|
On December 18 2010 05:09 gwombat wrote:
these two ideas are NOT mutually exclusive. if you are keeping your money low AND you are not queueing units, that's good macro. the whole point of keeping your money low is that you don't want money, you want units. and queued units eat money without giving you units right away. it's as if your money has vanished into a vortex. it'll be back later.
although i would say that for someone in bronze league who kept forgetting to make stuff, queueing units might actually be better than stubbornly saying "queueing is bad! no queueing!"
I agree that it can be an effective crutch. But players should realize that it IS a crutch for their broken macro. Just like a real crutch is an effective tool for a broken leg. Just make sure you aren't trying to convince yourself that your leg ain't broken or that thing you're leaning on isn't a crutch.
Also.. Queuing units is the SAME as not spending your money. You're just taking the dollars sitting in the top right corner waiting to be spent, and having them wait to be spent in a barracks, since you haven't really spent your money on a marine until he pops out ready to shoot stuff. Until he pops out, you can cancel the marine at any time and spend it on something else, just like leaving the 50 mins in the top right bank.
|
I don't think this is the way to teach it to anyone. It's important to do both of them at the same time, that's how you get good at macro. If you have the amount of money that you need to queue units to keep it low, then make more expansions and production buildings.
|
2500+ diamond t player here. i think that the two concepts are pretty similar, if not identical, because thinking about it, if you are queing units, then that means only part of the money you have spent is being used and in reality you have more money in your bank then it seems
keeping money low is basically saying that you should use as much of the money you have. and as i said earlier, queing units does not mean that the money is being used, hence we find that both the ideas are very similar to eachother.
|
Blazinghand
United States25551 Posts
I think we can agree that queued units don't do anything since they are in a queue. I disagree however about the unusefulness of production facilities. If you expo, or want to cut workers and make more units, or something like that, a spare rax is going to be more useful than a queue of rines.
|
On December 18 2010 00:47 Eeryck wrote:
The problem with the statement that queuing is ok, is that it smacks in the face the concept of optimized play.
Optimized play is pretty much (from the very start of the game):
1. Are units producing from every structure, if yes goto 2; else make units
2. Do I have enough supply, if yes goto 3; else make supply
3. Do I still have extra money, if yes build a building/get an upgrade; else goto 1
By switching to a mentality that at some point queuing is ok moves away from this concept of optimized play. I breaks the simple structure that guarantees the most units out at any given time for your given ability.
It creates a new mentality, that it is "optimal" to invest in things that are delayed in order to allow for a human error element. This breaks the mindset of optimal play. In this new mindset you will start to micro longer and longer and need to que more and more. For the time it works for you it will set up this huge bad habit, since it will work in the short term. Then when you advance behind it and hit a wall because your opponents will have stuck with optimized play what will you do?
Where optimized play already allows for the human error element. It tells you if you have extra minerals after units are producing and supply is met then you build a building. Done! No new thinking involved. The simple pattern remains simple and you will learn to do it more quickly over time.
I think it would be more precise to say that optimized play is when you're constantly producing units from every structure, and your next depot/pylon/overlord finishes just before you need it, and you have no extra money because you are constantly getting new production facilities and upgrades at the right time.
It's likely that no human is actually capable of perfectly optimized play, because there's always some little thing you could do better. The real question is, when you screw up, when you fall off the thread of perfectly optimized play, what is the smallest, least impactful mistake you could make?
Now, obviously it's best to have exactly one unit building in each production facility at all times, but if you're going to make mistakes (and you are - everyone does) which do you prefer: some amount of queuing (never more than one queued unit per structure), or the lost production time that inevitably results either from waiting for the previous unit to finish building before building a new one, or from being distracted at the precise moment that your round of production finishes? I tend to think that some small amount of queuing is better than losing any production time. Maybe I'm wrong about that, but I really don't think so.
|
Having actually read the OP, I think queuing is preferable to letting your money float, but it's like asking me whether I want to blast a shotgun into my left foot or my right foot.
