Platinum

[Dale]

- when to expand
Either expand during a battle or fast expand at the very beginning.
Also, after you've fended off an attack is also a good time.

- when to tech
Usually after you've fended off an attack,
or fast tech at the beginning.

- when to attack?
If you've scouted, and see an opening... attack.
If you've scouted, and see he has a bigger army,
either defend/counter or harass as best as you can.


- how do you deal with opponents that turtle (i've been losing to terran a lot)
As Zerg: Baneling bust, Nydus, Drop
As Protoss: Prism Warp-in, Prism Drop, Mothership Mass Recall.
As Terran: Thor, Nuke, Seige Tank, Medivac Drop

- when to upgrade
When the pressure he's putting on you is relatively low.
If he's putting pressure on you, you benefit more from making units.

basically... timing and macro
Practice! You can benefit a lot from using hotkeys, key binds,
and shift-movements. Use the UI to its fullest.

MOST IMPORTANTLY: Scout!!! That's the
only way to know if it's proper to make your move.

I hope this helps.

[Hammy]

No offense Dale but I think some of that advice is terrible...

When an opponent turtles, the rule of thumb is to overwhelm him with macro... which means you should macro up and expand (and tech in some situations). You're saying that the best way to beat a turtling opponent is to bust him, but that's not true: if he's turtling, that means he has a defensive edge, which means that you'll be investing a lot to kill him, and taking a risk. When I watch silver-player streams, the most frequent mistake they make is attacking when they don't have enough of an advantage, and then they just lose to the counter attack.
You need to watch out not to spread yourself too thin of course, since he could be planning a timing attack, and you have the following in mind: what could he be preparing, and what do I need to do to counter that? If you think of that all the time, you'll always be one step ahead. You might be spending a bit of ressources on something that he isn't doing, but that's the cost of playing solid.
Turtle busts are something pretty specific that shouldn't be the rule of thumb. Besides, when you all-in baneling bust him, you aren't really improving anyways.

-when to upgrade: NO! Upgrades need to be incorporated in your gameplan!! If you want to make a bunch of MnM you NEED a +1 upgrade, especially if you're marine heavy! The moment the +1 kicks in, you've got a decisive advantage. You're game plan could be to sync up +1w, with stim and with a nice mass of MnM. If he can defend that attack (say he has some photons and FFs his ramp), don't sacrifice your army, but expand and power up to MnMnM. With a strong army like that, you might not be able to break his front, but you'll have map domination, so you can macro up then.
I'm not saying you should always do this, but this is how you should be thinking.

-when to attack: Well yeah if you scout him and see that he's weak, you might as well try your luck, but actually your attacks should also be in your gameplan. When you scout him you get some intel on what he might be going for. When you understand the match-up, you know when he'll be weak with that build, and can plan accordingly.

Here's an example regarding "game plan":
Say, if you're in ZvP and he 3 gates, you can either go blingling, or focus on roaches, since you know that his robo/stargate will be greatly delayed. If you're going for roaches, plan a +1range to obliterate him.
If he's going 1-gate tech, you can expect either a fast stargate or a robo w/ quick imo/collossi. At that point it's crucial to get intel, since a void-ray rush could be countered with fast hydra den, while you can dominate a fast robo with muta. If I can't scout it, I try to be safe and start with hydra then transition if necessecary, all while expanding since I know 1-gate is no threat early game and minerals are not the limiting factor (gas is).

Do NOT take these examples too seriously, I'm just trying to illustrate what I mean by "game-plan".

Again, no offense Dale. I agree that scouting is crucial, but you're advice might actually encourage gamers to play so "randomly"... I'm not going to pretend I'm any good (I'm not at all, believe me), but this is my personal opinion on how to improve at this game^^

[Dale]

No offense Dale but I think some of that advice is terrible...
None taken.

