As part of my investigation into the use of ChatGPT in wargaming …
… I asked it to design me a set of PORTABLE WARGAME rules for the period 1880 to 1910, and what follows is the result.
Portable Wargame Rules: Colonial & Early 20th Century Warfare (1880 to 1910)
A simple, fast-playing, and portable wargame system for small-scale battles.
1. GAME OVERVIEW
This set of rules is designed for quick and engaging wargames set between 1880 and 1910, covering conflicts like the Anglo-Zulu War, Mahdist War, Spanish-American War, Russo-Japanese War, and early colonial skirmishes.
The rules prioritize playability while maintaining historical flavour.
Scale & Setup
- Map/Grid: Hexagonal or square grid (8 x 8, 10 x 10, or larger)
- Unit Representation: One unit represents a small detachment, platoon, or company.
- Dice: D6-based mechanics for simplicity.
2. UNIT TYPES & STATS
Each unit has the following attributes:
- Strength Points (SP): Represents combat f fectiveness.
- Units are eliminated at 0 SP.
- Firepower: Maximum range and fire dice roll.
- Movement: How many spaces the unit moves.
- Morale: Affects retreat and combat resilience.
Unit Type | SP | Firepower | Range | Movement | Notes |
---|---|---|---|---|---|
Regular Infantry | 4 SPs | 1D6 | 2 grid areas | 2 grid areas | Core unit, steady fire. |
Native Infantry | 3 SPs | 1D6-1 | 1 grid area | 3 grid areas | Fast but less powerful. |
Cavalry | 3 SPs | 1D6-1 | 1 grid area | 4 grid areas | Charge bonus. |
Artillery | 3 SPs | 1D6+1 | 4 grid areas | 1 grid area | Indirect fire, powerful at range. |
Machine Guns | 3 SPs | 2D6 | 3 grid areas | 1 grid area | Devastating firepower but immobile. |
Commanders | - | - | - | 2 grid areas | Boosts morale & activation. |
3. TURN SEQUENCE
Each turn consists of the following phases:
A. Initiative Phase
- Both sides roll 1D6. The higher roll decides who goes first. Re-roll ties.
B. Activation Phase
- Players alternate activating units (move/fire or fire/move).
- A unit may either:
- Move & Fire
- Fire & Move
- Charge into Melee (Cavalry & Native Infantry only)
C. Movement Rules
- Terrain affects movement (e.g., hills cost +1 movement, rivers require a bridge or crossing).
- No diagonal movement on squares; hex grids allow full movement.
- Cavalry may retreat 1 space after melee combat.
4. COMBAT RULES
A. Shooting
- Roll 1D6 (or more based on unit type) to determine hits.
- Target hit on 4+ (unless modified).
- Cover reduces hit chance (-1 to dice rolls).
- Condition Modifier:
- Target in cover -1
- Long-range fire -1
- Command nearby +1
- Example: British Regular Infantry fires at Native Infantry in cover. Rolls 1D6, but applies a -1 penalty.
B. Melee Combat
- Both players roll 1D6.
- Higher roll wins; loser retreats or loses 1 SP.
- If the result is a tie, both lose 1 SP.
- Cavalry gets +1 when charging.
C. Morale Checks
- A unit reduced to 1 SP rolls 1D6. On 1-2, it retreats; on 3+, it holds.
- Commanders nearby add +1 to morale rolls.
5. VICTORY CONDITIONS
- Control the battlefield: Hold key objectives for X turns.
- Break the Enemy: Reduce the enemy's force to below 50%.
- Survival: Defend a position until a set turn.
6. OPTIONAL RULES
A. Command & Control
- Commanders can activate 1 additional adjacent unit per turn.
- Without a commander, units require 1D6 roll of 3+ to activate.
B. Special Weapons
- Gatling Guns: Roll 3D6 but jam on a double 1 (misses next turn).
- Bayonet Charges: Infantry get +1 in melee when charging from 2 spaces away.
