AoS Army Harmonizer

[samheim]

I have a few questions:

• How do expect a list harmonizer program to be used?
• Does each player bring all their models, input them into the program, and the program writes a list?
• How would this work at a tournament?
• How would this work for summoning?
• Are you going to use a different list every match of the event?
• Do you input all players list at the start of the event or would this be done match by match?
• Do you feel list writing isn't a skill that needs to be showcased on an individual basis?

1) Players input what models they want to use to play the game with. If you have 1 person with 2000 points of Fire Slayers, and another person would only have a 500 point army. The end goal is that the army harmoniser would provide balanced stats so those 2 armies could play a balanced game against each other.

2) Yes.

3) You say what models you are playing with and then harmonise them. The harmoniser will have keywords like strong or fast. To make sure your flying unit does not become a tank or your light infantry become to tough.

4) Summoning is hard to factor in. Its somewhat subjective what the value is of summoning units at the board edge. This like everything else can be solved, even if no one knows how to factor it in yet. Far harder software problems are solved everyday.

5) You can use it however you like.

6) Either way. That would be for the user to decide how they want to use it. The harmoniser only harmonises.

7) No, the harmoniser makes games more exciting by first balancing the game more, and then providing generals with unique and impossible to foresee circumstances, like in real life. No longer can they always count on their kill battle line having a 3+ hit or always activates first. They have less potential, to cheese because the warscrolls are generated dynamically. In a tournament decisions would need to made on the spot! based on the new unit stats.

At the moment, its just who has best kill list and wins. This is more about, who is best at thinking fast and being good at strategy.

 

[Canas]

I want hard points values for everything

Which is impossible & pointless, as points are entirely dependent on context.

 

[LizardWizard]

1) Players input what models they want to use to play the game with. If you have 1 person with 2000 points of Fire Slayers, and another person would only have a 500 point army. The end goal is that the army harmoniser would provide balanced stats so those 2 armies could play a balanced game against each other.

2) Yes.

3) You say what models you are playing with and then harmonise them. The harmoniser will have keywords like strong or fast. To make sure your flying unit does not become a tank or your light infantry become to tough.

4) Summoning is hard to factor in. Its somewhat subjective what the value is of summoning units at the board edge. This like everything else can be solved, even if no one knows how to factor it in yet. Far harder software problems are solved everyday.

5) You can use it however you like.

6) Either way. That would be for the user to decide how they want to use it. The harmoniser only harmonises.

7) No, the harmoniser makes games more exciting by first balancing the game more, and then providing generals with unique and impossible to foresee circumstances, like in real life. No longer can they always count on their kill battle line having a 3+ hit or always activates first. They have less potential, to cheese because the warscrolls are generated dynamically. In a tournament decisions would need to made on the spot! based on the new unit stats.

At the moment, its just who has best kill list and wins. This is more about, who is best at thinking fast and being good at strategy.

If unit's warscrolls are adjuster based on what their opponent brings then you will just have players gaming the new system instead of the current one. You also drastically reduce army composition strategy.

A secondary affect would be that all matches would take longer. Fitting in a 2k tournament match in 2.5 hours can be hard enough even when all players have their rules memorized. If you are constantly changing the rules for each unit the games will take longer because players will be referencing their rules more often.

 

[samheim]

At that point you might as well just go and play with whatever nonsense you've made up yourself as you're no longer playing AoS.

Great! so your starting to see the potential of the system!

 

[samheim]

Which is impossible & pointless, as points are entirely dependent on context.

Context can be factored in very easily. You can take into account all the buffs, the table size everything.

 

[samheim]

If unit's warscrolls are adjuster based on what their opponent brings then you will just have players gaming the new system instead of the current one. You also drastically reduce army composition strategy.

A secondary affect would be that all matches would take longer. Fitting in a 2k tournament match in 2.5 hours can be hard enough even when all players have their rules memorized. If you are constantly changing the rules for each unit the games will take longer because players will be referencing their rules more often.

Yes, a sensible and well thought out point.

Nothing in life comes without costs. There is obviously going to be downsides to anything you do.

However, in real life, army composition does not exist anyway. Its also true to say, no two armies are ever truly balanced as well.

The thing is, AoS might seem very complex to many people. But it software terms it's like zero, zilch, nada nothing.

Hearts of Iron 4 would make AoS's rules, look like tic tac toe.