When an opponent turtles, the rule of thumb is to overwhelm him with macro... which means you should macro up and expand (and tech in some situations). Valid approach, but your expansions will initially come at the expense of your army. If he scouts, even a little, he'll attack your expansions before you see any profit. You're saying that the best way to beat a turtling opponent is to bust him, but that's not true: if he's turtling, that means he has a defensive edge, which means that you'll be investing a lot to kill him, and taking a risk. There are multiple ways to deal with the problem. I perfer the aggressive approach. If the guy is turtling, then either break his defenses or bypass them. Waiting and trying to macro expansions gives him time to mass (which is exactly what he wants). Turtling, than allowing you to out resource him is not the intention of your opponent. He will attack before he lets you see the resource profit of your expansions (unless he's a noob). When I watch silver-player streams, the most frequent mistake they make is attacking when they don't have enough of an advantage, and then they just lose to the counter attack.You need to watch out not to spread yourself too thin of course, since he could be planning a timing attack, and you have the following in mind: what could he be preparing, and what do I need to do to counter that? If you think of that all the time, you'll always be one step ahead. You might be spending a bit of ressources on something that he isn't doing, but that's the cost of playing solid. That's why I mentioned the scouting, defend/counter, and harassing. I see no disagreement between us on this point.

Turtle busts are something pretty specific that shouldn't be the rule of thumb. Besides, when you all-in baneling bust him, you aren't really improving anyways. I never mentioned the "all-in" part. If your approaching him from his front door, breaking his wall & getting your army past his wall and into the base before he can mass has to be a priority.

-when to upgrade: NO! Upgrades need to be incorporated in your gameplan!! If you want to make a bunch of MnM you NEED a +1 upgrade, especially if you're marine heavy! The moment the +1 kicks in, you've got a decisive advantage. You're game plan could be to sync up +1w, with stim and with a nice mass of MnM. If he can defend that attack (say he has some photons and FFs his ramp), don't sacrifice your army, but expand and power up to MnMnM. With a strong army like that, you might not be able to break his front, but you'll have map domination, so you can macro up then.
I'm not saying you should always do this, but this is how you should be thinking. My response to this was "When the pressure is low." Ask yourself when the pressure is usually low? At the start of the game, when you're going through your initial build order. The implication is that upgrades should be incorporated into your build order. I see no disagreement between us.

-when to attack: Well yeah if you scout him and see that he's weak, you might as well try your luck, but actually your attacks should also be in your gameplan. When you scout him you get some intel on what he might be going for. When you understand the match-up, you know when he'll be weak with that build, and can plan accordingly. You can't plan everything. You're metagame will only take you so far. Scouting and adapting is key. To be honest, I see no disagreement between us on this issue too.

Here's an example regarding "game plan":
Say, if you're in ZvP and he 3 gates, you can either go blingling, or focus on roaches, since you know that his robo/stargate will be greatly delayed. If you're going for roaches, plan a +1range to obliterate him.
If he's going 1-gate tech, you can expect either a fast stargate or a robo w/ quick imo/collossi. At that point it's crucial to get intel, since a void-ray rush could be countered with fast hydra den, while you can dominate a fast robo with muta. If I can't scout it, I try to be safe and start with hydra then transition if necessecary, all while expanding since I know 1-gate is no threat early game and minerals are not the limiting factor (gas is).

Do NOT take these examples too seriously, I'm just trying to illustrate what I mean by "game-plan".

Again, no offense Dale. I agree that scouting is crucial, but you're advice might actually encourage gamers to play so "randomly"... I'm not going to pretend I'm any good (I'm not at all, believe me), but this is my personal opinion on how to improve at this game^^
Again, no offense taken.

The thing is, everyone plays in a way that's best for him/her.
That's why I purposely left my responses vague, which
could be misunderstood as 'play randomly.'
That was not my intention.

And seeing as we agree on everything except
how to deal with a turtle, I think it's all gravey anyway.

