Spawn Larva could easily be fixed if it allowed you to stack it. I personally think it is a good mechanic and I play zerg.
Dear god man. Imagine mass queens!
ah yes, I see the confusion. I simply meant that you could queue them up. So like if I had forgot for awhile and my queen has 50 energy, boom I could poop two on there and then they would spawn consecutively
and I think we go over this automation argument every thread. There's a reason that SC2 has risen as an e-sport, and it's not because of automation
Yeah, I understand and I still love SC2 just the way it is.
My point is that if we already had these features it would be different. (money only deducts when a unit starts, queens can automatically inject, buildings can be told to repeat-produce) Nobody would say "hey, you know what would be great? If we could only produce one unit at a time efficiently, so we had reselect our base every 17 seconds and click again." Or "oh man you know what zerg needs? It'd totally be more fun if they had to select all their queens and manually click on their hatcheries ever 40 seconds."
joe k, this is more an sc2 thing than a game of the year thing so I'm bringing it here, but how exactly do you think sc2's "keyboard gymnastics" could be cut down?
For me at least, I'd add in Supreme Commander 2's "repeat" production option and queue system. As soon as something is finished, it looks at what you have queue'd next, then takes the money and builds it. If you have repeat, you can set it to constantly produce units or groups of units. In SC2, it'd be great if I could tell my reactor-barracks to constantly produce marines as long as I have money. Or my larvae to constantly produce roaches/hydras/overlords at a 2/2/1 mix.
Queens could have autocast on vomit larvae. Creep tumors could have autocast+direction to auto-spread. Give them a command to go build over *clicks* there, and they'll keep producing tumors until they reach there. Chronoboost could be set to auto-cast on a targeted building every time you had 25 energy.
If you really wanted to cut down on the finger gymnastics, you could also add in a stutter-step move command. You automatically move, stop, move, stop, move towards a direction. Engage at max range would be nice too. RA3's 'coordinate' waypoint system would be cool. You give a bunch of units commands, and hit activate, and they all move at once.
After that, the only thing I'd add in would be a programmable target priority system. So, when your units are attacking or stopped they target based on your chosen threat list, rather than the built in SC2 threat list. I'd have marauders automatically attack non-slowed stalkers, the nearest non-slowed zealot, then slowed stalkers, then everything else. Marines would always attack PDDs observers and banelings first, then whatever had the lowest health after that. Immortals would always attack armored, etc.
This is a good starting point. You could add Carriers are set to auto-cast make interceptors. The casting spells onto unit portraits that was removed because some koreans complained that it was too easy. Not being able to warp in units on the mini-map. No autocast on the vomit. No formations. No more complex orders than move, attack, hold, stop, patrol.
I'm not sure why people consider tedious repetitive tasks to be part of a strategy game, real time or not. Like when Dhal got all pissy when I said that it would be nice to pull down the replay bar menu for your side. check out your production tab. Check out your units tab. This is all easily available info that would make the game more reliant on strategy, and less on memorized rapid keyboard movements. Sure, its real-time, but usually anything resembling "strategy" goes to the wayside while you're off babysitting everything.
It's the reason that I *hated* War3. The keyboard gymnastics with your heroes, and the damn Creeps.
nobody is asking for a python API. Just make the damn thing do things that are repetitive nonsense. (hey, i need 25 sets of 'lings.... time to mash z, couldnt there be a combo that said select all, make all $foo?)
earlier Red Alerts and C&C's (not this latest one, it sucks ass) had more advanced features, such as the sidebar where you could just pick what units you wanted to be created, and it would make them from wherever was best, including making partial units as resources came in.
having hands that would pleasure your mother in less than 15 seconds is a neat skill, but I'm not sure where that became the focus of strategy games.
That would make macro nonexistent. Because hey, as soon as you build a rax, you don't need to pay attention to it anymore. It'll build the right ratio of marines to marauders. It won't queue. It'll only build it if you have the funds. You even suggest automatic target priority... are you serious? You want to eliminate micro too? Turn on the option to have my hurt stalkers blink back then re-engage. No need for stalker blink micro. Turn on the option to pick out only banelings out of a zergling/baneling ball so you can't screen them. Automating stutter-step? Are you high?
i recall hearing that the very hard ai cheats resources in so
edit: i think that the queen shouldnt have larva vomit at all
its badly done, and while automating would fix it in a sense, its avoiding the issue of it just being a bad macro mechanic in comparison to orbital and chronoboost
zerg macro mechanic should be done differently
how is it a bad macro mechanic?