You can model every piece of the AoS game in software. You can model the table, the dice, the spells, the distance etc etc.

You guys think an old fashioned board game, can't be accurately modelled when they are making virtual reality now?

This is super simple software to make. It's not GTA5. Its a few components and some stats. You have to do a little bit of thinking to how to crunch the numbers. That's it.

When making the software as well, you only build a small part of it at a time. First you try to build it to work with 2 armies, not the whole 22. and then through trial and error you adapt it so that it works better.

There is no magic involved. The magic comes when you have 22 battle tomes with random rules and no way to accurately system test those rules .

In software its very easy to test your system.

 

[ILKAIN]

Great! so your starting to see the potential of the system!

so your system is basically to not play AOS.... what the point then, just go make your own game. also you are literally reinventing the wheel. you want a guy with what is currently 2000 points of fyreslayers to be able to play a guy with what is currently 500 points of another army to be able to play fairly? that is LITERALLY WHY THEY HAVE POINTS! the fyreslayer player simply comes down to the 500 points limit.additionally the stats of the army would be different every time they played. same guy brings 2k fyreslayers and the second guy now brings 1k of his army. effectly per model the fyreslayers would be 2x as strong as they were the first game... so now you have players having to relearn their armies EVERY SINGLE GAME. I know we as Seraphon players have complained about the lack of an update in 4 years.... but there is a such thin as too many updates. no one in the world would buy models and play a game where those models had different stats every time. what you are proposing is effectively the same as when I play with the old school little green army men with my little brother. sometimes the grenadier throws nukes, sometimes he throws rocks, its all about whatever keeps the 4 year old happy....

 

[ILKAIN]

Yes, a sensible and well thought out point.

Nothing in life comes without costs. There is obviously going to be downsides to anything you do.

However, in real life, army composition does not exist anyway. Its also true to say, no two armies are ever truly balanced as well.

The thing is, AoS might seem very complex to many people. But it software terms it's like zero, zilch, nada nothing.

Hearts of Iron 4 would make AoS's rules, look like tic tac toe.

You can model every piece of the AoS game in software. You can model the table, the dice, the spells, the distance etc etc.

You guys think an old fashioned board game, can't be accurately modelled when they are making virtual reality now?

This is super simple software to make. It's not GTA5. Its a few components and some stats. You have to do a little bit of thinking to how to crunch the numbers. That's it.

When making the software as well, you only build a small part of it at a time. First you try to build it to work with 2 armies, not the whole 22. and then through trial and error you adapt it so that it works better.

There is no magic involved. The magic comes when you have 22 battle tomes with random rules and no way to accurately system test those rules .

In software its very easy to test your system.

Hi soldier here.... army composition one-effing-hundred percent is a real thing.

 

[samheim]

Hi soldier here.... army composition one-effing-hundred percent is a real thing.

No, for USA maybe, but I don't think for Iraq army they have the luxury of taking equal counter measures.

 

[samheim]

so your system is basically to not play AOS.... what the point then, just go make your own game. also you are literally reinventing the wheel. you want a guy with what is currently 2000 points of fyreslayers to be able to play a guy with what is currently 500 points of another army to be able to play fairly? that is LITERALLY WHY THEY HAVE POINTS! the fyreslayer player simply comes down to the 500 points limit.additionally the stats of the army would be different every time they played. same guy brings 2k fyreslayers and the second guy now brings 1k of his army. effectly per model the fyreslayers would be 2x as strong as they were the first game... so now you have players having to relearn their armies EVERY SINGLE GAME. I know we as Seraphon players have complained about the lack of an update in 4 years.... but there is a such thin as too many updates. no one in the world would buy models and play a game where those models had different stats every time. what you are proposing is effectively the same as when I play with the old school little green army men with my little brother. sometimes the grenadier throws nukes, sometimes he throws rocks, its all about whatever keeps the 4 year old happy....

The whole point is to design a system that can take into account over powered buffs. In software its very easy to, system test. The army harmoniser is just a system test for AoS, when you think about it.

One possible use is to play heavily mismatched armies against each other. But the harmonisation will work as a gradient. So the closer the 2 armies are to each other to start with, the less harmonisation will be applied.

If you like seeing at tournaments a few high performing lists, then obviously this will have no appeal. But if you want to play with models you like and have them not be useless, this idea may appeal more.

