I asked the ChatGPT for a suitable name for the game this AI-service suggested in an earlier blogpost. The ChatGPT suggested to call it's own rules for 'Tiny Troops: Fast and Furious Wargame Rules'. I wonder if these rules lives up to it's name?
I posted a link to my last blogpost about letting ChatGPT generate some easy to play wargaming rules on some wargaming related Facebook-groups I'm following, and it created some debate and seemed some what controversial to use AI for this purpose. I was curious about the suggested rules and wanted to try them out to see if they were so playable as they firstly seemed. To have something to compare them with, I decided to play by the same setup and scenario as I played in the previously reviewed One Page Rules Army Men Combat v2.0.
I usually play my games solo, playing both sides my self as good as I can. Since the ChatGPT was so involved already by writing and naming the rules, I wanted to see if it was (artificial-) intelligent enough to play one of the sides if I fed it with these situation, movements, developments, results and so on. Besides, using the ChatGPT this way may be less controversial than letting it generate the rules themselves perhaps? Since I have a soft spot for the UN-troops as almost 1200 Norwegians participated in this mission I let the ChatGPT play the Katangese Gendarmerie in this game. The ChatGPT needs to be persuaded to join a game as it's AI and don't want to play a game as it actually can't. It will however give guidance and advice for movements and actions for units in the game within the rules, so I'll let it 'guide and advice' all movements and actions for the Katangese Gendarmerie.
The scenario
I've placed this skirmish to the United Nations largest mission in both scale and operational scope during the Congo Crisis 1960-65, when UN troops 'peacekeeping' was more offensive 'peace enforcement'.
Situation:
Two UN vehicles from the UN base at Kamina were performing a reconnaissance mission on the roads between Kamina and the city of Kolwezi. Outside Kambezi they came under heavy mortar shelling from Katangese forces and they were probably both hit as the UN HQ at Kamina cannot reach them by radio anymore.
The UN HQ sends out search and rescue patrols to find the missing reconnaissance-unit, and it's expected that the roads in the area have taken damage and are impassable for other vehicles, and it's only reachable at foot.
The Katangese forces in the area are aware of the possible UN-wrecks and sends out patrols to find them, and to seize anything of value such as documents like the UN SOPs and codes for radio-communications.
Some houserules to this scenario seems to be required though; It takes 1 man 1 turn to search a vehicle, and it takes 1 man 1 turn to examine a casualty. To check a casualty's condition, roll a D6, and if the roll is the same value or higher than the turn number for that turn, the casualty is alive and can be medivaced. Casualties can be medivaced at half speed.
Forces:
- Katangese:
- 1 rifle-squad with 8 members, armed with 1 smg, 6 rifles and 1 LMG
- 1 rifle-squad with 8 members, armed with 1 smg, 6 rifles and 1 LMG
- 1 rifle-squad with 8 members, armed with 1 smg, 5 rifles, 1 sniper rifle and 1 LMG
- UN:
- 1 rifle-squad with 8 members, armed with 1 smg, 6 rifles and 1 LMG
- 1 rifle-squad with 8 members, armed with 1 smg, 6 rifles and 1 LMG
- 1 rifle-squad with 8 members, armed with 1 smg, 5 rifles, 1 bazooka and 1 LMG
Mission:
- Katangese: To find the immobilized UN vehicles and search them for valuable things such as secret UN documents, orders, maps, SOPs and radio-codes
- UN: To find the missing reconnaissance-patrol and save the members of the unit, and make sure no graded documents in the vehicles fall into the hands of the enemy.
AAR
Time: 3 hours (!)
Turns: 10
Result: UN victory
It all started very good, and the Katangese forces advanced towards the UN-vehicles slowly from cover to cover...
The UN forces also advanced very slowly, and due to limited firing-ranges the forces needed to move for several turns before they got within firing-range of each other. One of the UN-squads started to examine the casualties outside the jeep, and started to search the jeep, but was interrupted by the advancing Katangese gendarmerie.
