enemy, but as per the manual, they do try to continue on as much as possible, largely ignoring the fact that someone is trying to turn them into a heap of scrap metal. What I want them to do is stop when they come onto the enemy and fully engage them so you end up with a (hopefully) rather entertaining tank battle in open ground. However, because they aren't expecting to encounter anyone, and neither is the enemy as they too are on the move, there is no "fixed" enemy position, I can't really build in a stop to their dash across the map and switch to an all out assault ahead of time. I could do that of course, as the Computer AI isn't going to give anything away to the other side, but truth is, I can't say exactly when the meeting will occur. Anyway, I'll explain further once I get a bit more sorted out with Step 1 of this, but that's very, very close now. 

So, off for some more testing, although I do have some "real-world" (RW) stuff to attend to today, but at some point I'll come back with a map and full explanation of what I'm trying to do with all this. Regardless, this has been a real learning experience and everything everyone has contributed has helped me immensely (mostly for this scenario, but that will carry over into many other areas as well) so it's definitely not been in vain. I think, given this experience, I also need to get in touch with @BornGinger about his idea of overlays. I agree with @Combatintman that most AI Plans tend to be simple enough to not require it, but I think perhaps as this one evolves (and no doubt others upcoming), this sort of technique will prove invaluable. Were it not for RockinHarry's post here, I would definitely need his insights for that. No one can say there isn't a gold mine of intelligence and knowledge here!

Link to comment
Share on other sites

19 minutes ago, Glubokii Boy said:

I seem to recall that the manual actually has two different explenations for these timings. In one location it mentions something like this:

- Exit before is the first time the AI group will be allowed to start that order

and 

- Exit after is the time by witch the AI group will try to have completed that particular order (reached its objective location).

and in another section it mentions something like.

- Exit before is the first time the AI group will be allowed to start that order

and

- Exit after is the time the AI group WILL start that particular order even if the movement has not been triggered by other means (terrain trigger or AI-order trigger).

 

In my experience the later explenation seems to be more correct and the way it actually works.

And...Sorry ! It's not my goal to try and correct you Combatintman but i belive that the timing options are limited to half minute increments...(maybe you ment to write minutes and not seconds ? 😇)...Or is it indeed possible to set these timings at the desired second level ? If so...How ? 😎

 

 

 

No problem mate - to be honest I don't really understand the nuances of them or whether they have changed and it would appear, having just looked at the CMCW editor that they are 'exit between XXXX and XXXX in 30 second increments.  Around the time of the Market Garden module there was a huge debate in the Beta Board between some experienced designers about how they actually worked.  The end result of the conversation was half a dozen designers saying 'well this works for me' or 'that works for me' ... 'I never use both exit before and exit after' ... 'I always use both exit before and exit after'.  Now that isn't particularly helpful for people trying to learn this.  With regard to the timings I used - they were for example rather than actual increments.

So where does that leave us ... well it is a combination of finding what works 9 times out of 10 when testing for any given AI moves/actions that you want to put into your scenario.  The trick is to test and adjust as necessary.  I have no hard and fast rules for using the timings - I start with the premise of what I'm trying to do and seeing whether the settings I have put in actually work.  If they don't I change them until they do or find that thing just isn't going to work the way I want it to and change the premise or find a different way of doing it - such as employing a trigger (and I don't use them often by the way - because it usually takes me half a dozen attempts to remember how to use the bloody things).  For instance my two CMCW scenarios, Skirmish at Sichenhausen and Rumpenheim Rumpus have no triggers at all.  Rumpenheim Rumpus is has more than a battalion moving - so it is possible to use timed moves alone if you know what you're doing.  The trick here being to plan the thing so that you have an understanding of where everything is going to be in time and space.

Link to comment
Share on other sites

In my tests here, I've found the timing controls to be quite variable. In the first 10 tests I did, I had Tank A1 starting at 00:00 / 00:00 (Exit After / Start Before) and A2 starting at 00:30 / 01:00. In every case, both started at exactly the same time. I cannot tell you why. I checked that the tanks were, in fact, in separate groups and that they were A1 and A2, but still it didn't matter. Tank A3 did start after the 01:00 / 01:30 correctly. Go figure ;) .

