Topic: Soldier upgrade system needs improvement

Lobsterninja wrote: The soldier upgrade system needs improvement. My strategy currently is to set minimum occupancy on all military buildings in order to get units to go and train, then after most of the units are out training I set the military building occupancy back to max. Then I manually kick units out of training facilities when training appears to be complete. Obviously this is not consistent with the flavour of this game; one would expect a bit more automation.

Generally, it may need improvement, but Widelands is about economy, so we would need to do those improvements along the line of improving economy, but not enhancing warfare

Suggested modifications: 1. Any unit that can't benefit from additional training should leave the training facility.

That's good at a first thought, but bad at a deep insight. That would at the end mean too many new recruits created when the supply with higher level equipment ceases. Maybe there's a way to automate all those decisions, but that would at the end mean improving warfare very strongly. See my first reply about that.

1. In military buildings, have two occupancy settings (minimum occupancy and maximum occupancy). When the military building contains more than the minimum, and a training facility is requesting trainees, and a unit in the building needs that training, the unit most needing training is sent to the training facility.

That sounds like pure economy behaviour, but it may collide with your first statement, making endless runs of lowest trained but untrainable soldiers between the training camp and the military building.

1. If a military building is short of it's "minimum occupancy", it has precedence over training facilities.

That is a pure economic thing and could be implemented, I guess. Problem would then only be: If there is need for a Soldier, should it be created in the Headquarters or should it be drawn from a military building which is over the Minimum occupancy limit? One could do the default settings predefined per tribe to make it more interesting (Atlanteans have almost full occupancy as default, Barbarians less than half, Empire about half).

1. Even if a military building is full, it should request units of higher ability than are currently present. If such a unit arrives, the weakest unit should leave to make room.

This would improve warfare too much, I think. This should take manual attention.

CMake is evil.

Top

Quote

ixprefect wrote: Most of these things seems to be rather a matter of taste. Except:

Lobsterninja wrote: 1. Any unit that can't benefit from additional training should leave the training facility.

Isn't this already the case? Was I dreaming?

I've seen this only after the soldier has been fully upgraded. If the soldier can't be upgraded because of missing weapons or armory, he stays in the training camp. The suggestion was (if I understand it correctly) to evict soldiers at lower upgrade stages if the weapons or armory for the higher stages are not available. If this has been changed lately, I've not seen this working yet, but I've also not supervised soldier training in my games in the last four weeks.

I'm not so sure about the other things. I'd rather err on the side of keeping things simple for now. Setting another min/max configuration for each military building seems to be over-complicating things.

D'accord. Simple mechanics keep the player busy

CMake is evil.

Top

Quote

QCS wrote:

Lobsterninja wrote:

The soldier upgrade system needs improvement. My strategy currently is to set minimum occupancy on all military buildings in order to get units to go and train, then after most of the units are out training I set the military building occupancy back to max. Then I manually kick units out of training facilities when training appears to be complete. Obviously this is not consistent with the flavour of this game; one would expect a bit more automation.

Generally, it may need improvement, but Widelands is about economy, so we would need to do those improvements along the line of improving economy, but not enhancing warfare

I agree that Widelands is about economy. I don't think "enhancing" warfare contradicts this, as long as "enhancing" means simply "improving" rather than "making more significant". I am not making any proposals that make the military aspect more significant or more rich. I am proposing making it less frustrating, easier, and less subject to deadlocks. Just because this is an economic game does not mean that any kind of improvement in the military aspect is bad.

Suggested modifications:

1. Any unit that can't benefit from additional training should leave the training facility.

That's good at a first thought, but bad at a deep insight. That would at the end mean too many new recruits created when the supply with higher level equipment ceases. Maybe there's a way to automate all those decisions, but that would at the end mean improving warfare very strongly. See my first reply about that.

I'm fine with requiring manual effort to get military right, but as a universal rule, every significant aspect of the game needs to have either 1. good automation or 2. good control. If any part of the game has neither control nor good automation, it will lead to frustration. Yes, economy is more important than military, but that is beside the point. It's always frustrating when AI makes bad decisions unless you have the ability to override those bad decisions. On the other hand, bad AI is perfectly fine if it can be overridden by the player. Optimally the AI of every aspect of the game (including military) should be sufficient to prevent deadlocks in most cases.