7. EXAMPLE SCENARIO
Colonial Defence Situation: A British column is ambushed by a large force of native warriors. The British must hold their ground until reinforcements arrive.
- Map: Village in the centre, river on one side, hills on the other.
- British Forces: 3 Regular Infantry (4 SP), 1 Artillery (3 SP), 1 Machine Gun (3 SP), 1 Commander.
- Native Forces: 6 Native Infantry (3 SP each), 2 Cavalry (3 SP each).
- Objective: British must survive 8 turns without losing 50% of their force.
8. FINAL NOTES
- This system is easy to expand with more unit types (early tanks, armoured trains, etc.).
- Designed for solo, two-player, or small group play.
- Play on a small board (8 x 8 grid), making it portable and quick.
I find it interesting that ChatGPT has not just copied the text and mechanisms from the published versions of the PORTABLE WARGAME. For example, it has added Morale rules, and the Melee rules are very different from those I have written. It has also included ideas for Victory Conditions (something that I know some wargamer like) as well as some Optional Rules that have rather piqued my interest.
I was intending to play test Joseph Morschauser's FRONTIER rules in the near future, but I am now giving serious consideration to shelving that idea and play testing these rules instead.
Hi Bob
ReplyDeleteIt will be very interesting to hear your thoughts after you have had an opportunity to play test the computer rules. In particular any changes that you feel improve your own rules, but even more interesting any thoughts on those that you feel have not done so.
I think you are very brave to pit your own thoughts on rule writing against a computer created set.
I am a firm believer that there is no such thing as a perfect set of wargame rules. I believe that most wargamers develop their own ideas of what is "right", and that these often change dramatically over years of play testing. For me that is the beauty of wargaming, it is almost a living thing in the way that it develops within your own mind. Therefore I have little confidence that a computer can produce anything more than a compromise of all the rules contained in its memory.
But it may well be able to create a widely accepted set of wargame rules which take into account the most popular accepted elements. And in doing so could well come close to creating a large group of supporters.
It will never be for me, but I will still be very interested to reads your comments in due course
regards
Paul
Jan Leniston (Paul),
DeleteAt first glance, the rules that ChatGPT has created (I am not sure that I could use the term written in this case) certainly seem to work … but they are not complete enough for most wargamers. There certainly need some clarifications in places, and it mentions long-range fire but does not explain what constitutes long-range.
I had expected that what ChatGPT would create would be little more than a copy of bits drawn from my published rules but it did generate a few interesting changes that I’m sure it has ‘borrowed’ from other rules but which - nonetheless - are certainly worth further examination.
It was seeing ChatGPT ‘play’ a Matrix Game that made me realise that it had a potential role to play in wargaming, particularly for solo wargamers like me. It is developing and improving all the time, and I’m sure that within a year it will prove to be a worthy opponent … should I choose to go down that route.
Will ChatGPT replace wargame designers and rule writers at some point? That is an interesting question that I suspect will be answered in due course.
All the best,
Bob
That looks interesting, it might be worth some development. I would be interested in your thoughts after a play test. Incidentally, that is the mechanic I use for close combat in my Portable Wargames.
ReplyDeleteMark Cordone,
DeleteI agree that the rules are worth looking into.
The Close Combat system is interesting, and as you are already using it, I assume that it works effectively.
All the best,
Bob
It allows for faster combat resolution. The way I do it all relevant modifiers for terrain etc. Are applied, and if there is a tie the combat continues into the next players turn, but neither side loses any strength points until the combat is resolved. For 3x3 games only light infantry or cavalry have the option to retreat instead of taking a strength point loss.
DeleteMark Cordone,
DeleteThat’s a very simple and elegant mechanism! If you don’t mind, I will copy it!
All the best,
Bob
An interesting set of rules. How long, and how many tries did it take to get to this? A good starting point with plenty still to polish up to taste. I look forward to seeing what else you get up to with this kind of experiment.