and I think we go over this automation argument every thread. There's a reason that SC2 has risen as an e-sport, and it's not because of automation
the zerg macro mechanic (vomit) is easily the most powerful in the game. It creates exponential growth as soon as you have a queen (you double (plus) your unit production). Then another hatch&queen doubles that again. If you can manage to get to 4 bases, plus queens, you can produce 9x(plus a little) (28 larvae a vomit round vs 3) what you could off of one hatch, at once.
terrans have mules and reactors. protoss has chrono and gates. neither come close, unless you completely overbuild gates for late game replenishments.
i just think it is a high floor, low ceiling type of thing, which i personally dislike
it doesnt really warp the balance at higher levels, so its fine where it is, but i personally would like zerg's macro mechanics to be done in a different way
just remember, if you ever plan to go mutas, make sure you DO something with the map control they give you
take another expansion, drone up, whatever, but the single greatest terrified moment of a Protoss' game is when mutas hit their minerals and they only have ground troops like stalkers to defend. even if you do no damage, especially in the lower leagues, you will so frustrate your opponent that he will completely forget how to play.
This.
I used to have such a hard time just keeping my cool when banshees or mutas showed up in my base. Now I'm not really afraid of either, although, 2port is still a little hard to hold off.
Screw that noise! There is no way a human being can put out units as fast as that! NO FREAKING WAY!
If you macro right you should never fall behind the Very Hard A.I., actually. I do great against the Very Hard until I make a macro slip. I almost always beat the Zerg on Very Hard, 50/50 against my mirror, and I have a hella time beating the very hard terran. For some reason, one little slip-up on my macro vs terran is a way bigger deal than against Z or P.
i recall hearing that the very hard ai cheats resources in so
edit: i think that the queen shouldnt have larva vomit at all
its badly done, and while automating would fix it in a sense, its avoiding the issue of it just being a bad macro mechanic in comparison to orbital and chronoboost
zerg macro mechanic should be done differently
how is it a bad macro mechanic?
and I think we go over this automation argument every thread. There's a reason that SC2 has risen as an e-sport, and it's not because of automation
the zerg macro mechanic (vomit) is easily the most powerful in the game. It creates exponential growth as soon as you have a queen (you double (plus) your unit production). Then another hatch&queen doubles that again. If you can manage to get to 4 bases, plus queens, you can produce 9x(plus a little) (28 larvae a vomit round vs 3) what you could off of one hatch, at once.
terrans have mules and reactors. protoss has chrono and gates. neither come close, unless you completely overbuild gates for late game replenishments.
Huh. The rate your production increases is linear in the number of hatchery-queen pairs that you add on. The power of inject doesn't have to do with the amount of larva it adds. Overall production of a single zerg base compared to the comparable set of buildings for terran/protoss --- counting both workers and army --- is roughly equivalent. The power lies in the fact that all those larva can be dedicated to either drones or army, so a zerg's pure army production or drone production is better than the other two races, but their overall production is equal.
Kambing on
@TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
That would make macro nonexistent. Because hey, as soon as you build a rax, you don't need to pay attention to it anymore. It'll build the right ratio of marines to marauders. It won't queue. It'll only build it if you have the funds. You even suggest automatic target priority... are you serious? You want to eliminate micro too? Turn on the option to have my hurt stalkers blink back then re-engage. No need for stalker blink micro. Turn on the option to pick out only banelings out of a zergling/baneling ball so you can't screen them. Automating stutter-step? Are you high?
Actually, this version of the game already exists in not just one, but two forms!
Desert Strike and Nexus Wars
Personally, I love Desert Strike - I just play it to wind down from playing 1v1s, because I'm not nearly as interested in whether I win or lose and I get to watch all the battles I want. It's also not the shit show that Nexus Wars is...
Effectively you guys are arguing how much Real Time you want in your RTS.
On one side of the continuum you have Starcraft II as is, on the other you have battle chess (where it's entirely turn based but hey! the battles are animated!)
While improving the AI and self-designating priorities would be nice, programming fighting AIs is a different game than a RTS. Without the element of multitasking-focus distribution-unattended things screw up, it just isn't an RTS anymore. That being said, there are a fair number of default decisions that Blizzard make that take an unrealistic amount of micro to counteract (like SCV repair, which is finally being fixed.)
Dropping Loads on
Sceptre: Penny Arcade, where you get starcraft AND marriage advice.