NOW, @Combatintman, you just may have solved my quandary about Step 2 in my little plan :) . Triggers! I used them in "Night Moves" (coming to a "theatre" near you soon, I hope ;) ) and they worked very effectively! That might just be the ticket for this one, as long as a trigger will intercept and replace the initial order set. We'll see! I'm optimistic (hey, I said nothing said here is going to waste ;) ).

Edited by Canuck21
Link to comment
Share on other sites


The trigger function is indeed a nice feature. It comes in 2 flavours. The terrain triggers and the AI-order triggers. Both are very useful imo.
Up until recently i have only used the terrain trigger. I thought the AI-order trigger seemed a bit more complicated but with my latest scenario (The Bad Boys, CMSF2), released a few weeks ago, i decided
to take the plunge and try to add the AI-order trigger into the mix. I'm glad i did ! this is a really nice feature. The way it works is that you can specify any order you give to any particular AI group to work as a trigger for other AI groups (select - CAN TRIGGER in the dropdown menu). This trigger will be
activated as soon as that AI group starts to perform that particular order. This can be very handy when trying to coordinate the moves between your AI-groups
I guess it can be argued if it is the best choise to have these triggers activate when the AI group starts its order (as it currently works) or if it would be better to have the trigger activated when the AI group have actually finished that order. I'm currently of the opinion that the second option would be better
but i guess that this might depend on the situation.

Using only the game clock to direct the AI units movements will work well in most situations i guess but ones the AI troops comes under fire things can get messed up and the AI force may well lose its coordinated moves as some part of that force might fall behind due to taking fire. those units will simply not be
in the right place at the right time as originaly set out in the AI plan. Without any triggers to check this the other units in the AI force will still continue with the original plan (and timing) regardless of the fact that some vital parts of that plan might be missing due to being delayed by enemy fire.

One solution to safeguard against this or atleast improve the likelyhood that tha AI plan will remain in a coordinated fashion even ones the AI force comes under fire is to include these triggers imo. 

The difference between using terrain triggers and AI-order triggers is that with the terrain trigger ANY friendly unit will be able to set-off that trigger witch might be what you want in some situations. With the AI-order trigger you as the designer will be able to specify a particular unit (AI-group) to actually be in a particular location
in order to spring the associated trigger. Dependant on the situation one or the other might be better.

For example...Lets say you have 4 tanks advancing (each in their own AI-group)

- Tank 1 move out first with a dash order. It reaches its first location and switches to a static max assult order. This max assult order is set to CAN TRIGGER.
- Tank 2 and 3 are set to WAIT FOR TRIGGER and in the triggerlist you will now find - AI GROUP 1, ORDER 2 - (the max assult order from above) as an option to select for the trigger. Selecting this will see tank 2 and 3 wait at their current location until tank 1 starts its max assult order at its new location and
then they inturn start to perform their order (move to their first location).
- Ones tank 2 reaches its location it in turn will skip to a static max assult order set to CAN TRIGGER. This trigger - AI GROUP 2, ORDER 2 - is the used to send tank 4 on its way. This tank have been waiting (wait for trigger - ai-group 2, order 2) at its starting location. And now moves out.

Using these AI-order triggers will ensure that the other units are actually in place where they are supposed to be before the following units move out. Things stay coordinated even if one or more of the first tanks gets delayed by enemy fire. the following tanks will wait until the first ones are actually in place.
In the example above it might have been just as good or even better to use a terrain trigger for one (or more) of the moves. But this was just a simple example. 

But how long should the following tanks wait if some other tank do get delayed ? forever ? What if one of the tanks gets destroyed  and never makes it to its intended location ? Well here is where things gets a bit more complicated and the exit before and exit after times comes into play. 
Simply using the AI-order triggers and no exit before and after times set will not work. 
If one of the tanks gets destroyed the other ones will be waiting forever. Or even if the first tanks gets delay for...to long...the others will still be waiting for it to reach its location and the whole attack might get hopelessly delayed.
Adding the exit before and after values will be essential for this to work and prevent any complete failures of the AI plan.

The first value...exit before set the first possible time that the AI-group will be allowed to start this order...regardless of if the other unit, that have been set to trigger this unit, has started its order or not. The waiting unit will not start its order until this time is reached no matter what.
The exit after set the time that this unit will be forced to start to execute their order...regardless of if the triggering unit has indeed started its order or not. So this exit after will be sort of a safeguard to ensure that the attack will carry on even if the triggering unit fails in its mission (dead, delayed to long).