EDIT: Just to add to the discussion,

Lets assume that we have an expansion with 6 mineral nodes,
and 2 vespean geysers. The mineral cost to fully saturate this
expansion (without static defenses) using Zerg (the cheapest race) is:

A = Cost of Drones for Building
3 buildings x 1 drone per building = 3 drones
3 drones x 50 minerals per drone = 150 minerals

B = Cost of Hatchery.
1 hatchery x 300 minerals per hatchery = 300 minerals

C = Cost of Extractors
2 extractors x 25 minerals per extractor = 50 minerals

D = Cost of Drones for Mining Minerals
6 nodes x 3 drones per node = 18 drones
18 drones x 50 minerals per drone = 900 minerals

E = Cost of Drones for Mining Gas
2 Extractors x 3 drones per extractor = 6 drones
6 drones x 50 minerals per drone = 300 minerals

F = Cost of Overlord to supply Mining Drones
1 supply per drone x (18 + 6) drones = 24 supply
24 supply x (1 overlord per 10 supply) = 2.4 overlords ~ 3 overlords (rounded up)
3 overlord x 100 minerals per overlord = 300 minerals


TOTAL COST OF EXPANSION:
A + B + C + D + E + F =
150 + 300 + 50 + 900 + 300 + 300 =
2,000 minerals per fully saturated expansion.

Now, according to this SCL article, a fully saturated expansion
with 6 minerals nodes will yeild 13.67 minerals per second.
Assuming that we have all portions of the expansion spontaneously
blink into existence (taking 0 secs to build), it would still take
2 minute and 26 seconds before the expo repays it's cost.

Seeing as how most games are 10 - 15 minutes long, it would take
16% - 24% of the total length of the game to just PAY BACK the
investment you put in. Keep in mind, that's if the base takes ZERO
seconds to build. If we factor in build time, that number goes up.

That's more than enough time for even a noobish turtle to scout
out your expansion and roll through with his probably superior
army since he put those 2,000 minerals into building his army.

(Although, you probably have more gas for upgrade/tech though).




.

[moosh]

And seeing as we agree on everything except
how to deal with a turtle, I think it's all gravey anyway.

EDIT: Just to add to the discussion,

Lets assume that we have an expansion with 6 mineral nodes,
and 2 vespean geysers. The mineral cost to fully saturate this
expansion (without static defenses) using Zerg (the cheapest race) is:

A = Cost of Drones for Building
3 buildings x 1 drone per building = 3 drones
3 drones x 50 minerals per drone = 150 minerals

B = Cost of Hatchery.
1 hatchery x 300 minerals per hatchery = 300 minerals

C = Cost of Extractors
2 extractors x 25 minerals per extractor = 50 minerals

D = Cost of Drones for Mining Minerals
6 nodes x 3 drones per node = 18 drones
18 drones x 50 minerals per drone = 900 minerals

E = Cost of Drones for Mining Gas
2 Extractors x 3 drones per extractor = 6 drones
6 drones x 50 minerals per drone = 300 minerals

F = Cost of Overlord to supply Mining Drones
1 supply per drone x (18 + 6) drones = 24 supply
24 supply x (1 overlord per 10 supply) = 2.4 overlords ~ 3 overlords (rounded up)
3 overlord x 100 minerals per overlord = 300 minerals


TOTAL COST OF EXPANSION:
A + B + C + D + E + F =
150 + 300 + 50 + 900 + 300 + 300 =
2,000 minerals per fully saturated expansion.

Now, according to this SCL article, a fully saturated expansion
with 6 minerals nodes will yeild 13.67 minerals per second.
Assuming that we have all portions of the expansion spontaneously
blink into existence (taking 0 secs to build), it would still take
2 minute and 26 seconds before the expo repays it's cost.

Seeing as how most games are 10 - 15 minutes long, it would take
16% - 24% of the total length of the game to just PAY BACK the
investment you put in. Keep in mind, that's if the base takes ZERO
seconds to build. If we factor in build time, that number goes up.

That's more than enough time for even a noobish turtle to scout
out your expansion and roll through with his probably superior
army since he put those 2,000 minerals into building his army.

(Although, you probably have more gas for upgrade/tech though).




.

Wow, that's interesting stuff. Is there like a list of these kind of helpful articles somewhere?

My zerg game has gotten so much better thanks to some of the strategies i've been reading around everywhere. The build order i read somewhere for Spawning Pool at 13 made all the difference. Rest of the game is a matter of scouting and then teching appropriately. I don't really upgrade at this point. But i'll start incorporating that in.

