(resolved) Automation for units and buildings

General Discussion
22

And golf is still hockey (refined to the extreme), just without the team, time pressure, etc…
If you think MOBA is anything like RTS, are you sure you’ve actually played one? Just because the original one used an RTS engine, and they look cosmetically similar? MOBAs are brainless.

Attention management is important, I agree - just not to the extent you do. You still have to manage it. You’re assuming that the hole in time management caused by less macro won’t be filled by something else. If you’re needing to do something, distraction will work, whether it’s macro or managing multi-creeping or whatever.

Sure, macro disruption can be more devastating, eg if the player ends up forgetting to build - but having remembering/forgetting to do a menial task as a big part of your game is less fun than disrupting more mentally complex tasks (army management, planning, scouting).

I don’t agree. Upkeep is a gameplay feature. Queues are QoL - they could be implemented entirely client-side if you wanted (especially unpaid, but either way).

There’s kinda two separate purposes for queues:

  1. Long-duration: Put a lot of units in a queue so you don’t have to manually order them later.
  2. Tiny-duration: Add a unit slightly before it’s ready to train, so you get perfect timing and don’t lose a second or so hovering over the train key waiting for it to become available.

(1) is rarely good play. (2) is nearly always good play.

It’s a statistical side effect. It’s not prevention - it’s about ranking. How people with better adaptive-skill rank compared to better mechanical-skill.

Basketball considers height and ball-skills. If you somehow equalized everyone’s height (eg, via VR), you’d find that, on average, ball skills would increase (and height would decrease) for those succeeding - eg becoming pro.

There’s clearly nothing preventing good ball-skills from shining in basketball, but removing height as a factor would increase ball-skills further.

Here’s a more mathematical way of explaining it.
Say you have a population, and each individial has some amount of skill in A and B.
If you rank each individual on A×0.5 + B×0.5, the average A of the top say 10% will be lower than if you ranked purely by A (¹) - because some people with better A will end up below those with better B. The more weight you give B, the lower the average A will be in the top 10% of players. You could show this on a spreadsheet with random data pretty easily.

(¹): Technically it could be equal if A & B are perfectly rank-correlated for the top 10%, but chances are tiny.

I agree, if that’s how you want to use them. But nobody’s forcing you to.
See the top section of my reply here.
xhttps://us.forums.blizzard.com/en/warcraft3/t/minor-suggestion-automation-for-units-and-buildings/3043/15?u=xaxazak-1483

If you spend eons perfecting your talon strat and far less on dryads, your relative performance with talons will be higher than dryads. If a situation arises where these are the best two options, that difference in skill, coupled with habit (or reduced odds in considering dryads), will tilt you towards talons - even if the situation technically makes talons slightly worse.

This isn’t just theoretical - When I watch pros play, I’m always seeing this. They so often go for their comfort units and comfort heroes regardless of the situation.

In many cases, they’re not actually making the wrong decision - going for comfort picks increases their odds because they’re more practiced at using them (and everything related, such as macro timings to train them).

And that makes the game more boring - especially to watch.

But are they also controlling their double medivac harass on their opponent’s two expansions while scouting and expanding themselves? If they’re capable of managing X things at once, and one of them is macro mechanics, then removing macro mechanics gives them time for something else.

Mostly they’re unnecessary. But there are occasions where they’re beneficial (e.g. mass planes). So if you only use them when it’s beneficial (as good players will) there’s precisely zero downside.

Unless you want to argue there’s never a situation in which they’re useful. But then, is 12 some magic number? Why not 11? 7?

We both agree that using queues in this way makes you play worse (at least it should if the countering is hard enough). But is there anything wrong with that?

Firstly, the queues already exist, so the problem does too. Secondly, ways for people to play badly aren’t bad for the game.

If something’s bad practice, pros won’t do it. So if people claim they want something for QoL, and it’s bad practice so it doesn’t affect pro play, what’s the downside? QoL-ers are happy, pros are unaffected. If it leads some people to bad habits, I have no problem - it’s a competitive environment.

1 Like
23

@xaxazak. The truth is you don’t have the mind strength to multitask and adapt in Warcraft 3 games.

Why should blizzard do changes so you can work less and do less to keep up with more skill players ?

The thing is - multitasking is the biggest skill in Warcraft 3 and if you can’t do it - maybe play computer on normal or custom games with map triggers.

If you really wanted to improve - you would practice and go beyond your limit.

Some people don’t want to go beyond their limit but instead want blizzard to lower the game to their low ended micro skills.

