EPTObviously units have more stats than just HP and DPS, but it's nice to see what the "best" units in terms of pure hp and dps are.
Starcraft 2 units: a cost-effectiveness analysis. - Page 2
| Forum Index > SC2 General |
|
koppik
United States676 Posts
Obviously units have more stats than just HP and DPS, but it's nice to see what the "best" units in terms of pure hp and dps are. | ||
|
BigFatRoAcH
Japan90 Posts
[/QUOTE] I think you forgot about the one attribute: attack range. A tank, when sieged up can attack from 13 distances away and a roach can only attack from the range of 3. Throw in a few meatshileds(like a Thor or a couple of marauders) in front of the tanks and all your roaches would die before they can get to the tanks. | ||
|
carwashguy
United States175 Posts
On July 24 2010 14:19 koppik wrote:Ultralisks should be computed vs. armored, though. You sure? Liquipedia doesn't show that bonus. On July 24 2010 14:22 BigFatRoAcH wrote: I think you forgot about the one attribute: attack range. I only used the data I could reliably quantify. If you can think of a way, please don't hesitate to tell me. | ||
|
FC.Strike
United States621 Posts
On July 24 2010 14:25 carwashguy wrote: You sure? Liquipedia doesn't show that bonus. I only used the data I could reliably quantify. If you can think of a way, please don't hesitate to tell me. If liquipedia doesn't show the Ultralisk +armor modifier, it's out of date. I believe it's 15 damage to non armored, no idea what the top end is. | ||
|
ZomgTossRush
United States1041 Posts
Nonetheless, thanks for the charts as it can help more advanced players find interesting cost-effective unit compositions. | ||
|
Sadistx
Zimbabwe5568 Posts
On July 24 2010 14:01 carwashguy wrote: I need more convincing. Please show evidence. Your numbers are dps/R values in a vacuum. They ignore range, mineral/gas ratio, tech cost and usability on any given map. The evidence is reality - it's how you made the table. And your conclusions are wrong because they don't take it into account. | ||
|
nujgnoy
United States204 Posts
15 +25 armored | ||
|
FC.Strike
United States621 Posts
On July 24 2010 14:28 nujgnoy wrote: Pretty sure ultra does 15 +25 armored I wanted to say 40 damage to amored, and since I know that the low end is 15, that sounds right. | ||
|
Lennon
United Kingdom2275 Posts
Maybe this data could be used for future balancing. Remember that it's also about how you use the units. | ||
|
BigFatRoAcH
Japan90 Posts
| ||
|
fabiano
Brazil4644 Posts
In the end you will always find out what is cost effective after experiencing it. For example, I've never looked on Liquipedia what is cost-effective against hydras, but after playing games I found out that a bunch of dragoons alone could not deal with a bunch of hydras. Then I had to add some zealots/ht/templar and the mix would make my goons very cost-effective. I dont know if I made myself clear, or even If I am misunderstanding the concept of cost-effectiveness, but thats what I think. Dont stress yourself too much on numbers  | ||
|
koppik
United States676 Posts
Also, I think maybe it would be better to maybe have a separate column where gas is rated as more important as minerals. I'm not sure, but I think, from one base, you get 2.5 times the mineral income compared to the gas income. | ||
|
PrinceXizor
United States17713 Posts
| ||
|
nujgnoy
United States204 Posts
On July 24 2010 14:28 Sadistx wrote: Your numbers are dps/R values in a vacuum. They ignore range, mineral/gas ratio, tech cost and usability on any given map. The evidence is reality - it's how you made the table. And your conclusions are wrong because they don't take it into account. I think there's only so far that a calculation can do. But it's still helpful to explore how far numbers can take you. For example, siege tanks have a range of 13. But they have a 4 second siege mode. And their vision is only 11 iirc. So how would someone take all that into account? DPS, cost, dps/resource, these are things that are constant, whereas range etc are attributes that are influenced by the specific circumstances of a sitution a unit is used. Of course, this thread won't be the definitive end-all analysis for units, but I think these constant and quantified values are a good foothold to explore right now as we find a way to find more practical values for the units. I think everyone will have a distinct subjective opinion on the cost effectiveness of units after looking at the tables. But as for the OP, instead of outright labeling him wrong, credit should be given to finding the numbers and the initial inferences, from which we can make further conclusions or form different opinions. | ||
|
Gnarwhal
United States33 Posts
On July 24 2010 13:50 youngminii wrote: Yes, 1 gas is much more valuable than 1 mineral. Gas is limited to around 224~ mined per minute per base whereas minerals are mined at up to 800~ per minute per base not including mules which add a LOT of mineral mining. Lots of high tech units require a lot of gas and you can see that now 1 gas is a lot more useful than 1 mineral. You changed your point by saying in the end that one vespene is more useful than one mineral. This may be true, but does not determine value. One resource is one resource, equally valueable. | ||
|
DannyJ
United States5110 Posts
