Poll: Is a game with more units better?

[albator]

I don't think it is always a yes. In many game there is no point to have more unit cause you cannot efficiency control them, in spring you can !

yes you can ! :D

[luckywaldo7]

The background behind this poll is that it was used as an argument against MT.
http://springrts.com/phpbb/viewtopic.ph ... 62#p539858

maybe as note: more units doesn't mean better games, it just gets confusing.

So naturally I want to know what the community thinks.

Apologies for not reading through the whole thread, but IMHO isnt the question of "we should have more processing resource available" always has a "yes" answer? It doesnt matter what it is used for, either more units or smooth games, whatever the case is, if you have MORE headroom, then you are always better of than if you had none of it at all.

The statement only argues against the "more power = more units = games are as laggy". There were a few other valid questions raised against MT at the moment, which are not as straightforward to address as this one. But all in all, MT is the future and giving up on it is shooting yourself in a foot.

And yet, any day you will find that a bullet in the foot is better than a bullet in the head*.

MT would be glorious, but not at the cost of the engine actually working and being maintainable.

*(unless zombies)

[Prominence]

+1

If teammates leave after death or near death, then less unit per player means he, she, or they are at numerical disadvantage. Hence if the stalemate occurs, then it ends up being attrition for numerically disadvantaged player. Thus, it's not fair to have small unit caps like 1000.

[KDR_11k]

More units wouldn't really be worth it if implementing that would mean stripping the simulation down to AI War levels. That game has tens of thousands of units in play at the same time and thus has to trim them down to the barest minimum of complexity. An average attack involves around 2-3 thousand units per side.

+1

If teammates leave after death or near death, then less unit per player means he, she, or they are at numerical disadvantage. Hence if the stalemate occurs, then it ends up being attrition for numerically disadvantaged player. Thus, it's not fair to have small unit caps like 1000.

That's just an argument for implementing the unit limit differently (e.g. adjusting limits when team members leave) instead of a plain limit increase (which would have the same old problem just with bigger numbers).

[zerver]

-1 Cause lagout

you did not read the poll question! your vote has been forwarded to the paper shredder...

[KDR_11k]

I think he means network traffic. Multithreading wouldn't decrease that.

Also AI War gives us another argument: Large unit counts can't be meaningfully controlled, you pretty much run them on auto-pilot most of the time.

[PicassoCT]

Bien au contraire, monsieur KDR!

The truth is, we just have not developed a powerfull enough gui to put all those little strategys en large into place. Though spring is the spearhead there, a real large scale strategy gui, would be like paiting on the terrain, and connecting the colours with functions. Example?





This (actually more well drawn) represents a ambushzone. Units can ge assigned to guard the Reserve Zones, until either the ambush zone holds enough units or a trigger is walked over- then the Reservezone units lead a autoattack along the functionpaths into there defensezone- and empty the the ambush zone- afterwards retreating.

What is needed for it? A brush on the map, and some quite impressive logic to connect the various zones. What is the reward? Intricated behaviour and tactics, while having little work to set the whole thing up.. once (Do not make it a behavior saved per UNIT, im begging you)

So yes, there is work and room for future advances..

[zwzsg]

Picasso example is a tad too detailed, but in a broader way, it would interesting to investigate RTS controls where, instead of selecting units then telling them what to do, you create actions to be done, then your units perform them. Though I currently fails to name any, I know there are games in brother genres that already have such scheme.

[dimm]

Sounds like widget. In fact there was one for construction.

I also said "wut is the point besides it looking cool" so edit not delete. Lags the main reason i don't play btw.

[knorke]

Though I currently fails to name any, I know there are games in brother genres that already have such scheme.

globulation is somewhat like this, but it is not not really large scale.
http://globulation2.org/
has pro music though!

[PicassoCT]

But can you override it in case of emergency? Nobody wants the dungeon keeper feeling of powerlessness over the chaos called battleplan your minions improvise..


Actually, no to Sub_AI_Generals and TyconZones.. what i ment is a sort of programmable Plansystem.. AIs have nothing to do with it- maybe except for knorkes statemaschine AI. Maybe that is grasping it best. I want statemaschines. With states and exact conditions.

[varikonniemi]

No other engine can compete with that. That the very reason MT was awesome.

come off it. There will be another multi-threaded implementation

Is there something specific you refer to?

