Hi everyone! I'm proud to unveil the Solo Mode for Armada that I've been working on. It's a series of cards that create a deck which will drive the AI of an opponent, so you can play solo. All you need is a printer and some scissors, print it out, cut out the cards, and pair them up with some sleeved cards to create a deck... and you're ready to play! I recommend giving the AI player a handicap, perhaps an extra 50 points to make up for the fact that some of their choices will not always be smart for the AI. The PDF and rules are available on my website here:

http://www.crabbok.com/custom-content/armada-ai/

Along with other custom content I've created, such as the Nunchuck expansion, and Deep Space Nine.

Here's my video explaining it as well.

I'd love to hear what you think! Hopefully it'll help those of you who have that itch to play but sometimes have trouble finding an opponent.

It uses a deck of cards which help you handle virtually any situation that might arise during gameplay!

It can likely also be used to facilitate a cooperative game, where you and a friend team up against an opponent. Only issue I'd have right now is if the AI had more than 12 ships to control - in which case you'd have to come up with some other way to randomize activation.

So, I'm a bit curious, how does this balance against a person? The way I understand it, the cards make most of the AI actions random, I can't imagine that makes for a strong opponent.

Cool idea. !

It might be nice to make some sort of tiered or random choices when given equally good options.

--

Sometimes I feel like I just need some deployment options also. And a bit of randomness or help picking targets during simulated solo play

It is meant to be a dumb opponent more-or-less, and isn't EVER meant to replace a human opponent.

It counters this weakness two ways.

First, in that you'll give the AI more points than the Human, (for example, 450 vs 400 or even more)

and Second, there are bonuses that will happen on the cards. Additiona repairs, rerolls, and other things that will make this a different experience than a standard game.

I can't say I'm not curious. I'll definitely have to check it out. Thanks for sharing man!

I like the idea of the bonuses the AI can get. FFG did pretty much the same thing with their AI in the Imperial Assault App.

Hopefully I can find a time to try it out soon and give you feedback.

3 hours ago, Darth Sanguis said:

So, I'm a bit  curious, how does this balance against a pers  on?    

At first I thought the same but then this is not meant to replace another person but rather it is here to replace yourself while playing solo.

I tried to play Armada by myself few times but I know my next move so it wasn't that much fun. Now with this random decision device, it will be a bit more interesting and uncertain.

thanks

Cool I play Sails of Glory, the guys there made the same sort of AI and they work well.

Good idea

Awesome work @Crabbok ! Thanks for this!

Got two questions(hope I understood the rules right) As golden rule no. 2 is to not ram own ships, does the AI ship drop it's speed to 0 if it could avoid a collision with a friendly while doing so?

How about having AI ships maneuver toward the yellow dot of the front hull of their target ship, instead of towards the base of the target ship? May not be significant for small target ships but it could help the AI to lead itself abit when maneuvering toward larges or mediums. It's a pinpoint position to work with rather than a space covered by a base too(although I believe u mean the centre of that target ship when maneuvering towards it?)

Thanks again! Great content 👍

16 minutes ago, Muelmuel said:

Awesome work @Crabbok ! Thanks for this!

Got two questions(hope I understood the rules right) As golden rule no. 2 is to not ram own ships, does the AI ship drop it's speed to 0 if it could avoid a collision with a friendly while doing so?

How about having AI ships maneuver toward the yellow dot of the front hull of their target ship, instead of towards the base of the target ship? May not be significant for small target ships but it could help the AI to lead itself abit when maneuvering toward larges or mediums. It's a pinpoint position to work with rather than a space covered by a base too(although I believe u mean the centre of that target ship when maneuvering towards it?)

Thanks again! Great content 👍

Great questions!

1) Yes, if reducing to speed 0 is the only way to either A) avoid leaving the board, or B) ramming a friendly, they would do that, but then again, only if they had a NAV command, or token, capable of doing it.

2) Good suggestion. I've been playing it as the center mass of a ship, but I'll try focusing on the forward dot next time when applicable. (Often you are far away from your target so center mass and forward dot are virtually the same).

Thanks!

This does look interesting!

Actually, what I'd be really keen to try is AI vs AI, remove the bonuses and see what happens...

Played a game today using this system, and I enjoyed it. Certainly beats playing both sides!

A couple caveats up front: I'm not the most experienced player so ignored the bonuses, I used a thematic setup that in hindsight probably made the game easier for me based on the AI's limited movement choices, and I played without upgrade cards. My fleet (498 Pts) was an Ackbar MC80 Comd, MC75 Armd, Neb B Sp, MC30c Torp, and CR90A with 3 A-Wings, 3 X-Wings, and 2 Y-Wings. The AI (600 Pts) had a Motti Imp II, 2 Victory Is, and 3 Arquitens Comds with 9 Tie Fighters, 4 Interceptors, and 3 Bombers. The thematic set-up had the bulk of my fleet and the AI's Star Destroyers in opposite corners. The AI Arquitens deployed directly opposite my fleet, while my MC30c deployed opposite the AI's SDs. All of my ships started angled toward the Arquitens, while his were all angled toward the bulk of my fleet. I ended up winning by 107 pts, but it felt tighter than that as the AI only needed to damage my MC75 once more to destroy it.