I want you too know that if you intend to take this game seriously and want to be the best - you will have to learn micro macro and multitasking as well as body blocking and surrounding.

These mechanics are what separate a normal person from a pro.

The best thing to do is leave the game as it is and have people master the game.

1 Like
24

making things automatic, will only make the game casual, what happens with that? sure it may bring more casuals into the game, but as everyone knows, casuals move from game to game, in a couple of months they will move onto another game. and we will be left behind with this dumb automatic sh1t.
same happened to DBZF and the ridiculous auto-combo stuff, first couple of months it was the hot sh1t, and now there’s a tiny fraction of it’s initial players online.

if the game is too hard for you, just pick something else.

2 Likes
25

It boils down to a big yes, there is something wrong with having unpaid queues vs paid queues. It is a perceptively minor change that has a major impact on the game.

Unlimited unit selection could be considered similar QoL for Micro. The game would change if it were added. It would change how people approach functions in the game by diminishing something else.

You may be fine with diminishing unit production timings because you are not inclined to macro in the first place, which sets up a bias towards other perceived ‘meaningful tasks’ like scouting or double harass. The fallacy comes from dismissing unit production management as a menial task.

To make a reverse argument - what if we were to suggest scouting (as an example of a menial task that does not require full attention) be automated instead of having to click units to path around? What if we make base drop tactics/BM harass easier through more complex AI (auto windwalk when attacked or auto load into zeppelin when attacked). This allows low micro-skilled players to harass at the same level of pros through strategy rather than requiring mechanical skill to pull it off, and they can reserve their intellectual skill for macro. Would you accept this as a gaming option? I would argue against it.

Fundamentally changing the game for the sake of QoL is generally a bad idea.

1 Like
26

Did you read the whole conversation. I posted quite a few examples of multitasking that currently isn’t being done by pros. Scouting is often neglected, meaning pros still often don’t spot expansions for long times. Multi-pronged attacks are also rare (although that’s partly due to the square law of battles). Pros aren’t not doing this stuff cos they’re being lazy, it’s just that a significant part of their APM and attention is being spent on menial stuff.

I want multitasking. There are tons of tasks to be done - far more than are being done. Multitasking won’t somehow magically go away if you remove one task.

I never claimed I want an easier game - I actually claim the reverse - I want it harder. But I want the difficulty to be intellectual - making the right decisions should be the most important skill, expertly navigating a deliberately sub-optimal UI is not a skill I want to be part of the game.

Right now it’s pretty easy for mediocre strategists to do well by simply practicing mechanical execution of basic cookie-cutter strategies.

I want it harder. I know it’s counter-intuitive, but I believe that making the UI easier will make the game harder.

For me, the prime aim of QoL enhancements is to increase strategic difficulty.

1 Like
27

My arguments are nothing to do with me personally. If you need to know, micro is my weakest skill (esp saving and sniping injured units). But I’m mostly interested in game design issues rather than my own play.

Some games have this - basically, auto-wander - I am strongly opposed, because the game is making the decision of where to go for you (in this case, randomly).

IMHO zeppelin loading is badly designed, it needs delays (say 1s to load, one at a time, unit is disabled while loading). Perhaps even have attacks cancel loading. Auto-loading would suck, but instant manual loading also sucks. This is in a way similar to those sucky staves of preservation etc - click-to-save-a-unit mechanisms reduce strategic depth.

Autocast windwalk is probably a bad idea simply because it’s easily abused, regardless of other considerations. However, I do favor some other similar things, such as partial refund when under construction structures are destroyed (effectively, auto-cancel) - although this is mainly about latency, everyone wants to cancel at the last second so there’s no decision to be made, just a race between flying arrows and network packets.

I have the feeling that for most autocast spells, if they weren’t originally autocast, requests to make them autocast would be opposed here - but also vice-versa. Necromancers auto-raising skeletons - isn’t that just the computer playing your necros for you? Some buffs/debuffs are autocast, some aren’t (frenzy). If they had been, would anyone be asking for their removal?

3 Likes
28

From a game design approach, there are many side effects in regards to how Warcraft 3 is intended to be played. The intended way to play is a product of collective Warcraft 3 player expectations on how everyone should ideally play. This means if I am unable to select more than 12 units at a time, then everyone is unable to select more than 12 units at a time, and we know everyone has to adjust their micro to that standard.

Paid queues work because they have very direct and visual costs. You can only queue what you can buy. You only have as much gold as you have spent That is a part of design.

Unpaid queue can be deceptive. You are ‘purchasing’ units in a queue but only pay when they start production. Your actual resources are not reflective of the cost of units you have placed in queue. This leads to a disconnect between perception of unit production and amount of available resources.