3clipse: The key to any successful marriage is a good mid-game transition.
i recall hearing that the very hard ai cheats resources in so
edit: i think that the queen shouldnt have larva vomit at all
its badly done, and while automating would fix it in a sense, its avoiding the issue of it just being a bad macro mechanic in comparison to orbital and chronoboost
zerg macro mechanic should be done differently
how is it a bad macro mechanic?
and I think we go over this automation argument every thread. There's a reason that SC2 has risen as an e-sport, and it's not because of automation
the zerg macro mechanic (vomit) is easily the most powerful in the game. It creates exponential growth as soon as you have a queen (you double (plus) your unit production). Then another hatch&queen doubles that again. If you can manage to get to 4 bases, plus queens, you can produce 9x(plus a little) (28 larvae a vomit round vs 3) what you could off of one hatch, at once.
terrans have mules and reactors. protoss has chrono and gates. neither come close, unless you completely overbuild gates for late game replenishments.
Huh. The rate your production increases is linear in the number of hatchery-queen pairs that you add on. The power of inject doesn't have to do with the amount of larva it adds. Overall production of a single zerg base compared to the comparable set of buildings for terran/protoss --- counting both workers and army --- is roughly equivalent. The power lies in the fact that all those larva can be dedicated to either drones or army, so a zerg's pure army production or drone production is better than the other two races, but their overall production is equal.
i really don't think its linear... its geometric, and damnit, i've had graph/discrete math conversations with you, so maybe i'm missing something here, but 3->7->14->28 looks geometric to me.
If you macro right you should never fall behind the Very Hard A.I., actually. I do great against the Very Hard until I make a macro slip. I almost always beat the Zerg on Very Hard, 50/50 against my mirror, and I have a hella time beating the very hard terran. For some reason, one little slip-up on my macro vs terran is a way bigger deal than against Z or P.
The zerg AI is bad in general because it doesn't know how to harass with lings or mutas. So it almost always throws a ling/roach/hydra ball at you which all 3 races can stop easily.
The other two races are more proficient at straight-up attacking, so the AI fares considerably better with them.
Kambing on
@TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
0
Options
3cl1ps3I will build a labyrinth to house the cheeseRegistered Userregular
edited December 2010
It doesn't double. Every Hatch+Queen you add gives you 7 larvae every 40 game seconds. The progression is
3 (1hatch0queen)-->7 (1hatch1queen)-->14 (2hatch2queen)-->21 (3hatch3queen)-->28 (4hatch4queen), etc.
Adding additional queens adds no extra larvae; adding additional hatches without queens adds 3 larvae per 40 seconds.
EDIT: The real strength isn't that eventually the production becomes overwhelming, it's that when you're at maxed food and you save up a bunch of larvae you can end up with 60-80 larvae sitting around, so as soon as you lose units you can make new ones for 100+ food. To match that with Warp Gates Protoss need about 40, to match that with Barracks Terrans need about 20 reactor + 20 tech lab. Those are both massively more expensive investments, since larvae only cost energy and a Hatch can have up to 19 saved up.
2. The reason you are getting geometric is because you are assuming a double expansion after two bases. You can theorycraft if this is easier/more effective for Zerg vs. other races if you want, I don't have enough experience with it to judge, but if you consider the hatchery / expansion rate to be linear, your growth rate is linear. If you consider 4 hatcheries on 3 bases, it works in the short term (which may be all you need to win the game) but the resource collection rate will no longer allow geometric progression, although I'm pretty sure quadruple expanding is a rarity anyway =P.
Dropping Loads on
Sceptre: Penny Arcade, where you get starcraft AND marriage advice.
3clipse: The key to any successful marriage is a good mid-game transition.
i recall hearing that the very hard ai cheats resources in so
edit: i think that the queen shouldnt have larva vomit at all
its badly done, and while automating would fix it in a sense, its avoiding the issue of it just being a bad macro mechanic in comparison to orbital and chronoboost
zerg macro mechanic should be done differently
how is it a bad macro mechanic?
and I think we go over this automation argument every thread. There's a reason that SC2 has risen as an e-sport, and it's not because of automation
the zerg macro mechanic (vomit) is easily the most powerful in the game. It creates exponential growth as soon as you have a queen (you double (plus) your unit production). Then another hatch&queen doubles that again. If you can manage to get to 4 bases, plus queens, you can produce 9x(plus a little) (28 larvae a vomit round vs 3) what you could off of one hatch, at once.
terrans have mules and reactors. protoss has chrono and gates. neither come close, unless you completely overbuild gates for late game replenishments.