I found the attack priorities to work well in general. Because ships can attack from multiple arcs, I found the AI never had to pass up a juicy target entirely to engage a large or well shielded ship, although it might not always optimally apply coordinated fire commands or dials.

I felt that the AI's movement worked well for ships that like to engage head-on, like Star Destroyers, but was weak for ships that prefer flanking shots, like the Arquitens (or, I imagine, the MC80). For example, on the SD side, a single SD slowed down to engage my MC30c while the rest turned to intercept the bulk of my fleet head-on (makes sense). On the Arquitens side, meanwhile, the AI charged headfirst at my fleet at varying speeds. Initially, that meant they weren't able to leverage their flank shots, then it meant they ended up ramming my ships headfirst (bunch of wannabe Hammerheads!), and finally, it left them isolated at close range on the flanks of my MC80 and MC75. Regardless, I understand why this limitation exists (one deck streamlines the AI's decision-making) and honestly don't know how you could resolve it short of having a second deck with different movement priorities for ships that prefer to flank.

The only rule I didn't really like was the one that prevents the AI from using more than one defensive token on a given attack. I assume it's there to simplify the AI's decision making. Either way, I ignored it a couple of times when I felt it didn't make any sense (eg the AI was defending against the final activation of the round, drew a redirect to mitigate multiple hits, but had an unused brace).

Overall, really fun system to play with and kill an afternoon. I do have two questions @Crabbok :

Can you clarify if I was using the command dials as intended: I played so that the AI player did not stack command dials, but instead whatever card was drawn at the start of their ship's activation was immediately used as its command.
Did you ever playtest with the AI using multiple defensive tokens? I'm considering adding a second defensive token icon to a handful of the AI cards, but would like to hear your thoughts on that.

Thanks for the feedback!

Yeah the limitation of a single defense token is absolutely a weakness, and I realize that it will sometimes be the BIGGEST weakness of the AI - but I think the simplicity gained from that rule-set is worth it, and that's one of the biggest reasons to run higher point lists for the AI. (If they have 200 extra points, the fact that they might not brace doesn't seem all that bad! )

And yes, I believe you used the dials correctly. I don't actually SET dials at all and just reference the card for them - but I wanted to include in the rules that technically you set the dial - just in case a dial-affecting situation comes up.

And no, I haven't done a second token yet. The way I see it, if you want to try that you can do it 2 ways.

1) The lower left section of every token is left blank. This is for future proofing at the moment - allows me to easily add something if need be - a new mechanic, or whatever. For now, you could add an Asterisk to half of the cards, and when you draw an asterisk the AI uses a second token (If they can), and you'd draw more cards until you get another one that works.

or

2) simply decide which defense tokens you want the AI to use. But at that point, it's less an AI game and more a game of You vs Yourself. If you are comfortable making choices that counter your own attacks, then go for it!

This is actually a really fascinating and worthwhile read for anyone, especially the Americans among us.

9 hours ago, Canuck7 said:

I felt that the AI's movement worked well for ships that like to engage head-on, like Star Destroyers, but was weak for ships that prefer flanking shots, like the Arquitens (or, I imagine, the MC80). For example, on the SD side, a single SD slowed down to engage my MC30c while the rest turned to intercept the bulk of my fleet head-on (makes sense). On the Arquitens side, meanwhile, the AI charged headfirst at my fleet at varying speeds. Initially, that meant they weren't able to leverage their flank shots, then it meant they ended up ramming my ships headfirst (bunch of wannabe Hammerheads!), and finally, it left them isolated at close range on the flanks of my MC80 and MC75. Regardless, I understand why this limitation exists (one deck streamlines the AI's decision-making) and honestly don't know how you could resolve it short of having a second deck with different movement priorities for ships that prefer to flank.

What about a draw 2 choose 1 mechanic for movement? Then choose the one that makes more sense to the AI ship in question. There is one "flank" movement card in the deck but maybe 1 or 2 more cards could be changed/added in to increase the chance of AI flanking

13 hours ago, Crabbok said:

And no, I haven't done a second token yet. The way I see it, if you want to try that you can do it 2 ways.

1) The lower left section of every token is left blank. This is for future proofing at the moment - allows me to easily add something if need be - a new mechanic, or whatever. For now, you could add an Asterisk to half of the cards, and when you draw an asterisk the AI uses a second token (If they can), and you'd draw more cards until you get another one that works.