If your resources are low and you have high production (lots of barracks), you might not be able to save up for buildings or items. This system also prioritizes quickly-built low-cost units over units over costly units. In a situation where income only trickles in, you would inadvertantly be left with more cheap-cost casters produced while your big hitter melee units sit in a queue waiting to bank above 250g.

This has been communicated as intentional design, and a reason towards why queues are paid in advance. You can use queues as much as you want, you just have to fit it within your resource management. Unpaid queue has significant side effects to resource perception and management without adding a substantial benefit towards macro (beyond paid queues).

The perception of how much gold you actually have vs the amount you spent is complicated with an unpaid queue. It becomes difficult to predict unless you adapt your entire playstyle around that type of play. And frankly, this game has been around long enough where I see this feature having more unintended consequences as a ‘QoL’ addition. If WC3 were still in beta and a different game in the making, then sure we could test this out. However I think this game is set as is, and shouldn’t be complicated further with potential shifts in the meta (of resource management, in this case).

2 Likes
29

Decisions and strategy are also will you spend your attention in macro or micro, when you will do that how you will manage your time to do that. Strategy isn’t just attack from the hills and ambush the enemy like Helm’s Deep xD

When someone talks about more strategy less mechanics, that’s exactly what for me made the Blizz RTSs particularly war3 onward stand out. I don’t want to see you just did ‘This unit that counter that unit’ the end, strategy solved, Win. You have to do a series of mechanics decisions to achieve it. It does add more depth.

Not true that players don’t scout expands fast, it is not that hard to Shift queue some scout.

30

I’m not sure what you mean by this. Everyone plays by the same rules, that’s pretty clear, and we all adapt to those rules. But if the rules change, those things are still true. There are better and worse (or at least, more and less popular) rules, and there’s no need to keep bad or unpopular things - especially if you want more players (and more revenue).

I imagined unpaid as optional. You hold a key down while queueing and the units show up with a $ or something above their queue icon.

That’s true, if you use it in the long-term way - assuming the rules work that way. But if you only use it in the tiny-term way it’s unlikely you’ll have more than one thing queued.

You could use different rules though: Make it so the first unpaid item queued gets priority, or add some configurability.

Say you have 2 hotkeys A & B.

  • A builds in order (except when demands don’t overlap, eg 1st needs more wood but 2nd uses no wood and there’s enough gold for both).
  • B begins whenever resources are available as long as resources aren’t needed for A.

I’m not sure whether that’s already to complex though.

Again, this only affects long-term unpaid use. And, you could always show a separate value below, total unpaid costs.

1 Like
31

If the intended use is for limited short-term production… what is exactly wrong with paid queue? There is no advantage for having it unpaid that makes and sense as an option. It’s a clear advantage over anyone who has a paid queue, albeit one with resource perception consequences.

It is not the same rules the option is functionally different from an existing paid queue. That is like suggesting unlimited selection as an option.

When you say option, do you mean like a custom game mode that allows unpaid queues? Or on an individual basis? Because I still don’t grasp the concept you are suggesting here as I’ve pointed out many conflicts with how resource management actually works in WC3. The cost of buildings and the high dependancy on items can be heavily impacted by misinterpreting available resources.

1 Like
32

Short term unpaid is useful when you’re waiting for the last few resources. Costs 100, you’ve got 90 and going to get +10 in a second or two. You save a few 100ms and a second or so of “hover” time, depending on latency and reflexes. Small savings, but not insignificant (and they often compound) - every bit helps.

Paid queue is for when you have the resources now and don’t want production delayed by accidentally using those resources elsewhere and/or don’t want a misleading UI value. (It’s quite similar to ordered (~FIFO) unpaid, except for the UI values).

Both paid and unpaid are available in melee, switched by holding a (customizable) modifier key (probably shift by default).

So unless you use the modifier key while adding to a queue, everything works like it does now.

1 Like
33

Short term unpaid is useful when you’re waiting for the last few resources. Costs 100, you’ve got 90 and going to get +10 in a second or two. You save a few 100ms and a second or so of “hover” time, depending on latency and reflexes. Small savings, but not insignificant (and they often compound) - every bit helps.

But there are no effective differences to paid queue other than small savings. Small savings should be the reward of good macro timing.

Paid queue is for when you have the resources now and don’t want production delayed by accidentally using those resources elsewhere and/or don’t want a misleading UI value. (It’s quite similar to ordered (~FIFO) unpaid, except for the UI values).

