|
I was curious about which attack upgrades actually had a measurable difference (ie. number of hits) against various other units. So I made a program to simulate matchups and out came a giant chart.
Extremely well balanced, almost every upgrade is useful against every other unit. Kinda amazing actually. Most of all the first upgrade (+1) makes a big difference.
Yes it takes into consideration special things like hardened shields, number of attacks, zerg regen, etc, and it's up to date with patch 15.
Some interesting ones like +1 on mutas dropping stalker hits to kill from 20 to 17! that and +2 on roaches being less than stellar in a ZvZ matchup (no effect on lings or on hydras as compared to +1).
"Upg" column is number of hits, "Time" column is how many seconds it takes to kill one, "Win" column is if this unit beats the other in a 1v1.
Will probably do a "cost for cost" chart later on.
EDIT: Added armor upgrades and shield upgrades. Only on the XLS file for now. As expected, armor/shield upgrades are really only useful against low damage units like zerglings or marines, and have no effect in some cases, for example, Zerg against Terran mech.
Here it is in Excel: http://www.tacticstudios.com/downloads/sc2_upgrades.xls
Here it is in Text: http://www.tacticstudios.com/downloads/sc2_upgrades.txt
Here is the source code: http://www.tacticstudios.com/downloads/sc2_src.rar
Karma
|
Really nice work man!! :D
|
Tanks, that was actually just the thing I was looking for  A couple of things are wrong though as according to your charts Sieged Tanks can shoot BC's/ Banshees and Vikings while Thors are apparantly unable to hit air 
Maybe you did the 'ol Switcheroo?
|
Might want to fix your text file - Long names push all the columns to the right.
|
On June 06 2010 10:52 Saechiis wrote:Tanks, that was actually just the thing I was looking for A couple of things are wrong though as according to your charts Sieged Tanks can shoot BC's/ Banshees and Vikings while Thors are apparantly unable to hit air Maybe you did the 'ol Switcheroo?
He fixed whatever you were seeing. I can't find the discrepancy.
|
Saechiis for the thors and vikings I split them into two: Thor "Air" and Viking "Air". I should probably couple these together. For the tanks I'm not sure where you're seeing that, maybe reading it backwards? The ones on the left are the attackers, right is the defender. For me it shows:
Siege Tank Viking Air Unhittable
Siege Tank Banshee Unhittable
Siege Tank Battlecruiser Unhittable
|
Good work but does it take into account bonus damages? For example, I see it says Baneling takes 2 hits to kill a Zergling, which is false.
|
Very cool. I'm going to use this.
Any chance we can get workers put into it too? I'd be curious to see which harassing units could kill workers in less hits with an upgrade or two.
Also, you should freeze the top row using 'freeze panes' so it stays there while the rest of the sheet scrolls.
Thanks much =)
|
Wooops Good catch Ocedic. Yes it takes those into account but things that killed exactly (ie. 35 dmg to 1 zergling) didn't count correctly. Fixed!
|
"Thor Mutalisk Unhittable"
It's definately there in the Text version, the Siege tanks being able to hit air is fixed (or I'm blind ).
Edit: Oooh, I get it now ... but why the Thor vs Thor Air split? Seems logical that it uses Air attacks vs air and Ground attacks vs ground. Besides my confusion and blindness, great job 
|
wow kudos to you for doing this work
|
Missing Queen. Otherwise, amazing information.
|
Great job, but I think you accidentally set Banelings as a light unit in your program. Also, I think that you set reaper's HP to 40 instead of 50 (that's what the data looks like). Other than that, its pretty sweet.
|
On June 06 2010 11:29 sk` wrote:
Missing Queen. Otherwise, amazing information.
I would also like to see overlord and workers
|
You also have to take zerg regeneration into account, since zerg units always regenerate a miniscule amount of health(it's displayed as 1 extra health, since the game display rounds health up, but it's actually less) between any two attacks and since the engine is linear, you can't have two attacks happen at exactly the same time, so the only case where the regeneration does not apply is one hit kills.
For example: mutas die to 7x2 phoenix/viking attacks, instead of 6x2(at equal upgrade levels).
|
Wow. Mad props, this is absolutely amazing!
I'm really really looking forward to when you get the costwise compariosn up and running. Though it would probably be somewhat wrong since you will have large battles then, and then splash damage would be hard to interpet correctly, and given the fact that ranged units get off more hits as melee units are approaching. But give it your best shot and I'm looking forward to reading your results. Also would be nice to see if this thread has correct in that ultralisks have no costwise ground-counter.
I would also like to see a column for defensive upgrades. I guess this wouldn't be too much work, you just add 3 more columns with -1, -2 and -3 upgrades, corresponding to the case where the defender has the armour upgrades. There would actually be cases not covered where you have both attack and armour upgrades (not all cancel one-for-one), but for simplicity I think you should assume that there are only armour or attack upgrades present, not both.
Anyway. Great work and keep it up!
|
Wow! If you don't mind me asking how long did it take you to do this?
|
Nice chart indeed.
You could also add Ultralisk +2 armor and maybe point defense turrent/infested terran, etc.
|
thanks for all your work! :-)
|
Seriously awesome. How about static defenses?
|
|
|
|
|
|
EPT
