|
On July 25 2010 08:23 uberdeluxe wrote:Show nested quote +On July 25 2010 08:18 eivind wrote:On July 25 2010 07:51 Piousflea wrote:
---------
MYTH #4: "Thors hard-counter Ultralisks": BUSTED
Without any upgrades, one Ultra will beat one Thor with ~90 HP left. If the Thor uses Strike Cannons, it will kill the Ultra with 10 HP left.
With 3/3 upgrades, the Ultra will beat the Thor with ~106 HP left. If the Thor uses Strike Cannons, the Ultra wins with 1 HP left.
If multiple Ultras and multiple Thors are involved, the Thors lose horribly due to splash damage, regardless of Cannon usage.
Throw in any of the 100 chokes, units in way, buildings in the way, minerals in the way and the ultralisks will lose hard. Though I guess it is useful to know that in an ideal word ultralisks can kill thors. If you time your attacks well and have good creep coverage, you can usually pull off a nice offensive vs mech terran(where you see thors the most, since your facts aren't only pumping tanks) when the super immobile mech army is finally required to move out, and you can get a nice surround. Plus, mind control on the thors for bonus points 
Yeah, thors arent the biggest problem due to Infestor & speedlings, but the biggest problem is small maps combined with tanks.. but that discussion does not fit here  I should definetly try getting ultralisks more often though, but most of the times the game is already decided.
|
#3 and #7 are interesting. #3 tells me that Infestor-supported Hydralisks are a valid counter to BCs, since a ground Zerg is almost guaranteed to have more minerals than an air Terran (look at the recent IdrA vs TLO game2 KOTB).
#7 in particular is coooooool. I'm starting to think that Zerg players should have a minimum of 4 Queens, even on 2 bases, if they are being pressured by a Terran. Queens cost no gas, can transfuse and creep, and synergize well with Spine Crawlers and Zerglings... And they protect you from air cheese (provided you have detection against Cloakshees)! Whats not to like?
|
This is a UnitTester thread, so "no micro" is assumed. Obviously none of these are "real game" situations. So yes, everything is attack-move.
It is easy to test viking vs phoenix scaling. For 1v1, 3v3, 5v5, and 10v10 battles the phoenixes win more often than they lose. Vikings cannot "shoot and scoot" vs phoenixes, it's 2.75 speed and stop-before-shooting versus 4.25 speed and moving-shot. Large numbers give the vikings a chance to kill a phoenix or two before the phoenixes get in range, but this is canceled out by the viking's tendency to waste shots on overkill.
As far as capital ships go, battlecruisers are easy to test, broodlords and carriers are very hard to test. BLs tend to give insanely variable test results, if the broodlings make a tight wall the ground units don't stand a chance, if they don't then BLs are in trouble. Carriers are currently bugged in UnitTester so you have to build interceptors manually and testing takes forever. (hitting the button to auto-populate interceptors can allow one carrier to have 12-16 interceptors)
BCs vs Hydra
- 1 BC will always win against 5 hydras, and always lose against 6 hydras.
- 3 BCs will usually win against 15 hydras, and often lose against 16 hydras. (variable)
A 1 BC : 5 hydra ratio seems fairly constant
- Unburrowing hydras under BCs is not a good idea. The unburrowing animation takes long enough that BCs can actually fire ~2-3 shots before the hydras start shooting.
BC vs Stalker
- 1 BC will always win against 4 stalkers, and always lose against 5 stalkers.
- 3 BCs will usually win against 12 stalkers, and will sometimes lose against 13 stalkers.
This places the BC:Stalker ratio at just over 1:4.
- Unlike Burrow, Blink helps significantly. Blinking 12 stalkers at 3 BCs will win.
BC vs Marine
- Without shield or stim, 1 BC beats 13 marines and loses to 14 marines.
- With shield only, 1 BC beats 11 marines and loses to 12 marines.
- With stim only, 1 BC beats 11 marines and loses to 12 marines.
- With shield+stim, 1 BC beats 10 marines and loses to 11 marines.
- 3 BCs vs marines (with shield+stim) is variable. If the marines get a good concave, as few as 30 marines can kill the BCs. If the marines don't get a great concave, BCs can kill as many as 32 marines.
1 BC : 10 marines with shield+stim is fair.
However, unlike stalkers or hydras, marines have less range than BCs. This means that well positioned BCs can melt marines without taking damage.
|
|
|
Awesome thread!
Thought
Muta>Viking>Corruptor>BC
I guess its
Corruptor>BC>Muta>viking (bottom of the food chain! Who'd have thought?)
Not totally sure 3 muta>3 viking but the mineral:gas ratio is different so ehh.
Also means that for air dominance battles it goes like
Muta>Void rays>Corruptors>Everything but Void rays>
Looks like corruptor/muta with armour upgrades might work pretty well on an island map.
Then we have raven/viking vs void rays vs corruptor/muta i guess.
Not sure corruptors lose to void rays either though. I know corruptors rape when void rays aren't charged (+2 armour vs 5 damage) but when they get charged they die really fast.
|
On July 25 2010 08:46 Piousflea wrote:
Carriers are currently bugged in UnitTester so you have to build interceptors manually and testing takes forever. (hitting the button to auto-populate interceptors can allow one carrier to have 12-16 interceptors)
No they are not? Using the "storm" icon works fine for me if you are using this Unit Tester. You just have to use an a-move instead of a global attack command, otherwise all the interceptors form a doom missile that shoots out like cannon. Just select them all and load them up, and they will 1 by 1 fill up with interceptors.