Combining triggers (terrain or AI-orders) with a working set of timings will result in a really well functioning AI plan i belive. Even when under fire. Or atleast as good as the current game engine will allow.

The challange is to get all these things to work together though...It's not all that easy or even neccesary i guess but if done right you will have a nice AI-plan.

The way i did it in The Bad Boys scenario was that i first designed the AI plan using only the timing options and rewiewed it in scenario author mode with the player forces set to hide to ensure that the AI would meet no opposition. Ones the AI plan worked as i liked using only timings and unopposed i added
triggers to the plan. Both AI-order triggers and terrain triggers. I then tweaked the initial timings of the orders that had triggers added to them to allow for exit before slightelly before the intially set times and exit after to slightelly after the intially set times for those orders. Then i played
the 'player' side in scenario author mode to se how things played out and tweaked some of the order-timings as needed. This actually was not all that complicated that i had feared that it might have been and i'm really happy with how the AI-plan turned out. The AI in this scenario is capable of conducting a two pronged
attack followed by a fighting wihtdrawl in a rather nice way imo.

Link to comment
Share on other sites

Oh wow, I think you're on to something here! I've seen those triggers as well, but also haven't used them. Still, this has some serious possibilities. Which module is "Bad Boys" in? I'd like to play it, then afterwards have a look at the AI plans if that's ok with you?

Really interesting post here - thanks for this. This is getting better and better :) .

Link to comment
Share on other sites

hello..

The scenario is for CMSF2...link below.

The Bad Boys - AI lead US troops tries a snatch/rescue mission against local ISIS forces in Syria (new scenario). - CM Shock Force 2 Maps and Mods - Battlefront.com Community

Use the link in the last post. The updated one 😎..

You are offcourse free to have a look at it in the editor. Keep in mind though that i'm not one of the 'scenario designing pros' but rather a happy amateur. 😉

The AI plan do work though..atleast in my testing. Hopefully it will help you with some of your ideas. 

Keep on struggeling...scenario designing gets more intresting the more you learn ☺️.

 

 

Link to comment
Share on other sites

Nuts, I don't have that game. Nevertheless, I'll be picking your brain regardless :D . I'm working on Step 2 now that I have the units moving in an orderly manner solved (thanks to all the help I've had here!!), and I can easily see a very good application of exactly what you described above wrt the AI Triggers. I'll be using both types from the looks of it, and the AI Trigger function will be a really key feature in this set. I'll be back (where have I heard that before 🤭?). 

Edited by Canuck21
Link to comment
Share on other sites

The good old suggestion...To start simple ! certainly holds true here when trying to learn triggers and timings.

What i'd suggest you do is something like this...

- Start a new scenario

- Use the map you originally start with when opening the editor without adding anything.

- Add a single platoon to one of the sides and asign each individual squad as its own AI group.

- Add a small terrain trigger.

With this set-up try to get the AI to move this platoon forward in the way you want...

Try using the AI-order triggers to get the squads to advance in a staggered fashion one after the other.

experiment with the exit before and after times to see how those changes play out and how they allow or not allow the squads to move forward. When you got the platoon advancing nicely using the AI-order triggers add the terrain trigger to the mix and see if you can get that to works the way you want...

 

With a simple set-up like this you can run many tests quickly and change any detail you like in a very short time.

 

Link to comment
Share on other sites

That's almost really what I have with this scenario. Actually, I've tried out the terrain trigger and it's working perfectly for what I want. That will be for the lead tank orders, then when he gets triggered, I'll use that to trigger the others. I think that will be exactly what I need, but so far, so good :) . I have used terrain triggers before and those I understand fine. I just have to coordinate a few things with it. Now, one question I could ask you here is, which takes precedence - a timed order (such as "go to 'action spot Y' after 05:00 and before 05:30) OR, a trigger order (there are bad guys over there, go to "action spot X" and shoot them)? Let's say the trigger order occurs at 04:59 and then of course, the time for the next order comes up. Which one will get used?