[Hammy]

It's fun to have those numbers down but they are completely biased:
-Who ever said you have to fully saturate your expansion? (2 drones per node is optimal btw, especially as Zerg. With other races you might want to maynard, but even then you have to consider that factor).
-What about the production output you receive from the larvae?
-Gas? May I remind you that gas is probably the main reason you expand to begin with? Expanding to yellow minerals is fantastic, but just an added bonus, but in most games gas is the limiting factor. There are some exceptions of mineral based strategies of course... It would be interesting to add the gas mining values to your "model".
-Those 2k minerals (actually more like 1600minerals or 1750minerals) aren't spent instantly either...

Calculations like these are extremely misleading. I hope that moosh was refering to the macro article and not this model. The whole basis of a strategy game is put aside here... Although, with adjustements, I think this is interesting to note while optimizing builds at a very high level (which sadly, we aren't doing here), this is theorycraft pushed to the extreme...

When do you expand: when you can afford to! If your survival is always pending on a thin thread, you can't invest any ressources to anything other than army, or you'll get run over before you can capitalize on it. That means that even though expansions are expensive, you only invest what you can affort at that exact instant. The difficulty is finding out when you can invest in expansions or tech, instead of pure muscle. In the end, that investement will pay back slowly but surely and allows you to take an advantage within minutes.
Sometimes it's better to pump extra units to finish off your opponent, sure. That's absolutely true, but you can't forget about defenders advantage.


Also, you say games are 10-15 minutes long as though each of those minutes were strategically identical. Actually "advantage/disadvantage" shifts so quickly in a game (for example, a shift occurs the moment your first collossus pops out, it doesn't even need to deal damage) that you can easily balance out a new expansion.
Why do players often expand as they are moving out to attack?
-Because the units you could be spawning with those 300/400mins wouldn't be involved in the battle anyways because they would be crossing the map
-Because that's a moment where you have an advantage (otherwise, in most cases, you wouldn't be moving out)
-The battle, even if not decisive, will lead to damage on both sides, which means both armies will be smaller in size:
=> This reinforces defenders advantage in case of a counter-attack: with smaller armies, your neo-spawned units will contribute more to a battle if he were to counter attack your new expansion, while his units will have to run through the map.

Sure, there are different playstyles, but in the end solid playstyles are what pay the most, while risky playstyles might win you games here and there, but you can lose to "inferior" players with them. Also, if this game is anything like BW, as builds are more and more optimized, solid players will likely have an edge.

TL: DR: In the end I ramble a bit about unrelated stuff, but my point is that, while these values are interesting, they don't really benefit the average player and might actually be very wrongly interpreted and hinder a newbie's progression.
That being said, and I think this sums up my opinion rather nicely: I one-base play rather frequently for timing-attacks to finish my opponent early on (great for fast laddering), but I need to switch my gameplan very quickly if I feel like my attack might fail, hence losing me the game.

[Dale]

It's fun to have those numbers down but they are completely biased: I disagree. All numbers are verified. The logic is sound.
-Who ever said you have to fully saturate your expansion? (2 drones per node is optimal btw, especially as Zerg. With other races you might want to maynard, but even then you have to consider that factor). Agreed. 2 drones per node as the optimal saturation was also the conclusion of the SCL article. I expect the 2 drone/node model to pay itself off quicker, but the 3 drone/node model will yield more minerals per second once it's been paid off (the expansion will exhaust quicker as well).
-What about the production output you receive from the larvae? Agreed. This will provide a tactical advantage for Zerg. Thanks for pointing that out.
-Gas? May I remind you that gas is probably the main reason you expand to begin with? Expanding to yellow minerals is fantastic, but just an added bonus, but in most games gas is the limiting factor. There are some exceptions of mineral based strategies of course... It would be interesting to add the gas mining values to your "model". I mentioned gas at the last line of the post.
-Those 2k minerals (actually more like 1600minerals or 1750minerals) aren't spent instantly either... The less drones you make, the more time it takes to pay off the expansion. That time will not be wasted by your opponent.
Calculations like these are extremely misleading. I hope that moosh was refering to the macro article and not this model. The whole basis of a strategy game is put aside here... Although, with adjustements, I think this is interesting to note while optimizing builds at a very high level (which sadly, we aren't doing here), this is theorycraft pushed to the extreme... I disagree. Expanding is an incredibly important tactical decision. Nothing but benefit can be gained by crunching the numbers. I did that as fairly as I could using conservative numbers. It sums up the time/mineral/unit cost of expanding pretty nicely.