Huh. The rate your production increases is linear in the number of hatchery-queen pairs that you add on. The power of inject doesn't have to do with the amount of larva it adds. Overall production of a single zerg base compared to the comparable set of buildings for terran/protoss --- counting both workers and army --- is roughly equivalent. The power lies in the fact that all those larva can be dedicated to either drones or army, so a zerg's pure army production or drone production is better than the other two races, but their overall production is equal.
i really don't think its linear... its geometric, and damnit, i've had graph/discrete math conversations with you, so maybe i'm missing something here, but 3->7->14->28 looks geometric to me.
Hrm. I suppose this might be a "growth relatively to what" thing, but:
1 hatch larva production = 1/15 larva per second
1 hatch+queen larva production = 1/15 + 4/40 larva per second
So the amount of larva you produce overall assuming perfect macro is:
1/15n + (1/15 + 4/40)m larva/second
where n is the number of bare hatches you have and m is the number of hatch/queen pairs you have. And the above equation is linear in the number of hatches you are producing from.
Kambing on
@TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
It doesn't double. Every Hatch+Queen you add gives you 7 larvae every 40 game seconds. The progression is
3 (1hatch0queen)-->7 (1hatch1queen)-->14 (2hatch2queen)-->21 (3hatch3queen)-->28 (4hatch4queen), etc.
Adding additional queens adds no extra larvae; adding additional hatches without queens adds 3 larvae per 40 seconds.
I'm in complete agreement, now plot production capacity over time (i almost always see a double expand from 2->4 or a macro hatch and a 3rd base (4 total)), and let me know what the curve looks like.
3cl1ps3I will build a labyrinth to house the cheeseRegistered Userregular
edited December 2010
But that requires math
EDIT: Also, a double hatch or a 3 bases 4 hatch makes it still a linear progression, because you're just adding 2x instead of 1x at that point; the discreet unit is still x, not 2x.
I think people argue that vomit larva is a bad mechanic not because you have to do it but because there's no choice or option involved in the matter
realistically, you pretty much should be MULEing every time too but you have the choice of scouting/detecting or fixing a blunder with supply calldown as well
and clowns can chronoboost whatever the fuck. of course, you need to be chronoboosting just as much as you need to be MULEing and all, but T and P have options and choices
queens just gotta spit on shit every 40 seconds. it's whatever, though
It doesn't double. Every Hatch+Queen you add gives you 7 larvae every 40 game seconds. The progression is
3 (1hatch0queen)-->7 (1hatch1queen)-->14 (2hatch2queen)-->21 (3hatch3queen)-->28 (4hatch4queen), etc.
Adding additional queens adds no extra larvae; adding additional hatches without queens adds 3 larvae per 40 seconds.
I'm in complete agreement, now plot production capacity over time (i almost always see a double expand from 2->4 or a macro hatch and a 3rd base (4 total)), and let me know what the curve looks like.
To be fair, almost any hatch before pool build will move you from 4 larvae a minute to 20 larvae a minute almost instantly.
zerg rush on
0
Options
3cl1ps3I will build a labyrinth to house the cheeseRegistered Userregular
edited December 2010
Whenever people talk about Chronoboost, all I can remember is that like every pro I watch just doesn't use it.
Protoss might be better off if people actually spent their bloody chronoboost.
Hrm. I suppose this might be a "growth relatively to what" thing, but:
1 hatch larva production = 1/15 larva per second
1 hatch+queen larva production = 1/15 + 4/40 larva per second
So the amount of larva you produce overall assuming perfect macro is:
1/15n + (1/15 + 4/40)m larva/second
where n is the number of bare hatches you have and m is the number of hatch/queen pairs you have. And the above equation is linear in the number of hatches you are producing from.
i agree with your math, it looks to describe the larvae/second, but I'm not sure that it describes the point I'm trying to make, which maybe I'm completely off on this - its happened before, but I think that looking at it from game time elapsed as your X axis, and larvae production potential as your Y axis, I still think that you get a exponential curve.
(editted to reflect reality of fast expand)
1 Hatch:
3 larvae/45 s (lets call this L)
2 Hatch
6 larvae/45s (2L)
Add a queen:
10 larvae/45 s (3L+1)
Add another queen
14 larvae/45 s (4L+2)
Expand+queen
21 larvae/45 seconds (7L)
Macro Hatch+queen
28larvae/45 seconds (9L+1)
So, for 1500 minerals (not factoring in pool, or OL costs), you now have a little over 9x of your original production potential. If the time between new hatcheries/queens is equal (and I would argue that 3&4 go down near simultaneously) , you are looking at a geometric increase in larvae production potential over game time.