I'm thinking of going with option 1, but with a second caveat - the second token will only be used if it isn't already exhausted. Might prevent the AI from burning through tokens too quickly. Will try to get another game in tonight and share how it goes!

I think if you're going to start adding more "smart" decisions you may want to reduce the point advantage you give the empire. My fear is that at a certain point, you're basically just playing against yourself. This project was meant to be a whole new type of opponent that would surprise you. sometimes in bad ways, other times nasty ways.

{I'll just post here if I may to remind folks of my own AI system - and Solo Campaign}

http://daveswargames.blogspot.com/2018/05/defenders-of-gelron-sector-encounter-at.html (which could work with Crabbok's system)

https://www.dropbox.com/s/4jv0dijeg320j83/Star Wars Armada AI.pdf?dl=0

And the thread in Off-Topic were these things can be discussed...

Nice looking system! I'll be honest I hadn't seen this before. It certainly appears more intuitive and less up to chance than mine - which may even make it more ideal for evenly matched fleets!

Yes - it’s not stated explicitly - but they head towards the focal point in formation.

Happy to talk further about my system in my thread - let’s not interfere further in Crabbok’s.

I haven't had a chance to to try either AI system yet, but my initial thought upon reading this one was that wouldn't it be advantageous in almost all cases to aim for a double-arc shot versus straight charging?

Unless you want to ram/block, which is totally a viable strategy situationally, I'd think going for the double arc is usually a good idea.

I checked and @LTD seems to have a similar provision in his model (which I do plan to try!).

Also, another thought, the AI model works in X-Wing for several reasons, one of which is that each ship has preferred actions for that specific ship. Now given, the movement systems are totally different between X-Wing and Armada, but would it help to add some kind of ship-specific logic to movement?

Eg if you only have reds, don't charge head on. If you have red and blue, try to get only into blue range, etc?

You could probably pretty easily categorize ships into a small number of categories based on their dice/arcs so that movement makes more sense.

Initial thoughts:

Flankers - focus on side-arc shots at range.

Brawlers - focus on close-in shots (double-arc when possible)

Support - attempt to perform shots at range, then move into obstructed position for safety (eg flotillas)

Just finished trying out your system. I'll probably post a full battle report this weekend, but here are some initial thoughts.

For context, I ran a dual interdictor Konstantine fleet against a Rebel MSU list. I tried to run it as true to your rules as possible.

It may have just been my one game, but I suspect control based lists are extremely effective against the AI. Which makes sense as it's self-control is obviously limited. Basically I was able to perfectly manipulate the AI with K speed control and whatnot. Allowing me to cause the Rebels to ram each other like crazy while trundling one after another into the meat grinder. And honestly it was amazing. I found it extremely satisfying to try out a non-competitive list idea and actually get to see it happen in a way that's rare against a real opponent.

On the note of speed control, I used the gravity well to force some of the opponents ships to deploy at speed zero. I just house ruled that if an AI was ever at speed zero it would default to a nav command while still following the other guidance on the drawn card. Otherwise half the ships might have done nothing all game.

Despite how well i was able to control it the AI still threw some surprises at me, like Salvation hammering the side of my flagship all game, using up all of my defensive abilities to keep it alive.

I didn't have any issues with the AI only using one defense token, but it was an MSU, so often only one token was valuable or they'd die either way.

I did tweek the squadron attack priorities. If a squadron had bomber I made it default to attacking enemy ships first instead of squadrons. Otherwise a pair of Y-wing would have been flailing uselessly against a scatter ace. It also made it so that the card bonus that gave enemy squadrons Heavy actually did something.

Lastly, kind of the Golden rule I went by in any areas where the AIs decision was unclear is to always pick the decision that was the worst for me as the player. It's something the Imperial Assault App does, and it worked well when breaking ties between which ship the enemy would Target or how it would move. I didn't super analyze the decisions as much as just go with my initial gut reaction, so it didn't feel like I was just playing against myself.

Another pro I noticed, is that the game seemed to go fairly quick, since I was the only one taking time to plan.

I seriously had a blast playing, but think there could be just a few more tweaks without making things too complicated.

I haven’t tried this system yet, but it will definitely make being an antisocial introvert easier 😂 , and i definitely prefer ur simplicity just for a quick, fun scenario. Props for making the cards look good too!

8 hours ago, NightAngel47 said:

On the note of speed control, I used the gravity well to force some of the opponents ships to deploy at speed zero  . I just house ruled that if an AI was ever at speed zero it would default to a nav command while still following the other guidance on the drawn card. Otherwise half the ships might have done nothing all game.   

This is a fantastic idea and honestly one that I didn't anticipate. Great ruling!