[A] Starbow - Page 370

On August 26 2013 18:56 Hider wrote:


Well I looked at the difference in the current Sbow econ between 2nd worker income and 3rd worker income, which currently is this;
0.5 - 0 = 0.5

If the difference is even larger (e.g. if it is 1/1/0.25 --> then its 0.75), then players will just be rewarded for taking even faster bases.




This is the new suggested? This indeed is a drastic change, and will definitely require mineral income per path reduced from 8 to 7, and probably also just 8 mineral pathes in main and natural.

But ofc it will make possible to take bases at a slower rate.

On August 26 2013 19:13 Zaphod Beeblebrox wrote:
I think an ideal ratio would be something like 1/0.8/0.5

This is hard to implement, but it could be possible if we change worker AI somewhat. I belive that workers atm will stand still at a mineral patch and wait if another worker is currently mining it and there is no other worker waiting there (and there are no patches without assigned workers).
This has the effect of making this second worker not waste any time trying to optimize where it mines - this optimization happens automatically once the worker pair gets into rythem. The AI also makes workers know how many patches there are being mined, so they will always spread ideally untill the second round is produced.
If we instead made workers move to find new patches instead of waiting, this will lower the effectiveness of the second worker, allowing us to make workers slightly faster and make a third worker actually contribute to income.

Would this work, or am I just being silly right now?

And yet we never reach a state where Protoss core units feel good and interesting enough vs the other races. It is the same dilemma over and over again.

The main problem IMO is mass Mutalisks. How can P deal with mass Mutalisks, especially when he must defend 4+ bases with clumsy Dragoons? Unlimited selection of Mutalisks can fly between bases easily. Mass Corsairs and mass Cannons is the only way to go? Mass Null Wards? (They will be fixed by XiA)

A design related problem though can probably be fixed with a design related solutions.

So the values you now suggested for the Immortal + Stalker have been properly tested in the unit tester and in the editor? You have set a desired outcome and you have tweaked those values so we now know this will be good? Or are your numbers just another blind suggestion?

Yes. And how is that solution?

On August 26 2013 19:59 Zaphod Beeblebrox wrote:
If we model the Immortal mostly after the Dragoon we will have most of the interesting interactions the Dragoon had. Only problem then is the Stalker...
I said it before, make it a tad more fragile and give it medium(?) armor type - then adress its mobility and ability to be agressive.

I like the idea of Stalkers working best in groups of 3-8 due to their supirior mobility to all other protoss units (maybe give a speed boost with blink). Being fragile they will avoid tanks, lurkers and big groups of units in general, but with strong attacks be brilliant at handling Mutalisks and opposing mobile hit squads in general.

I did not change my system I just corrected an error. Like I said numbers can be tweaked. My most important point however, is that if you want the third worker to actually produce any income then the second worker will always double you income compared to having just one.

so we avoid scenarios where zealots rape speedhydras and where there is no reward for adding in stalkers. The basic unit relationships must work pretty well before the game can be fun to play.

If dragoon is added, then we need to see how it fare vs tanks, Vultures/marines/maurauders etc in large and small scales, w/ and without support. Severeal tests has to be made.

Speed Zealots are suppose to "rape" speed Hydras. (Unless Hydras have a critical number)

Yeah of course. But in that case I can look at the BW relationships. See how the relationships are there between Goon - Tank - Vulture - Hydra etc, and just steal it. Its a good foundation that works. It will probably need some minor tweaks to fit with the SC2 engine and with some other factors of the game.

But I do not say that the Dragoon must or will be added. I am just starting to seriously consider that solution, since there are so many problems connected to the Stalker + Immortal. (Even though Decembers new Immortal solution will help with some things at least.)

On August 26 2013 20:08 Hider wrote:


Hmm normal/medium armor type. Interesting idea. Will make it be able to tank a bit better vs tanks. Prob not enough to make it good enough. If combined with a small tweak/buff to blink, then yes. With new econ mobility units receives an indirect nerf, so a blink buff wouldn't be break the game offensively. However, that is likely to also make them too good at defending harass.

Default SC2 eco isn't terrible in terms of versatility. Main problem with it is that you make too much money.

On August 26 2013 17:40 Hider wrote:
Anyway, I am gonna list up my suggested patch notes;

Protoss
- Stalker damage increased from 14 vs normal to 16 vs normal.
- Stalker receives new type of abilitiy that buffs blink in in the intended usage. It will likely need quite a bit of testing to make it feel right and balanced.
- Reaver splash damage nerfed significantly.
- Archon HP/shield reduced change from 10/300 to 140/140 which will make EMP less important/strong vs protoss, and this will make help seperate when you want to make HT's and when you want to make Archons, as I personally dislike that EMP's kinda counter both units.
- Warp prism movement speed reduced from 3.75 to 3.2 Upgrade increases speed from 3.2 to 3.5


Zerg

- Dark Swarm replaces the Sc2 binding cloud with 75% range reduction.
- Lurkers might neeed a small HP buff as new Immortals are quite strong.
- Ultralisks might need a HP buff with Dark Swarm removal/nerf and immortal buff.

Terran
- Turret cost increased from 100 minerals to 125 minerals (or damage further reduced). Turret works okay'ish at the moment vs zerg and protoss (as you don't want to overmake them), however with new economy I fear that instead of investing into additional expos, you will use the spare minerals on a faster turret wall off which will reduce action in all matchups. Given that Sentinel openings and warp prism openings have been nerfed a bit, this change seems fair for early game as well.

5B) With the current balance, protoss can just rape banshees by researching speed on Scouts, but if that ability is removed, he will instead prefer to mix in blink stalkers to try and defend vs the banshees.