I'd say, a better option than queuing is if your money is getting high, just send one scv to expand, and several more to build more production fac. Even if your macro is imperfect (as it often is), having a better INFRASTRUCTURE that you can use even occasionally is preferrable to having a crappy one that you use constantly. Obviously, it only depends how much better your infrastructure is, and how well you can macro to begin with, but the higher infrastructure is definitely a better investment than queuing.
|
i like this post ^^ cause im some way those buildings actully have a posiblity to do something. where as teh qued units can do nothing. and if you do run low on money because you have to many production failitys. your atleast keeping your money low.. and its not like you built an obscean amount of buildings. just a little more then enough.
|
On December 19 2010 03:56 Matrijs wrote:Show nested quote +On December 18 2010 00:47 Eeryck wrote:
The problem with the statement that queuing is ok, is that it smacks in the face the concept of optimized play.
Optimized play is pretty much (from the very start of the game):
1. Are units producing from every structure, if yes goto 2; else make units
2. Do I have enough supply, if yes goto 3; else make supply
3. Do I still have extra money, if yes build a building/get an upgrade; else goto 1
By switching to a mentality that at some point queuing is ok moves away from this concept of optimized play. I breaks the simple structure that guarantees the most units out at any given time for your given ability.
It creates a new mentality, that it is "optimal" to invest in things that are delayed in order to allow for a human error element. This breaks the mindset of optimal play. In this new mindset you will start to micro longer and longer and need to que more and more. For the time it works for you it will set up this huge bad habit, since it will work in the short term. Then when you advance behind it and hit a wall because your opponents will have stuck with optimized play what will you do?
Where optimized play already allows for the human error element. It tells you if you have extra minerals after units are producing and supply is met then you build a building. Done! No new thinking involved. The simple pattern remains simple and you will learn to do it more quickly over time. I think it would be more precise to say that optimized play is when you're constantly producing units from every structure, and your next depot/pylon/overlord finishes just before you need it, and you have no extra money because you are constantly getting new production facilities and upgrades at the right time. It's likely that no human is actually capable of perfectly optimized play, because there's always some little thing you could do better. The real question is, when you screw up, when you fall off the thread of perfectly optimized play, what is the smallest, least impactful mistake you could make? Now, obviously it's best to have exactly one unit building in each production facility at all times, but if you're going to make mistakes (and you are - everyone does) which do you prefer: some amount of queuing (never more than one queued unit per structure), or the lost production time that inevitably results either from waiting for the previous unit to finish building before building a new one, or from being distracted at the precise moment that your round of production finishes? I tend to think that some small amount of queuing is better than losing any production time. Maybe I'm wrong about that, but I really don't think so.
Not true. With every production cycle you spend putting units in queues where they aren't actively producing, the more money and production time you lose. Essentially, after a certain number of production cycles, you will have wasted a certain amount of minerals/time that would have been equivalent to setting up an additional structure and producing out of it, even with gaps.
If you routinely have 1 unit queued for at least half a production cycle (e.g. You create a marine and halfway as it's done you create another) it'd be better to stop queuing and put up another barracks or a reactor.
If you think about it this way, let's say your income is 125 minerals per minute and your production facilities are capable of making one marine and one scv every minute. Ignoring the fact that your income rises (unless you're fully saturated) with each production cycle, we have a perfect situation. Every 4 cycles, we must use 100 minerals to build a supply depot, and every cycle, we gain 25 minerals.
Now let's imagine instead that our production is the same but our income is 175 per minute. This suggests that we're able to queue one marine every cycle, then one SCV, etc. After 3 cycles we've suddenly queued 3 units; 150 minerals. What if we just made another barracks after 3 cycles? We'd increase production, and we'd have 2 marines and 1 SCV every cycle. Something tricky happens and we have to cut marine production from one barracks for a few seconds after each cycle to get minerals for supply, but this is still better than queuing.