Link to comment
Share on other sites

2 hours ago, Canuck21 said:

That's almost really what I have with this scenario. Actually, I've tried out the terrain trigger and it's working perfectly for what I want. That will be for the lead tank orders, then when he gets triggered, I'll use that to trigger the others. I think that will be exactly what I need, but so far, so good :) . I have used terrain triggers before and those I understand fine. I just have to coordinate a few things with it. Now, one question I could ask you here is, which takes precedence - a timed order (such as "go to 'action spot Y' after 05:00 and before 05:30) OR, a trigger order (there are bad guys over there, go to "action spot X" and shoot them)? Let's say the trigger order occurs at 04:59 and then of course, the time for the next order comes up. Which one will get used?

There´s always ONE time frame for orders to get executed. The "after" and "before" (or "exit between" - "...and" as it´s now). If setting 05:00 to 05:30 unit can move at 05:00 at the earliest and will move off at 05:30 anyway. So a trigger sent at 04:49 gets ignored as is any later then 05:30. You can set trigger activation range according to play testing results then. If you want the trigger to kick in earlier, then set "after" ("exit between") to earlier than 05:00. If you want the unit move (or behavior change) to solely depend on a friendly trigger, then set earliest execution to 00:00 (mission start) and latest ("before", "...and") to 1 minute after known mission end time. (depends on "strict" or "variable" mission time setting).

So if the triggering unit never starts its triggering order (gets destroyed or routed before etc.) then the trigger recipient will never execute its next order in THIS case. So then might be usefull setting "before" ("..."and") to a time where the next order should get executed in every case. The bailout time from trigger dependency so to say.

If you´d like seeing every possible trigger beeing used in an AI attack mission then check this one out. Makes use of both, friendly and enemy terrain triggers, as well as unit triggers. Also used is "retreat", "face" and "suppressive fire" option in various combinations.

https://www.thefewgoodmen.com/tsd3/combat-mission-battle-for-normandy/cm-battles-for-normandy/cmbn-v4-mg-vp-you-enter-germany-introduction/

 

Link to comment
Share on other sites

Well, after 3 8 hour days of experimenting and trying all sorts of combinations, I don't think what I want to accomplish is possible at this time. It may well be I'm not catching on to something, but what I was shooting for (if you'll pardon the pun) was an all out tank battle in the desert. Now the scenario would work (I suspect) extremely well H2H, but with AI it seems both sides just sit back and shoot at each other until there's only 1 left standing (maybe). There's no jockeying for position (unless you program it, and I have no idea how you'd program that since the human player could simply counteract any movements the AI made, and pick the AI guy off). To be honest, I'm not terribly familiar with armoured tactics so maybe it was a case where they grab a position and fire away, I really don't know. I would have thought you'd try to get a flanking or better yet, a rear shot if possible, or defensively try to get hull down, but again, I just don't know. I'm not even sure CMxx is set up to do all tank battles. So, I'm going to work this one a bit more but if/when it gets released, it will likely just be a H2H scenario. It is, obviously, purely fictional as the desert war was over before Operation Husky, which is when CMFI started, so perhaps there won't be much interest in it anyway. If anyone knows me to be wrong about CMxx not being suitable for tank battles, please say so. I may revisit this down the road a bit and tweak it somewhat.

Regardless of all this, the help I've received here will pay dividends in spades for other scenarios I'll make down the road, which will be more combined arms situations rather than just tanks. So, there was no time wasted in this thread, as far as I'm concerned. I learned a huge amount about how to program the AI that will carry over to all my scenarios, so I thank everyone who contributed here. Believe me, I've read, or am reading everything here and it's helped immensely, as I say. Now if you'll excuse me, some guy from the Department of the Environment for Tunisia wants to talk to me. Something about a littering issue that he seems to think I have something to do with? :D 

Link to comment
Share on other sites

As the game currently stands there is an *optimal* scenario size. Make your scenario too small and one mistake can end the battle before it starts. Make the scenario too large and you run out of AI orders slots for your units and need to compromise on how you issue your AI movement commands. Make the scenario too complex and you'll be unable to write AI orders that anticipate what the player is likely to do. You can be ambitious in your scenario making and oftentimes they work well. Just keep in mind ambitious scenarios come with a downside.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Unfortunately, your content contains terms that we do not allow. Please edit your content to remove the highlighted words below.
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...