I think people argue that vomit larva is a bad mechanic not because you have to do it but because there's no choice or option involved in the matter
realistically, you pretty much should be MULEing every time too but you have the choice of scouting/detecting or fixing a blunder with supply calldown as well
and clowns can chronoboost whatever the fuck. of course, you need to be chronoboosting just as much as you need to be MULEing and all, but T and P have options and choices
queens just gotta spit on shit every 40 seconds. it's whatever, though
zergs op
the choice is drone or units, inject or spawn tumor/transfusion
Orikae! |RS| : why is everyone yelling 'enders is dead go'
When I say pop it that means pop it
0
Options
MaratastikJust call me Mara, please!Registered Userregular
edited December 2010
RE: more automation in SC2...
SC2 is not just a strategy game...it is a multitasking game. Who can keep up with doing more things all at once. Believe it or not some people find the level of multitasking present in SC2 appealing (I do for instance). SC2 strikes a nice balance between pure strategy and pure multitasking. This is part of the fun of SC. Sure some people don't like it, but then maybe you should play those other games with more automation rather than complaining about SC. People who feel the lack of automation gets in the way of the strategy are missing out on the point of a lot of SC2's gameplay. Half of the point/fun is in the number of tasks you have to keep up with. I wouldn't have SC2 any other way.
EDIT: Also, a double hatch or a 3 bases 4 hatch makes it still a linear progression, because you're just adding 2x instead of 1x at that point; the discreet unit is still x, not 2x.
joe k, this is more an sc2 thing than a game of the year thing so I'm bringing it here, but how exactly do you think sc2's "keyboard gymnastics" could be cut down?
For me at least, I'd add in Supreme Commander 2's "repeat" production option and queue system. As soon as something is finished, it looks at what you have queue'd next, then takes the money and builds it. If you have repeat, you can set it to constantly produce units or groups of units. In SC2, it'd be great if I could tell my reactor-barracks to constantly produce marines as long as I have money. Or my larvae to constantly produce roaches/hydras/overlords at a 2/2/1 mix.
Queens could have autocast on vomit larvae. Creep tumors could have autocast+direction to auto-spread. Give them a command to go build over *clicks* there, and they'll keep producing tumors until they reach there. Chronoboost could be set to auto-cast on a targeted building every time you had 25 energy.
If you really wanted to cut down on the finger gymnastics, you could also add in a stutter-step move command. You automatically move, stop, move, stop, move towards a direction. Engage at max range would be nice too. RA3's 'coordinate' waypoint system would be cool. You give a bunch of units commands, and hit activate, and they all move at once.
After that, the only thing I'd add in would be a programmable target priority system. So, when your units are attacking or stopped they target based on your chosen threat list, rather than the built in SC2 threat list. I'd have marauders automatically attack non-slowed stalkers, the nearest non-slowed zealot, then slowed stalkers, then everything else. Marines would always attack PDDs observers and banelings first, then whatever had the lowest health after that. Immortals would always attack armored, etc.
Uh, creep tumor and transfuse are both awesome. If MULE is good because you have the option to scan, then inject has to get bonus points for those.
In the mid- and late-game, you will be injecting with your hatch queens 100% of the time, so there is no choice there. In the early game though, there is some choice involved that hasn't been "figured out" yet as to what is optimal, e.g., mixture of injects and tumors for your first two queens.
Kambing on
@TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
0
Options
VariableMouth CongressStroke Me Lady FameRegistered Userregular
edited December 2010
maybe folks should just remember to chrono
I remember to mule because I hit my OC hotkey every 15 seconds or so to see if I'm producing. same as everything else... actually maybe more or less often depending. but when I see enough energy I mule unless I need to scan or need to save scans.
I don't get how toss checks their nexus and sees energy and doesn't use it. if you don't know what to use it on, use it on the nexus.
it's just weird to me. everyone does it... they can't be missing the energy that's there so why not spend it.
anyway... yeah, it's a game where your multitasking has to be up to a certain level before the awesome strategy of the game really comes into play. that doesn't make it less strategic, it just also makes it a real time challenge.
Uh, creep tumor and transfuse are both awesome. If MULE is good because you have the option to scan, then inject has to get bonus points for those.
In the mid- and late-game, you will be injecting with your hatch queens 100% of the time, so there is no choice there. In the early game though, there is some choice involved that hasn't been "figured out" yet as to what is optimal, e.g., mixture of injects and tumors for your first two queens.
that's what the third queen in the day9 build is for. Tumors, injects, and defense. :-)
i recall hearing that the very hard ai cheats resources in so
I didn't see anyone correct this yet, but no, only Insane cheats resources. I believe they get 7 minerals per worker rather than 5, or somewhere thereabout. You can see this by playing a quick game against an insane AI and then looking at the income tab in the replay, or maybe even putting an insane AI against a very hard AI and just spectating (can you do this?)