On July 24 2010 14:37 Gnarwhal wrote: You changed your point by saying in the end that one mineral is more useful than one vespene gas. This may be true, but does not determine value. One resource is one resource, equally valueable. hrm? Where does he say 1 min is more useful than 1 gas? | ||
|
carwashguy
United States175 Posts
On July 24 2010 14:33 koppik wrote: Ultralisks are definitey 15(+25 armored). Updated spreadsheet. Thanks for the heads up. On July 24 2010 14:33 koppik wrote: I'm not sure, but I think, from one base, you get 2.5 times the mineral income compared to the gas income I'll sleep on this one. On July 24 2010 14:28 Sadistx wrote: Your numbers are dps/R values in a vacuum. They ignore range, mineral/gas ratio, tech cost and usability on any given map. The evidence is reality - it's how you made the table. And your conclusions are wrong because they don't take it into account. You misunderstood me. I meant: please show evidence which refutes my inferences. I will update accordingly. | ||
|
Sadistx
Zimbabwe5568 Posts
On July 24 2010 14:35 nujgnoy wrote: I think there's only so far that a calculation can do. But it's still helpful to explore how far numbers can take you. For example, siege tanks have a range of 13. But they have a 4 second siege mode. And their vision is only 11 iirc. So how would someone take all that into account? DPS, cost, dps/resource, these are things that are constant, whereas range etc are attributes that are influenced by the specific circumstances of a sitution a unit is used. Of course, this thread won't be the definitive end-all analysis for units, but I think these constant and quantified values are a good foothold to explore right now as we find a way to find more practical values for the units. I think everyone will have a distinct subjective opinion on the cost effectiveness of units after looking at the tables. But as for the OP, instead of outright labeling him wrong, credit should be given to finding the numbers and the initial inferences, from which we can make further conclusions or form different opinions. You misinterpreted my post. I never said his numbers are wrong, only the conclusions. When not expecting an air threat, heavily favor Roaches to Hydralisks. Blizzard must want us to build Battlecruisers in the late game. They're a great bargain. lol... On July 24 2010 14:41 carwashguy wrote: You misunderstood me. I meant: please show evidence which refutes my inferences. I will update accordingly. Your inferences were made ignoring factors that are crucial to the cost effectiveness of units in any particular game. I listed those factors in my previous post. | ||
|
Tempora
United States78 Posts
TvZ, almost invaluable in early game if opponent uses zerglings, also EXTREMELY effective at harrassing mineral lines. Lame? afraid not xD good use of micro makes them pretty scary in PvT too, because they are so much faster than zealots. also FORCES opponent to make stalkers which are in turn eaten up by mauraders that you knew to produce BECAUSE you went hellion. its the only reason im good with all three races, or else i would never play terran, ever. ever. ever.... | ||
|
Calamity
Canada161 Posts
On July 24 2010 14:01 carwashguy wrote: I need more convincing. Please show evidence. These values, they can help a lot but some of your conclusions about each units based on this data is just not accurate. For example, Battlecruisers are amazing but you need a fusion core, time to build the battlecruiser, and the upgrades you would have for air units at the time. If you transition from bio or mech to battlecruisers, you wouldn't have 3 / 3 upgrades for air while your enemy most likely would. Thors are strong units that do tons of damage to air and ground units, yet again time and the slow movement of Thors make them weak in some situations compared to a Bio army. Hellions are great for killing workers. You're values don't factor in speed, and ability to micro them via kiting. Hellions under micro can kill an entire expo full of drones with the pre-igniter upgrade. As for spine crawlers, they might be cost effective, IF they moved and attacked. They can only attack while static which would explain why they're so strong. They are also limited by creep I think? Phoenixes might be slightly better against air units than the stalker, but stalkers are from warpgates which can be easily produce since you don't need make stargate. It's also easier to transition from stalker to phoenix. Some do make sense though, like how facing a bunch of armored units immortals and void rays will crush all. Massed voids will probably crushed every Terran armored unit except for a bunch of Yamato Cannon Battlecruisers (even then you'll need a crapload of BCs). But if Terran had a Marauder + Ghost army. Sure there's a bunch of armored units so i must use Immortals. You will get killed since Ghosts will just emp you're immortals and they're basically dead meat (dead metal) to the marauders. This reminds me, spells of several casters in this game would make me to believe that they are more effective than this data would suggest. AoE, how many units are caught in the spell, the cooldown, the energy per damage ratio, number of times attack can be used, if it does any damage at all (sentries) etc. | ||
| ||