[Anarchid]

Example?

Latest wesnoth has some of that, btw, called "whiteboard":

[KDR_11k]

Picasso example is a tad too detailed, but in a broader way, it would interesting to investigate RTS controls where, instead of selecting units then telling them what to do, you create actions to be done, then your units perform them. Though I currently fails to name any, I know there are games in brother genres that already have such scheme.

I'm not sure that'd be better than what we currently have, it's perhaps useful for much slower genres or awkward systems like phones or tablets but I don't see how it'll help with dealing with larger forces.

Telling your units to go or attack somewhere as a blob isn't hard with what we have, getting the various unit types to do different things gets tougher. And I don't mean just "arty goes here, tanks go here", I mean quickly repositioning units as the need arises.

[Beherith]

Like Majesty?

[Anarchid]

Dwarf fortress is an rts.

[PicassoCT]

Picasso example is a tad too detailed, but in a broader way, it would interesting to investigate RTS controls where, instead of selecting units then telling them what to do, you create actions to be done, then your units perform them. Though I currently fails to name any, I know there are games in brother genres that already have such scheme.

I'm not sure that'd be better than what we currently have, it's perhaps useful for much slower genres or awkward systems like phones or tablets but I don't see how it'll help with dealing with larger forces.

Telling your units to go or attack somewhere as a blob isn't hard with what we have, getting the various unit types to do different things gets tougher. And I don't mean just "arty goes here, tanks go here", I mean quickly repositioning units as the need arises.

The problem is that telling your units to do so- is taking up precious ressources (attention) which could be elsewhere.

Imagine it like aset of pre-setup able automised unitpumps.

The beauty of it that they wont attack like a "blob" - instead you get a well coordinated ambush from multiple reservespots. And you can even have multiple ambushzones drawing from one or two reserve zones, thus resulting in units which basically auto distribute to whoever or howmany are triggered first. The beauty of it is basically that you can program such quite easily. State A, State B, connection+condition and voila. I find it fascinating that this used for traditional Waiting/Repair/Ambush or Ferrysystems is considered optimal, but once it reaches the actuall battleplanning its a evil controll taking system. Well its not, but thats not the point.
But then what is?

The time to program stuff, that exists in abandunce at gamestart?
The fear that a opponent could make auto-counter attacks while you micromelee with him? Maybee its the idea of the "brilliant general" that is still in my head, and that wont let go of dumb grunt-behaviour (autonomy even in strong limitations? NEVERETTE!) down there on the battlefield.

[dimm]

I find it fascinating that this used for traditional Waiting/Repair/Ambush or Ferrysystems is considered optimal, but once it reaches the actuall battleplanning its a evil controll taking system. Well its not, but thats not the point.
But then what is?

Traditionally its just getting rid of boring time waste while your version may infringe on actual fun.

[Anarchid]

Automated, pre-programmed taskforces are a very real thing also in a game called Space Empires IV, especially if you play simultaneous mode. You cannot directly influence tactics in that mode (eliminating some hilariously overpowered moves), but you can spend a real lot of time pre-planning your moves, testing them against enemy units and tactics in a simulator, etc.

All invisible, and unconventional.
Final strategy, all so clear to me

Then you send your fleets into battle for results that you cannot foretell and influence. This letting go of control, throwing of dice after you think you cannot improve your plan further, is actually part of the fun: your preparations are finally put to the ultimate test.

Now is time to go to War.

What would that mean applied to an RTS? I think that would mean that you get a lot of time in your game to pre-program those things somewhere. Such an RTS cannot be the hectic microfest like sc or, say, zk. If anything, it would be...

An Ender's Game :)

[knorke]

But can you override it in case of emergency?

Of course not. The very idea of Globulation is to have only vague control. If units wander off or do not behave perfectly efficient that is part of the game. Since that is expected it is not as frustrating as one might think, it is part of the game.

I find it fascinating that this used for traditional Waiting/Repair/Ambush or Ferrysystems is considered optimal, but once it reaches the actuall battleplanning its a evil controll taking system. Well its not, but thats not the point.
But then what is?

Traditionally its just getting rid of boring time waste while your version may infringe on actual fun.

I think there is a point where games have so many helpers/widgets that players end up controlling those helpers more than actually playing the game.
Too complex UI is just as bad and distracting as a too crude one.