I've done the replay thing and I am 100% positive that insane AI gets more money than you do or other AIs do.
Uh, creep tumor and transfuse are both awesome. If MULE is good because you have the option to scan, then inject has to get bonus points for those.
In the mid- and late-game, you will be injecting with your hatch queens 100% of the time, so there is no choice there. In the early game though, there is some choice involved that hasn't been "figured out" yet as to what is optimal, e.g., mixture of injects and tumors for your first two queens.
that's what the third queen in the day9 build is for. Tumors, injects, and defense. :-)
Nah these decisions happen before a non-hatch queen goes up. For example here are valid openings for your first two queens
1st queen inject main -> 1st queen inject main -> 2nd queen inject natural
1st queen inject main -> 1st queen tumor main -> 2nd queen inject main
1st queen inject main -> 1st queen inject natural -> 2nd queen tumor main
The first typically gives you too much larva than you have resources to spend and limits the ability of your main queen to help defend an early push. The second causes your drone distribution to be skewed (more at main than nat). The third gets your early creep for defense and an even distribution of drones at main and nat but your queen's energy pools are out of sync.
Kambing on
@TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
i recall hearing that the very hard ai cheats resources in so
I didn't see anyone correct this yet, but no, only Insane cheats resources. I believe they get 7 minerals per worker rather than 5, or somewhere thereabout. You can see this by playing a quick game against an insane AI and then looking at the income tab in the replay, or maybe even putting an insane AI against a very hard AI and just spectating (can you do this?)
I've done the replay thing and I am 100% positive that insane AI gets more money than you do or other AIs do.
Very hard AI is probably about platinum level.
I corrected it, but no one pays attention to my posts because I don't have a sc2sig profile. =(
Kambing on
@TwitchTV, @Youtube: master-level zerg ladder/customs, commentary, and random miscellany.
Posts
Dear god man. Imagine mass queens!
...larvae
ah yes, I see the confusion. I simply meant that you could queue them up. So like if I had forgot for awhile and my queen has 50 energy, boom I could poop two on there and then they would spawn consecutively
Yeah, I understand and I still love SC2 just the way it is.
My point is that if we already had these features it would be different. (money only deducts when a unit starts, queens can automatically inject, buildings can be told to repeat-produce) Nobody would say "hey, you know what would be great? If we could only produce one unit at a time efficiently, so we had reselect our base every 17 seconds and click again." Or "oh man you know what zerg needs? It'd totally be more fun if they had to select all their queens and manually click on their hatcheries ever 40 seconds."
This is a good starting point. You could add Carriers are set to auto-cast make interceptors. The casting spells onto unit portraits that was removed because some koreans complained that it was too easy. Not being able to warp in units on the mini-map. No autocast on the vomit. No formations. No more complex orders than move, attack, hold, stop, patrol.
I'm not sure why people consider tedious repetitive tasks to be part of a strategy game, real time or not. Like when Dhal got all pissy when I said that it would be nice to pull down the replay bar menu for your side. check out your production tab. Check out your units tab. This is all easily available info that would make the game more reliant on strategy, and less on memorized rapid keyboard movements. Sure, its real-time, but usually anything resembling "strategy" goes to the wayside while you're off babysitting everything.
It's the reason that I *hated* War3. The keyboard gymnastics with your heroes, and the damn Creeps.
nobody is asking for a python API. Just make the damn thing do things that are repetitive nonsense. (hey, i need 25 sets of 'lings.... time to mash z, couldnt there be a combo that said select all, make all $foo?)
earlier Red Alerts and C&C's (not this latest one, it sucks ass) had more advanced features, such as the sidebar where you could just pick what units you wanted to be created, and it would make them from wherever was best, including making partial units as resources came in.
having hands that would pleasure your mother in less than 15 seconds is a neat skill, but I'm not sure where that became the focus of strategy games.
Joe's Stream.
the zerg macro mechanic (vomit) is easily the most powerful in the game. It creates exponential growth as soon as you have a queen (you double (plus) your unit production). Then another hatch&queen doubles that again. If you can manage to get to 4 bases, plus queens, you can produce 9x(plus a little) (28 larvae a vomit round vs 3) what you could off of one hatch, at once.
terrans have mules and reactors. protoss has chrono and gates. neither come close, unless you completely overbuild gates for late game replenishments.