 

[GreenyRepublic]

Samheim, for the benefit of myself and any other software developers here, please go ahead and write this system you are all-so convinced is 'very simple' to make, and host it on some repository somewhere so we can all see for ourselves just how this thing would work. You seem very keen to remind us of how trivial this thing would be, so surely this isn't much to ask.

 

[LizardWizard]

Samheim, for the benefit of myself and any other software developers here, please go ahead and write this system you are all-so convinced is 'very simple' to make, and host it on some repository somewhere so we can all see for ourselves just how this thing would work. You seem very keen to remind us of how trivial this thing would be, so surely this isn't much to ask.

Would love to see it run.

Bring this from theoretical to actual. It is much easier to trouble shoot a real thing than an imagined one. As a moderately well preforming tournament player I would be happy to lend my input into a finished project.

 

[samheim]

Would love to see it run.

Bring this from theoretical to actual. It is much easier to trouble shoot a real thing than an imagined one. As a moderately well preforming tournament player I would be happy to lend my input into a finished project.

Obviously time is an issue. I will try to do something basic. It will not work at first, it will be crap to start with. But it might be able to evolve into something useful.

Even if that is completely different to what the idea is now. The whole point is, I don't know exactly how to do this. No one in the world does, unless they have already done it. I have a vague idea about solid points values, randomly generated stats and establishing hard points values of buffs.

But I could be completely wrong. We can only really start with the end goal. Which is to match 2 opposing forces.

 

[LizardWizard]

Obviously time is an issue. I will try to do something basic. It will not work at first, it will be crap to start with. But it might be able to evolve into something useful.

Even if that is completely different to what the idea is now. The whole point is, I don't know exactly how to do this. No one in the world does, unless they have already done it. I have a vague idea about solid points values, randomly generated stats and establishing hard points values of buffs.

But I could be completely wrong. We can only really start with the end goal. Which is to match 2 opposing forces.

Well, if it works then it works. If it doesn't you will have an answer.

 

[Aginor]

Sure, go ahead.
There are at least two other people on these forums who have worked on simulators for AoS and have tackled some of those problems - with different techniques and success. It is not as trivial as you make it seem.

Go and make it. Take your time.

 

[samheim]

Sure, go ahead.
There are at least two other people on these forums who have worked on simulators for AoS and have tackled some of those problems - with different techniques and success. It is not as trivial as you make it seem.

Go and make it. Take your time.

Yes, sure. My idea is to first start on a very small scale.

I am going to try to balance the units of the Seraphon start collecting, I own and the Khorne Blood bound. About 5 or 6 unit types each.

My method is to first work out the points values for each statistic on the warscroll. WS, BS and so forth.

I will detach the abilities and put them into a list, they will each be given a points value.

The harmonisation will work like so.

First of all it will find the difference in points values between the 2 forces. Then it will split this difference in half.

It will then go shopping from a set lists randomly (add ability or + 1 to something) until it has made up the difference in points.

Conversely it will subtract things from the over valued list until it match's.

I will use my own points system and not the AoS points system, as I want the points system to be extremely hard set.

 

[bOdziO Wolf]

Sounds like something a couple of good heuristic algorithms should be able to solve. I recommend to study Evolutionary Algorithms family and try to apply one of the approaches. Translation of an actual possible game outcome probability of success to a starting list aka game simulation is probably the most tricky part to implement and it will be probably the 90% of work to be done. Other 10% is adjusting and all the fun, like renting a 128 CPU machine on AWS

 

[GreenyRepublic]

The whole point is, I don't know exactly how to do this. No one in the world does.

...You have described this whole task as 'trivial' more times than I can count?

 

[ILKAIN]

No, for USA maybe, but I don't think for Iraq army they have the luxury of taking equal counter measures.

counter measures are not "army composition" you are talking two very different things.

 

[Canas]

Context can be factored in very easily. You can take into account all the buffs, the table size everything.

yeah no... context is a super difficult thing to get right...

You can model every piece of the AoS game in software. You can model the table, the dice, the spells, the distance etc etc.

This is super simple software to make. It's not GTA5. Its a few components and some stats. You have to do a little bit of thinking to how to crunch the numbers. That's it.

Have you ever made any actual complex software? Or modelled something halfway complex? Or made a (video)game, even a simple one? Cuz you seem to have absolutly no idea what makes a game, or software, complex.

But yeah, as others have said, if it is so trivial, go ahead and make a poc and stick it on a public github or something. The result should at least be interesting to look at.