Btw, a immortal, with the help of a warp prism can kill a thor (even if it has a 250mm cannon upgrade) with a little drop and lift micro.
|
wow the Thor and Ultra results shocked me quiet a bit....
|
God damn it I dind't realize that youtube video was a fucking spoiler, I just watched it thinking "wtf is this."
|
Ultras rape all Terran ground. Terran has no answer for Ultras, Mech is the worst. The best thing to do as a terran if there are ultras in play is to mass marauders and hellions (for the lings that are always paired with ultras) and try to stim kite them while getting BC's with gas. Ultras are freaking good.
|
Viking range against phoenixes is basically meaningless due to the fact that phoenixes don't have to stop to shoot...
|
On July 25 2010 09:21 lu_cid wrote:
Viking range against phoenixes is basically meaningless due to the fact that phoenixes don't have to stop to shoot...
If your viking is over other anti air units or a turret, extra range is very powerful.
|
Well if we start throwing other units into the equation...
|
Very very good post. Seems like you put alot of effort into it as well, testing several different scenarios for each myth. I tip my hat to you sir, genuinely.
|
On July 25 2010 08:51 Chronopolis wrote:Show nested quote +On July 25 2010 08:46 Piousflea wrote:
Carriers are currently bugged in UnitTester so you have to build interceptors manually and testing takes forever. (hitting the button to auto-populate interceptors can allow one carrier to have 12-16 interceptors) No they are not? Using the "storm" icon works fine for me if you are using this Unit Tester. You just have to use an a-move instead of a global attack command, otherwise all the interceptors form a doom missile that shoots out like cannon. Just select them all and load them up, and they will 1 by 1 fill up with interceptors. Btw, a immortal, with the help of a warp prism can kill a thor (even if it has a 250mm cannon upgrade) with a little drop and lift micro.
Hm, I found out what is bugged. With one carrier the storm icon works fine. However, if you have 2 or more carriers and hit storm, it sometimes generates interceptors that are not linked to any carrier that will attack-move all the way across the map by themselves, and can even keep flying if the carriers are destroyed.
|
On July 25 2010 07:51 Piousflea wrote:
MYTH #4: "Thors hard-counter Ultralisks": BUSTED
Without any upgrades, one Ultra will beat one Thor with ~90 HP left. If the Thor uses Strike Cannons, it will kill the Ultra with 10 HP left.
With 3/3 upgrades, the Ultra will beat the Thor with ~106 HP left. If the Thor uses Strike Cannons, the Ultra wins with 1 HP left.
If multiple Ultras and multiple Thors are involved, the Thors lose horribly due to splash damage, regardless of Cannon usage.
Nope. I am sure someone else has pointed it out but... sure, one on one in an open field the ultra will win. But you completely neglect too many factors, like how effective an ultra is to a terran blob. You probably ain't going to reach the thors before 1-2 ultras go down, and even then his thors can blast away at your 1-2 ultra that can engage in melee while the others look on. The hellions makes for good meatshields for thors&tanks.
|
How are you testing this? Do groups both start within range? Some examples:
#1: If Phoenix start within their range 4, they should win up to around 7, then Viking should win up to around 13. However, Vikings should get 1 shot off before being engaged.
#2: Again if Corruptors start within range they should beat Vikings. Kill time is lower by ~1 second. However the Vikings should again get 1 free shot, which closes that gap.
#4: Ultas starting in melee will beat Thors. If starting at range the Thors will get 2 free shots (122 damage). The kill time difference is only 1.3 seconds, which is roughly equal to the close difference. Likewise if the Thor waits until firing a single shot to start the Cannons they should kill the Ultralisk before dying even with full upgrades.
Now, in regard to calling them hard counters.... no. But I have tested with the parameters described above and noted a discrepancy on these points. When testing 8v8 Viking/Corruptor for instance the Vikings won handedly. 2 Corruptors died before the first Viking died and there were 3 Vikings left at the end. When testing an extremely large number (30+) the Corruptors did start to pull ahead. The problem stems from Vikings 1 shotting Corruptors from out of the Corruptors range. Same goes for Phoenix, but the extra speed makes it a little more fair (with minor micro you should only lose 2 Vikings in an 8v8). Phoenix also started to win earlier. From 1-7 Phoenix wins and at 12+ Phoenix wins.
Thor vs Ultralisk is interesting. The shooting time of the Thor doesn't interfere with the casting. So if both start out of range the Thor takes a single shot, then immediately goes into the 2 second cast time, then the channel. When this is done Ultralisks die before the cast time finishes (the second to last hit kills them). There's also a flaw with your logic in regard to large battles and splash. Why do the Thor need to take splash at all? They can easily spread out and negate the splash. Furthermore at any large number it becomes nye impossible for Ultras to win. The big funny one is 8v8. Try 8v8 and focus fire with the Thors. It will be 6v8 by the time the first Ultras can attack. It will be 5v8 before all Ultras are in range. It will be 2v8 before the first Thor dies and at the end 6-7 Thor will be standing. With some quick testing I was able to beat 12 Ultra with 8 Thor.
Edit: Noted you're saying 'no micro' so it's a non-game situation, but the point stands.
|
excellent thread! i hope the OP keeps findings up to date on the first page.
|
Good read and nice job doing all the testing, it's good to know all the myths and rumors that linger around the strategy forums. ;D
|
|
|
The unit tester is definitely a great way to get a feel for how units matchup against each other, but it's hard to make hard and fast rules with it. Everything changes when you throw the human element into it as well as other units.
But great job, let's see more of this! Love seeing comparisons and learning small things about units.
|
|
|
|
|
|
EPT