Thus, after 6 production cycles, if we queue we've made 6 marines and 6 SCVs, with no money. If we use that money for a barracks we're looking at, let's say, 7 marines and 6 SCVs, with money. (pretend it takes 2 cycles to complete the barracks)
EDIT: in this situation, let's say we want to maintain constant SCV production. This means we need to use the barracks at around 67% efficiency to maintain enough minerals to not get supply blocked every 3 cycles. Thus, this means that, for the first production cycle after the barracks goes up we'll have 2 marines at 1:20 and 1 SCV, with 1 SCV 33% percent complete. We should also have roughly 33 minerals saved; this will go towards a supply depot after the third cycle.
|
On December 19 2010 03:56 Matrijs wrote:
The real question is, when you screw up, when you fall off the thread of perfectly optimized play, what is the smallest, least impactful mistake you could make?
The least impactful response when your macro is imperfect is to make more production structures. Sometimes expanding with the extra money is the better choice, and I actually kind of prefer this option because it can allow you to turn your macro mistake into a big advantage, but even if he busts it and you die, I think it can make that game more educational for you.
|
On December 19 2010 03:56 Matrijs wrote:Show nested quote +On December 18 2010 00:47 Eeryck wrote:
The problem with the statement that queuing is ok, is that it smacks in the face the concept of optimized play.
Optimized play is pretty much (from the very start of the game):
1. Are units producing from every structure, if yes goto 2; else make units
2. Do I have enough supply, if yes goto 3; else make supply
3. Do I still have extra money, if yes build a building/get an upgrade; else goto 1
By switching to a mentality that at some point queuing is ok moves away from this concept of optimized play. I breaks the simple structure that guarantees the most units out at any given time for your given ability.
It creates a new mentality, that it is "optimal" to invest in things that are delayed in order to allow for a human error element. This breaks the mindset of optimal play. In this new mindset you will start to micro longer and longer and need to que more and more. For the time it works for you it will set up this huge bad habit, since it will work in the short term. Then when you advance behind it and hit a wall because your opponents will have stuck with optimized play what will you do?
Where optimized play already allows for the human error element. It tells you if you have extra minerals after units are producing and supply is met then you build a building. Done! No new thinking involved. The simple pattern remains simple and you will learn to do it more quickly over time. I think it would be more precise to say that optimized play is when you're constantly producing units from every structure, and your next depot/pylon/overlord finishes just before you need it, and you have no extra money because you are constantly getting new production facilities and upgrades at the right time. It's likely that no human is actually capable of perfectly optimized play, because there's always some little thing you could do better. The real question is, when you screw up, when you fall off the thread of perfectly optimized play, what is the smallest, least impactful mistake you could make? Now, obviously it's best to have exactly one unit building in each production facility at all times, but if you're going to make mistakes (and you are - everyone does) which do you prefer: some amount of queuing (never more than one queued unit per structure), or the lost production time that inevitably results either from waiting for the previous unit to finish building before building a new one, or from being distracted at the precise moment that your round of production finishes? I tend to think that some small amount of queuing is better than losing any production time. Maybe I'm wrong about that, but I really don't think so.
First of all, producing units from every structure is not a necessary part of optimized play. There are times where you want to be spending your money on more production or on an expansion, because it will benefit you later with a higher unit count. Best in BW did this really well (his macro was the "Best" >_>)
Also, I don't know why you really have to choose. You should just work on getting your timing down to where you queue for no more than half a second on one end and have idle time by no more than half a second on the other, at least for the units you typically make (I don't think that a Protoss needs to be overly concerned on how long it takes a Carrier to finish, really). If you can't at least do this with workers, you're probably doing it wrong.
|
This thread doesn't warrant 15 pages of responses. What do you do when your money is really high to bring it down? You could queue up 4 rounds of marines, and then oops 30 seconds later you're back to your money being really high, but you have nowhere else to queue units. Or, you could build enough production facilities to match your income.
|
|
|
|
|
|
EPT