The paid aspect is not just so that you don’t ‘accidentally use resources intended for other use’. It’s an advance payment. That is the cost of using a queue. The point of WC3’s macro design is based on meaningful macro choices. If you want to save 100ms, then you queue using up resources. If you don’t have those resources then you need to work on your resource management. That is how the game is played.

Both paid and unpaid are available in melee, switched by holding a (customizable) modifier key (probably shift by default).
So unless you use the modifier key while adding to a queue, everything works like it does now.

That simply doesn’t work. I believe you are seeing this feature as how you would use it; as a short term limited queue. I’m looking at this feature with Murphy’s law in mind, and I don’t see the benefits outweighing the flaws considering paid queue does everything you’re asking for given you have the resources to do it. For the specific situation of wanting to queue and not having the resources to do it, I think it’s better to work on that aspect rather than depending on a function in the game.

With the situations that I’ve presented such as unit production priority, resource cost not being reflected and the overall limited use that queues have altogether, it’s better to keep the paid queue the way it is and let players use it only as a stepping stone towards not having to use it at all. Ideally, the goal everyone should strive towards is perfect macro timing where you are using resources to build units manually. That is the ideal way to play the game, not by using queues to save 100ms or shave a bit of time from waiting for resources.

I see unpaid queues as a gateway towards macro laziness. This is a huge incentive towards all-in or early rush strategies for players with low macro attention. Queue up a barracks full of units, rally point to hero and rush. Normally, a player of lower skill may be caught up in the heat of battle and forget to reinforce when necessary, forcing them to switch strategies or learn to macro better. Unpaid queue would have an unintended effect of incentivizing bad strategies like all-in full queue rush because a player has little incentive to learn how to macro properly. Because this is an official feature in the game, they believe this is the intended way the game should be played since there is no perceived cost/punishment for fully queuing a barracks and forgoing macro while in battle. This is not something ‘abusable’ with a paid queue, since all costs are upfront.

1 Like
34

Blizz has been pushing automation since 2002. It is most likely we will have automation in buildings in Warcraft 4 or Starcraft 3 for the reason the thread starter wanted it.

As a.i. develops well eventually have self stragtic units. You’ll select a large group and a-key them to move/attack, but as they encounter the enemy they’ll auto fall into a battle formation like a flock of birds. A siege tank will auto path its way to the high ground to siege. Blizz will probably make an auto cast a.i. button.

35

if you want automatic bullsh1t, then just play AI vs AI and think you are one of them lol

36

This is probably close to the essence of our disagreement - I have no interest in valuing that skill (given that it devalues other skills - see the basketball explanation). There’s also the latency issue.

Again, I have no interest in valuing that. As you’ve made the decision already, IMHO there’s no reason to force the execution of that decision to be any later.

Sounds menial, mindless, dated, etc… Again, new players won’t value that much, spectators don’t value it, and old players who want a primarily strategy game don’t value it. Many of the ones that do value it do so because they’re invested in it (pros especially) - they’ve developed strong macro reflexes.

Somewhat related - SC2 pros are getting beaten by AI already - primarily because while AI still sucks at deep strategy (especially positional), the current crop of pros don’t have exceptional strategic skills to beat them with either, and the AI does the menial stuff flawlessly.

A RTS with well designed units and structures should give players ways to beat mindless attacks like this. Rapid but expensive base defense, coupled with decently hard mid/low tech counters to the rush unit, should fend it off - although the rusher should probably still have an advantage if they adapt. But, before then, rushes should be scouted. XP also helps here if the enemy is spamming units, to a small degree.

It’s be nice if Blizzard chose to balance their game specifically to encourage clever play, but I don’t think they will. Even TFT, aside from the new units, made it worse.

They could improve things by:

  • Increasing the hardness of counters (possibly partly through tweaks to the damage table, eg weakening siege vs unarmored).
  • Reducing the flexibility of units (eg. siege tank AA is such a bad idea - maybe make it cheap but negligible, or just replace it with a different skill more suited for their main role).
  • Making teleport far harder (TPs cost 600+, and possibly replace all teleport-type staves).
  • Slightly exaggerate the speed difference between fast and slow units.

Continuous all in full queue rush should not gain them many ranks. IMHO, let sucky players suck in whatever sucky manner they choose.

Also, maybe adding a small cancel tax for units (say 10%) might help to punish this playstyle. (Make it free for say 5s perhaps, so as to not punish accidental hotkey typos).

IMHO, the line to draw is all about decisions. These suggestions mostly don’t ask the AI to make decisions for the player (and when they do I usually oppose it). Instead, they just let the player make decisions earlier if they choose to.