Joe's Stream.
play sup com
if you want sc2's mechanics
play sc2
and for vomit
i just think it is a high floor, low ceiling type of thing, which i personally dislike
it doesnt really warp the balance at higher levels, so its fine where it is, but i personally would like zerg's macro mechanics to be done in a different way
i love the design of the queen as a whole though
FTFY
This.
I used to have such a hard time just keeping my cool when banshees or mutas showed up in my base. Now I'm not really afraid of either, although, 2port is still a little hard to hold off.
If you macro right you should never fall behind the Very Hard A.I., actually. I do great against the Very Hard until I make a macro slip. I almost always beat the Zerg on Very Hard, 50/50 against my mirror, and I have a hella time beating the very hard terran. For some reason, one little slip-up on my macro vs terran is a way bigger deal than against Z or P.
Huh. The rate your production increases is linear in the number of hatchery-queen pairs that you add on. The power of inject doesn't have to do with the amount of larva it adds. Overall production of a single zerg base compared to the comparable set of buildings for terran/protoss --- counting both workers and army --- is roughly equivalent. The power lies in the fact that all those larva can be dedicated to either drones or army, so a zerg's pure army production or drone production is better than the other two races, but their overall production is equal.
Actually, this version of the game already exists in not just one, but two forms!
Desert Strike and Nexus Wars
Personally, I love Desert Strike - I just play it to wind down from playing 1v1s, because I'm not nearly as interested in whether I win or lose and I get to watch all the battles I want. It's also not the shit show that Nexus Wars is...
On one side of the continuum you have Starcraft II as is, on the other you have battle chess (where it's entirely turn based but hey! the battles are animated!)
While improving the AI and self-designating priorities would be nice, programming fighting AIs is a different game than a RTS. Without the element of multitasking-focus distribution-unattended things screw up, it just isn't an RTS anymore. That being said, there are a fair number of default decisions that Blizzard make that take an unrealistic amount of micro to counteract (like SCV repair, which is finally being fixed.)
3clipse: The key to any successful marriage is a good mid-game transition.
http://www.youtube.com/watch?v=e4MSN6IImpI
i really don't think its linear... its geometric, and damnit, i've had graph/discrete math conversations with you, so maybe i'm missing something here, but 3->7->14->28 looks geometric to me.
Joe's Stream.
The zerg AI is bad in general because it doesn't know how to harass with lings or mutas. So it almost always throws a ling/roach/hydra ball at you which all 3 races can stop easily.
The other two races are more proficient at straight-up attacking, so the AI fares considerably better with them.
3 (1hatch0queen)-->7 (1hatch1queen)-->14 (2hatch2queen)-->21 (3hatch3queen)-->28 (4hatch4queen), etc.
Adding additional queens adds no extra larvae; adding additional hatches without queens adds 3 larvae per 40 seconds.
EDIT: The real strength isn't that eventually the production becomes overwhelming, it's that when you're at maxed food and you save up a bunch of larvae you can end up with 60-80 larvae sitting around, so as soon as you lose units you can make new ones for 100+ food. To match that with Warp Gates Protoss need about 40, to match that with Barracks Terrans need about 20 reactor + 20 tech lab. Those are both massively more expensive investments, since larvae only cost energy and a Hatch can have up to 19 saved up.
1. Awesome video! Thanks for linking it.
2. The reason you are getting geometric is because you are assuming a double expansion after two bases. You can theorycraft if this is easier/more effective for Zerg vs. other races if you want, I don't have enough experience with it to judge, but if you consider the hatchery / expansion rate to be linear, your growth rate is linear. If you consider 4 hatcheries on 3 bases, it works in the short term (which may be all you need to win the game) but the resource collection rate will no longer allow geometric progression, although I'm pretty sure quadruple expanding is a rarity anyway =P.
3clipse: The key to any successful marriage is a good mid-game transition.
Hrm. I suppose this might be a "growth relatively to what" thing, but:
1 hatch larva production = 1/15 larva per second
1 hatch+queen larva production = 1/15 + 4/40 larva per second
So the amount of larva you produce overall assuming perfect macro is:
1/15n + (1/15 + 4/40)m larva/second
where n is the number of bare hatches you have and m is the number of hatch/queen pairs you have. And the above equation is linear in the number of hatches you are producing from.
I'm in complete agreement, now plot production capacity over time (i almost always see a double expand from 2->4 or a macro hatch and a 3rd base (4 total)), and let me know what the curve looks like.
Joe's Stream.