6b) Assuming the protoss player stabilizies/kills the banshees with his stalkers, they become quite oboselete with current balance. However, with some kind of offensive small squad threat to them, they will be usefull for harassing relatively weakly defended outer expos. Thus they can now function as an anti deathball unit.

I believe this type of balance/dynamic is a lot more fun, and rewards scouting/reacting/decisionmaking, harass, "fun micro" (rather than spamming) and anti-deathball play too a larger extent.

But what exactly have we done? Immortals have been UP vs mech ever since it was implemented (maybe there was a period with absorb where it was fine). If we had spent more time in the unit tester before adding it into the game, I think criticism would be quite different.

Further, the stalker + zealot domination problem vs zerg is also purely a probelm attributed to lack of time spent in the unit tester. What needs to happen is that we define the desired outcome, and then test and tweak untill we are there. Regardless of what units would be in the game, we will be in a mess with loss of complaints when balance is so flawed as it currently is now.



Indeed. So the two factors to consider here (assuming stalker gets balanced in PvZ) is
1) Mass mutalisk PvZ that likely will be super frustrating/boring to play against for protoss without the blink stalker
2) The boringness of the blink stalker in TvP mech.

The former might be something that can be fixed by sentinel at robo and giving nullsphre a big buff vs mutalisks.
The latter is IMO less of a issue if it only relates to TvP mech and if you can be offensive with all other units. But it will also be harder to fix it.

Buffing stalker damage and blink is playing with fire. If stalkers get better vs hydra then we will see blinktimings, no doubt about it, if the stalkerball is big enough lings wont do shit as they wont have the surface area, hydra's wont do shit because you can keep trading without losing stalkers, muta's obviously will not be the answer either => you're forced into a lot of lurkers => safeguard deals with that + you're really mobile, there are probably enough places on the map to go around the lurkers and surpass terrain unless we only have maps with chokes => map design limited.

I will still take scouts over stalkers vs banshees any day, banshee can abuse terrain, bait out blinks, then you have to wait for blink cooldown, etc. While with the scout, i can just let it follow you so you actually never will be able to really harass anything.
I don't know why you think chasing banshee's is "fun micro", just like defending vs muta, blinking in and out of bases correctly and scouting the movements so you don't get caught offguard at some angle and meanwhile try to scout what the enemy's next move is, is one of the harder non fun things to do. Look at WoL pvz, big mutaballs were one of the lamest things to see, toss couldn't move out unless he went for the baserace.

About turret rings: again, if someone doesn't want to be dropped, he will make sure he doesn't get dropped by all means, what you aim for is some trigger that makes you unable to make more than X turrets per base. And anyway, as said before, if a player makes many turrets, change your plan and abuse the fact that he spent a lot of minerals on base defenses, even if it means it is 'less exciting'. Showing players what you are capable of if they over-invest in base defenses will make them stop doing so, the power is in your hands.

Early zealot/stalker works because they both deal good enough with lings, the dragoon however is not good vs lings => clear weakness and with dec's awesome new attack system the zerg can micro his hydra's vs the zealots and chase dragoons with lings.

You cannot give an extra ability to the stalker to be better vs small groups of tanks, then we are back at the problem where terran just died vs blink-openings, unless you require an extra upgrade that probable comes late enough for the P to choose better harassment options.

I also agree with the fact that blinding cloud counters siege tanks harder than dark swarm and i don't think it's a realistic spell for the defiler as well, its casting range would need to be too great for them to land anything without being instakilled by siegetanks first.

On August 26 2013 22:47 Foxxan wrote:


mind explaining this?
---------------

a big flaw, of sc2 economy is the second worker - 1/1~/?


overall this means you rather have 16 workers on one base than 8workers spread on two bases (8min field)
A reason additional bases is not taken for the minerals after the normal 3base is that supply = death or life.
The normal is 65 workers on 3base, and if you take a 4th or 5th its for the gas purely.

One big reason for this is not the economy initself though, its that supply needs for your deathball

Ive said this before and i will say it again anyway:
You have 16 workers
Two bases
8 mineral fields on each

If the income encourage that you spread the workers out (eight on each base), this adds a good depth for gameplay

I feel this is very important!

On August 26 2013 23:25 decemberscalm wrote:

You are right about its flaws. But with SBOW economy you've got no income curve. It is 100% when you add first worker per patch. Second worker is always something like 64%. No third worker. Always 1:.64.

With SC2 eco, once you reach 16 workers (yes you are correct, first two workers are 1:1) then you get a curve. Each worker brings less and less money. Optimal saturation for SC2 is roughly 24. Third worker brings less and less each one you add.

This is similar to how in BW, past 1 per patch each added worker adds less and less each time, but you can still over saturate.
You cannot over saturate in Starbow. Most you can do is make sure you've got 1 worker per patch, and then add on top of that.



edit:

@Economy

Me and Xiphias have tried out many different methods.
Nothing works for what we want due to smart pathing.

Unless I can find anything better I am pretty sure we will need to resort to worker bounce to improve the economy.
We played with tons of different methods.
If you have a third worker able to mine, the second worker is always 100% efficient. (SC2)
If you have a second worker mining under 100% efficiency, the third worker cannot mine. (Starbow).
We played with harvesters that delay before returning their money, still no good.

This method looked interesting. They experiment with different distances for the mineral fields to give different efficiency levels. I don't think I can properly test it when they don't have an explanation on how to set it up.
http://www.teamliquid.net/forum/viewmessage.php?topic_id=393800


So in summary, worker bounce seems the way to go for now.