I would argue that this principle is just as important for military as it is for economy, and improvement does not have anything to do with the relative emphasis of military vs. economy. Emphasis and quality are orthogonal concepts.

In military buildings, have two occupancy settings (minimum occupancy and maximum occupancy). When the military building contains more than the minimum, and a training facility is requesting trainees, and a unit in the building needs that training, the unit most needing training is sent to the training facility.

These items were not meant to be taken separately, but rather as a system.

That sounds like pure economy behaviour, but it may collide with your first statement, making endless runs of lowest trained but untrainable soldiers between the training camp and the military building.

That's only if you interpret my post based on the assumption that it is stupid If you read more carefully at the complete proposed system, you will see that it is not as you describe. The purpose of setting two levels is to allow the player to have control over the dilemma of choosing between keeping the building defended and getting training done. By having two numbers like this, it is possible to design a system that allows a players entire military to eventually become fully trained without micromanaging. Yes, this would improve the military aspect of the game, but this is actually not a bad thing, since the improvement would in fact reduce the emphasis on military by reducing the need for micromanagement. In this case quality goes up while emphasis goes down. An improvement in the military in this case makes it easier for the user to focus on economy.

If a military building is short of it's "minimum occupancy", it has precedence over training facilities.

That is a pure economic thing and could be implemented, I guess. Problem would then only be: If there is need for a Soldier, should it be created in the Headquarters or should it be drawn from a military building which is over the Minimum occupancy limit? One could do the default settings predefined per tribe to make it more interesting (Atlanteans have almost full occupancy as default, Barbarians less than half, Empire about half).

It would be drawn from a military building which is over the Minimum occupancy limit, of course. The whole point is to get all of the soldiers trained without micromanaging. Thank you for noticing that it is a pure economic thing. In fact my entire proposal is essentially economic. The idea is to move units around in such a way as to convert less valuable units into more valuable units, so in that sense it is economic.

Even if a military building is full, it should request units of higher ability than are currently present. If such a unit arrives, the weakest unit should leave to make room.

This would improve warfare too much, I think. This should take manual attention.

I don't understand the idea that we should avoid improving warfare. I know it is an economic game, therefore the military aspect of the game should be kept simple (from the users point of view). But aside from that, troop allocation needs improvement because it currently resembles herding cats. I don't think the quality of the game is harmed by improving the military aspect of the game so long as it doesn't make the military more significant or more complex. I'm only proposing making the military aspect easier. In fact, I would argue that improving automation of warfare, while decreasing micromanaging of warfare, serves to increase the emphasis on economy.

Manual attention would be okay, I suppose, except that manual attention is currently inadequate. The player has almost no control over troop placement. All the user can do now is eject troops randomly and hope they go where they are wanted.

One of the things I like about the game is that there is plenty of automation, which alleviates much of the micromanagement found in other games. Ideally this automation would extend somewhat more to troop placement. As a minimum requirement, it should be possible for all units to eventually become trained automatically if the necessary buildings and resources are in place. This means there needs to be some mechanism by which training facilities pull troops from buildings and vice versa based on current skill level.

Top

Quote

ixprefect wrote: Ah, I understand now. Well, actually I don't want automatic eviction just because a helmet happens to be missing. That would be extremely irritating for a player who is in the later stages of the game and really does want to upgrade soldiers fully, but has a currently slow-going resource production. After all, you can always evict soldiers manually, but you can't manually pull soldiers back once they've left the building.

Okay, I agree with that. However, the current system does not result in all your units getting trained fully. It only maximises those units that happen to make it to a training facility. There is no mechanism, automatic or manual, to get all of your units trained. My main point is that we need a system that makes units move around in such a way that everyone gets trained eventually. Again, this would not increase the emphasis on military, but on the contrary would decrease the emphasis on military by making the user not have to worry about it as much. Paradoxically, the more automation of the military, the less "emphasis" on the military.

True, you can evict soldiers manually, but this is not very useful since you can't control where they go, and they are not very smart about choosing a destination. Of course I would certainly not suggest adding the ability to directly control the destination of evicted units since that would seriously go against the whole flavour of the game for obvious reasons. Better would be to handle soldiers in a manner consistent with the rest of the game. Specifically, units should automatically move in such a way that their value increases.

Anyway, I think I'll take a look at the code....

Top

Quote

Lobsterninja, try to write a good blueprint at launchpad widelans home. I'm sure that there will be improved and reviewed by developers easily.

Top

Quote