It seems like the Katangese forces focused on trying to defeat the UN forces first, and search the vehicles afterwards. The continuous Katangese fire harassed and interrupted the first aid given to the 2 casualties outside the jeep and the search inside the vehicle it self as the UN had to return fire.
After exchanging fire for a while, the Katangese forces started to take a toll and had several casualties. It seemed like the vehicles was out of reach for them now, and they were fighting for their lives instead. This made it possible for some of the UN-troops to start searching the jeep again, but none of the casualties could be saved by this time/turn.
One UN squad charged into hand-to-hand combat and eliminated one of the Katangese squads.
The few remaining members of the other Katangese squads were soon under so heavy fire, that the remaining 2 Katangese squads were soon eliminated as well. The UN-forces started to search the other vehicle.
While the UN-forces seized all the classified documents from the vehicles the very last Katangese survivor was retreating. -actually more fleeing the scene.
Pros
- Nice mechanics for unsaved hits and outcome of melee. On the other hand these effects probably prolonged the game.
Cons
- Several of the cons were already pointed out in the comment-section section in my last blogpost by readers and fellow wargamers. Firstly the rules don't specify the games turn sequence. I just rolled for initiative every time activation was needed, using yellow dice for the Katangeese forces and Greens for the UN. High roller got to activate a unit.
- When morale tests are needed seems unclear in the rules, and I just followed the example of other rules by testing a unit's morale if it was halved.
- The rules states that a LMG can fire twice per turn, but does not says if it fires two times during a turn or if it just fire once with 2 dice. I played by rolling 2 dice for the LMG to indicate a higher rate of fire to a specific target. On the other hand the HMGs cannot 'fire twice' within a turn, but they fire on a longer range and has a higher hit probability with a 3+. These rules makes it impossible for a HMG to inflict multiple damage on a target, as it probably would do in real life.
- I don't know why the rules says that sniper rifles can only fire once per turn, as all weapons except for LMGs can only fire once per turn.
- The movements are very limited, and especially if moving through rough terrain. This makes the advancements very slow. It doesn't help playing with the ChatGPT either as it's very time-consuming to both read and write with the ChatGPT.
- I think the firing range for rifles are too short, as it equals only 3 fully movements. Even the bazooka has a longer firing-range than a rifle which would not be the case in real life.
- The rules have no benefits for being in cover, and does not make it harder to hit targets in cover.
- No rules stating that heavy weapons needs more soldiers to serve them. Here all weapons can be operated by a single soldier.
Conclusion
It is a lot you can use ChatGPT for; making rules and act as an opponent. I asked if it could write this blogpost and a battle report as well, but the suggested text made no sense, so I had to do that my self.
Despite of the name of the game, it was not so Fast or Furious. -Especially not fast if your opponent is an AI which you need to write and describe every move on the game to, and take time to read how it will react on them when activating. I found this time-consuming, even if some of the comments by the ChatGPT were amusing.
Would I do it again? I think the cons outnumbers the pros for these rules, so I'll probably not doing another game with these rules. I know AI-apps take everything "from here and a little there" found online and combine it all into an answer, but I don't think this game combination was very successful though even if it looked good at first glance. I think the rules have so many basic shortcomings, that I'm not even going to put them among the other free wargame links and download section of this blog. I guess they can be tweaked further to the better by continuing asking the ChatGPT for further rules and details trying to improve its shortcomings. I did not initially do that as I wanted to try the ChatGPTs capacity to initially suggest a ruleset without having to follow up and give too many corrections and further specifications. Perhaps I should?
Would I play with a ChatGPT as an opponent again? -More likely; Perhaps it works better for even smaller games and even simpler scenarios?
To quote the ChatGPT at the end: "Based on the information provided, it appears that the situation is unfavorable for my squad, and your units have gained the upper hand. I accept the outcome, and it seems that your forces have successfully completed their mission. Well played!" (...) "it was a pleasure playing with you".