EDIT: Also, a double hatch or a 3 bases 4 hatch makes it still a linear progression, because you're just adding 2x instead of 1x at that point; the discreet unit is still x, not 2x.
realistically, you pretty much should be MULEing every time too but you have the choice of scouting/detecting or fixing a blunder with supply calldown as well
and clowns can chronoboost whatever the fuck. of course, you need to be chronoboosting just as much as you need to be MULEing and all, but T and P have options and choices
queens just gotta spit on shit every 40 seconds. it's whatever, though
zergs op
(fixed)
Did I do it right?
Steam ID
To be fair, almost any hatch before pool build will move you from 4 larvae a minute to 20 larvae a minute almost instantly.
Protoss might be better off if people actually spent their bloody chronoboost.
i agree with your math, it looks to describe the larvae/second, but I'm not sure that it describes the point I'm trying to make, which maybe I'm completely off on this - its happened before, but I think that looking at it from game time elapsed as your X axis, and larvae production potential as your Y axis, I still think that you get a exponential curve.
(editted to reflect reality of fast expand)
1 Hatch:
3 larvae/45 s (lets call this L)
2 Hatch
6 larvae/45s (2L)
Add a queen:
10 larvae/45 s (3L+1)
Add another queen
14 larvae/45 s (4L+2)
Expand+queen
21 larvae/45 seconds (7L)
Macro Hatch+queen
28larvae/45 seconds (9L+1)
So, for 1500 minerals (not factoring in pool, or OL costs), you now have a little over 9x of your original production potential. If the time between new hatcheries/queens is equal (and I would argue that 3&4 go down near simultaneously) , you are looking at a geometric increase in larvae production potential over game time.
Joe's Stream.
the choice is drone or units, inject or spawn tumor/transfusion
SC2 is not just a strategy game...it is a multitasking game. Who can keep up with doing more things all at once. Believe it or not some people find the level of multitasking present in SC2 appealing (I do for instance). SC2 strikes a nice balance between pure strategy and pure multitasking. This is part of the fun of SC. Sure some people don't like it, but then maybe you should play those other games with more automation rather than complaining about SC. People who feel the lack of automation gets in the way of the strategy are missing out on the point of a lot of SC2's gameplay. Half of the point/fun is in the number of tasks you have to keep up with. I wouldn't have SC2 any other way.
the discrete unit is your basic hatch 3/45s.
Joe's Stream.
all of that sounds terrible to me
you remember to spend it in the very beginning, and then when you need tech/units in a hurry.
maybe a secondary function for nexus energy would be nice... shield regen... emergency scan....
Joe's Stream.
In the mid- and late-game, you will be injecting with your hatch queens 100% of the time, so there is no choice there. In the early game though, there is some choice involved that hasn't been "figured out" yet as to what is optimal, e.g., mixture of injects and tumors for your first two queens.
I remember to mule because I hit my OC hotkey every 15 seconds or so to see if I'm producing. same as everything else... actually maybe more or less often depending. but when I see enough energy I mule unless I need to scan or need to save scans.
I don't get how toss checks their nexus and sees energy and doesn't use it. if you don't know what to use it on, use it on the nexus.
it's just weird to me. everyone does it... they can't be missing the energy that's there so why not spend it.
anyway... yeah, it's a game where your multitasking has to be up to a certain level before the awesome strategy of the game really comes into play. that doesn't make it less strategic, it just also makes it a real time challenge.
real time strategy.
that's what the third queen in the day9 build is for. Tumors, injects, and defense. :-)
Joe's Stream.
Smite: Plutoniumwombat
I didn't see anyone correct this yet, but no, only Insane cheats resources. I believe they get 7 minerals per worker rather than 5, or somewhere thereabout. You can see this by playing a quick game against an insane AI and then looking at the income tab in the replay, or maybe even putting an insane AI against a very hard AI and just spectating (can you do this?)
I've done the replay thing and I am 100% positive that insane AI gets more money than you do or other AIs do.
Very hard AI is probably about platinum level.
Nah these decisions happen before a non-hatch queen goes up. For example here are valid openings for your first two queens
1st queen inject main -> 1st queen inject main -> 2nd queen inject natural
1st queen inject main -> 1st queen tumor main -> 2nd queen inject main
1st queen inject main -> 1st queen inject natural -> 2nd queen tumor main
The first typically gives you too much larva than you have resources to spend and limits the ability of your main queen to help defend an early push. The second causes your drone distribution to be skewed (more at main than nat). The third gets your early creep for defense and an even distribution of drones at main and nat but your queen's energy pools are out of sync.
I corrected it, but no one pays attention to my posts because I don't have a sc2sig profile. =(