When do you expand: when you can afford to! If your survival is always pending on a thin thread, you can't invest any ressources to anything other than army, or you'll get run over before you can capitalize on it. That means that even though expansions are expensive, you only invest what you can affort at that exact instant. The difficulty is finding out when you can invest in expansions or tech, instead of pure muscle. In the end, that intvestement will pay back slowly but surely and allows you to take an advantage within minutes. Agreed. The above number crunching supports the statement in this paragraph.

Sometimes it's better to pump extra units to finish off your opponent, sure. That's absolutely true, but you can't forget about defenders advantage.
Agreed. No one is forgetting the defenders advantage, the disagreement we have, I think, is how best to overcome that advantage.

Also, you say games are 10-15 minutes long as though each of those minutes were strategically identical. I didn't mean to imply that. Either I misspoke or perhaps you misunderstood. Actually "advantage/disadvantage" shifts so quickly in a game (for example, a shift occurs the moment your first collossus pops out, it doesn't even need to deal damage) that you can easily balance out a new expansion.
Why do players often expand as they are moving out to attack?
-Because the units you could be spawning with those 300/400mins wouldn't be involved in the battle anyways because they would be crossing the map
-Because that's a moment where you have an advantage (otherwise, in most cases, you wouldn't be moving out) Agreed. That's some good analysis/advice right there.
-The battle, even if not decisive, will lead to damage on both sides, which means both armies will be smaller in size:
=> This reinforces defenders advantage in case of a counter-attack: with smaller armies, your neo-spawned units will contribute more to a battle if he were to counter attack your new expansion, while his units will have to run through the map. Agreed.

Sure, there are different playstyles, but in the end solid playstyles are what pay the most, while risky playstyles might win you games here and there, but you can lose to "inferior" players with them. Also, if this game is anything like BW, as builds are more and more optimized, solid players will likely have an edge. Agreed.

TL: DR: In the end I ramble a bit about unrelated stuff, but my point is that, while these values are interesting, they don't really benefit the average player and might actually be very wrongly interpreted and hinder a newbie's progression.
That being said, and I think this sums up my opinion rather nicely: I one-base play rather frequently for timing-attacks to finish my opponent early on (great for fast laddering), but I need to switch my gameplan very quickly if I feel like my attack might fail, hence losing me the game. Agreed.

This discussion has been very productive.

EDIT:

Stuff my analysis doesn't take into account, that has not yet been mentioned.
1. Minerals spent on the expansion are done over time.
2. Minerals earned by expansion are done over time.
Even before/after the expansion has been built, units can be made with incoming money.

[Hammy]

What I mean is that it would have been nice to have the actual values of gas collection versus mineral collection, not just a note saying that you would have more.
In fact, I think a key concept behind expanding is how much minerals you're investing to collect gas instead. Sure the expansion pays for itself in a certain time, but how much gas are you gaining out of it during that time, and after? The strength of a zerg is usually jauged by how many gases he's mining from, not how many bases he has.

Also, I wasn't implying that your numbers were "wrong", but that your parameters were unrealistic. Hence my points on worker saturation, exact gas payback and other parameters.

It's been an interesting discussion indeed!

PS: I try not to extrapolate too much on what you say, but when I see:

Seeing as how most games are 10 - 15 minutes long, it would take
16% - 24% of the total length of the game to just PAY BACK the
investment you put in.

I can't help but notice that you put aside many parameters for the sake of the point you're making. Such an analysis is arguable and the numbers I quoted here aren't strategically relevant.
=> That's why I said you seem be considering each minute as strategically identical.
Maybe that's not what you meant, but that's what this analysis implies.