ReplyDeleteWEK 3,
DeleteChatGPT produced these rules almost instantly in reply to the simple question ‘Please write a set of Portable Wargame rules for the period 1880 to 1910’.
They do need a bit of work to tidy them up, but even as they are they are perfectly usable.
All the best,
Bob
BOB,
ReplyDeleteA good set of rules- straight forward and easy to read and learn.
Kev Robertson (Kev),
DeleteI agree. They certainly are a simple and hopefully easy to use set of rules.
All the best,
Bob
Hi Bob,
ReplyDeleteA while ago I played around with an AI for roleplaying games. I found that it did it's best work when I didn't accept it's first attempt, but queried parts of it or made suggestions. Then, after it had improved it, the final results were often rather good. However, when I tried it for historical wargaming, it informed me that the Incas rode llamas into battle as cavalry, so I learnt to deeply mistrust anything it presented to me as fact!! I gave up on it but perhaps the latest versions have improved.
Steve
Steve (Steve),
DeleteI’m interested to read that you’ve also tried using AI for role play games … and thanks for your advice about not taking its first suggestions. I understand from the person who ran the session at VCOW that it will learn the more times it ‘plays’, and that giving the program as much information as possible helps to ensure that it won’t stray too far from reality!
All the best,
Bob
I've tried to use ChatGPT to generate some rules a couple of years ago. The rules it suggested looked good on the paper, but when I test-played them they didn't work so good as their first impression. When posting this in some wargaming FB-groups I got the idea that some consider the use of AI as controversal as they consider what AI suggests as 'theft'. You asked a question and it suggested your own rules, so what do you think?-Is it 'Theft'??
ReplyDeleteI used the ChatGPT to generate the rules for the 'Wargame in a Matchbox-Challenge'. They also looked good on the paper, and I tweaked them a little with the use of AI. The rules kind of looked good, but the more I play them, the more I think that they are just too simple...
As I'm mostly a SOLO-gamer I've used ChatGPT as an opponent. Be prepeared for a lot of writing, but it's very unpredictable as an opponent. It did actions I wouldn't if I'd play both sides my self.
Now I usually uses ChatGPT to suggest wargame scenarios, names for wargames and 'cover-photos' for blogger posts.
I'm looking forward to read your experiences playing the AI suggested version of your own rules.
Best wishes.
Roger,
DeleteI get the impression from people who are much more knowledgeable about these things than me that ChatGPT and similar programs are improving on an almost daily basis, and that if I was to ask it to design another set of rules to the same parameters in a month, the results would be better.
The ‘theft’ of intellectual property is nothing new to me. Far as I’m concerned, I know that several so-called ‘free information’ organisations have made the text of my books freely available online without my permission. I was even accused by one of being a ‘rich, fat cat author who was making money from people who could not afford to buy my books’.
If one looks at almost all published rules, they tend to borrow ideas and mechanisms from previously published sets of rules. Is this ‘theft’? Possibly … but plagiarism is probably a better way to express it.
It sounds as if the rules you asked ChatGPT to develop were more of a starting point for the set that you’ve ended up using, and if you then continue to develop them into something better, they will become more human than AI in origin.
I must admit that I have yet to use ChatGPT to act as an opponent, but I will bear your useful comments in mind. As to producing scenarios etc., that sounds like a very useful way forward that I need to investigate further.
All the best,
Bob
Can we find a way for the AI to *play* the game? That would be interesting! Google's AI Studio can "see" your PC desktop so I tried showing it a PC hex 'n' counter wargame with the intention of asking it to suggest some next moves. The results weren't very satisfactory but it's early days. AI Studio is free if you want to try it yourself.
ReplyDeleteZardoz,
DeleteThanks for the suggestion. I suspect that as AI improves - and it seems to be doing so at a phenomenal rate - using such programs to act as an opponent in a game will become much more feasible. When that happens, it will be interesting to see what impact it will have on wargaming.
All the best,
Bob