So let’s actually try to play this thing (the rules are on my previous AI post here). After reading through them a few times I realised a lot of the issues could be solved by playing with single figure units like a small scale skirmish game, rather than the squad based game I had initially imagined. So here is my Imperium force made from my Chaos collection:


And here are my Mechanised Legion:


Both forces came to around 200pts each and this seemed like a good level to play at, taking an hour to get through.
Without a proper scenario, I made up something simple. 3 objectives were on the board. Whoever held most at the end of turn 4 would win the game.

I immediately had second thoughts about the turn structure:
Each turn is divided into phases:
- Initiative Phase – Players roll a D6; the highest roller chooses who goes first.
- Movement Phase – Players take turns moving their units up to their Move (M) value in inches.
- Shooting Phase – Players take turns attacking with ranged weapons.
- Melee Phase – Units in base-to-base contact engage in close combat.
- Morale Phase – Units that lost models check for morale.
It’s clear that this is a mashup of a 40k Igo-Ugo style turn and an alternate unit activation style turn. While the roll off suggests a 40k style turn, the description of the individual phases suggests alternate activations. My initial instinct was to have alternate activations within those four phases, with the higher roller going first in each phase. But thinking everything would get bogged down, especially with single figure units, I opted to have each player use all of their figures at once in each phase, so simply put:
Roll off for side A and B, then:
Side A Moves, Side B Moves
Side A Shoots, Side B Shoots
Side A Fights, Side B Fights
Morale Tests.
This worked well enough, and seems like the middle compromise between the three possible ways to interpret the rules. This would be a pointless exercise if I help the AI too much, so I’m trying to stick as closely as I can to rules as written.

Of course it’s difficult to tell after only playing a single game, however I did start to think that the Imperium may be a little under powered. The Walkers were a particularly tough unit for them to deal with.

The Walkers begin at defence 5, so you need 5+s to hit them. However if you shoot, they gain +1 defence because of their Energy Shields, but then if they also have cover their defence also goes up by 1. Also… if they move half or less, or just stay put they get another +1 defence.
This is largely why I added rules for rolling 7+s on D6s. Rather than just having it be impossible. However even with that rule, it was difficult for the Imperium to even hit them, let alone kill them.

Once the Warriors had closed the distance they fared a little better in melee, however the Walkers were still brutal in melee.
While the fight in the centre raged on, the Assault Templars got gunned down by the drones on one flank and the Sentinels and Striders tied themselves up in melee on the other.

Here’s where things wrapped up. The Striders had killed the Sentinels and were working their way back to the centre. The Middle was contested, by mostly held by the robots, and the Mechanised Legions home objective was secured by Drones. It was a pretty convincing victory on the Legions part.
The scattered Space Warriors was due to the morale rules. When a unit is at half strength (I went with half wounds), they take a morale test. This may have them carry on as normal, carry on but suffer negatives to their attacks, or retreat to the closest board edge. And ‘closest’ is what caused issues, as units may try to retreat through the enemy lines (which is what happened here), so I think they should probably head back towards their starting edge, or to speed things up maybe just vanish entirely.
It was a fun enough game, but that was probably to be expected. Once Donald Featherstone had figured out ‘roll to hit, roll to wound’ there’s a reason it’s been used in 95% of wargames, because it works and it’s just fun. But this did feel like I playing One Page Rules… maybe too much. I’m tempted to fix the core rules and have my own 40k lite set to play with, but I’m not exactly feeling that buzz to get building and painting armies specific to this like I have with Havok. I think this is because when it comes to oddities of various other rule sets I know a person made their odd decision for a reason, they play tested something, or disliked something in another rule set. When it comes to AI, when something is odd… you just stare into the void. Did it make that choice for a reason? You simply can’t know. And if it did, make that decision for a reason, was it to make a better game, or was it to mimic some rules it doesn’t actually understand?
This all does make me wonder though… will we get to the point where a company produces a solid core set of rules and then you could go to their website and have an entire faction, unique to you generated with AI, including the 3D sculpts, which you could download and print yourself or have printed and sent to you. It would be a pretty whacky idea, but kind of cool… a sort of infinite wargame universe… provided the AI wouldn’t just produce billions of slightly different space marines and tyranids.
Fascinated by the concept of the AI designed bespoke faction for you
Wonder if it would write the background fluff as well 🙂
Interesting stuff.
I find the whole AI thing fascinating but a bit scary. Having wargames rules written by AI and them (more or less) working is fascinating.
I do kinda like your personalised army idea, I wonder how long we’ll be waiting for that?!
I reckon a company will attempt it within the next 5 years. I mean not having to write rules or sculpt miniatures… if it really took off, you’d just be printing money. Although I’m guessing you’d need to be the first one out of the gate, as the market would probably get flooded immediately.
I would imagine so. Might be difficult to not have it contradict itself across multiple armies, but that never stopped GW.
In theory AI should have the memory to remember what it said and not repeat it but I suspect that it doesn’t understand enough to understand contradictions
Yeah it’s been very much like that with the rules. It doesn’t seem to comprehend what they actually mean in practice.
Fascinating 🧐