IMBALANCED! - Introduction - Page 39

Excuse formatting... i gotta sleep.

I seriously doubt colossus are truly an OP problem in ZvP. I spent a good hour testing stuff and will share results below.

Logically, P have to beat Z with a gateway timing or colossus 95% of the time... so since all most people lose to is colossus, it would be a target for criticism. Can't criticize high templar b/c P never beats Z with it in a typical match.

Besides nothing in tournaments really showing an imbalance problem with colossus, there are several unit compositions that do fine against colossus + stalker. If Zergs can find a way to tech switch to heavy banelings, a few lings, + medium to light roach late game, that'd be the way to go IMO. Why?

1. Against a splash ranged army, you want unit variety so that splash doesn't kill much. In other words, first volley hit some lings miss others, next volley might hit some banes miss others, next might hit roaches, miss others, etc.

2. Banelings strangely are possibly the best answer vs colossus stalker. Especially with baneling drops, + corruptors or empty OLs to draw fire, during

the fight. With the unit tester stuff... about 100 food worth of army vs about 100 food... zerg trades armies pretty well if heavy baneling (40 to 55 banelings I think). The way banelings and lings go almost single file around roaches helps them avoid splash A LOT. FF is the only scary thing... OL drops and/or a couple of Ultras to break FF help a ton. If they don't have any FF (or if your ultra break most of them) it's gg.

Overlord drops seem pretty great in the fight b/c stalkers have medicore dps. And if they try to focus fire the lings/banes coming single file get on them pretty quick. If stalkers try to blink micro away, the colossus are left vulnerable due to their bad speed.

Another option is heavy heavy corruptor + mixed ground, mostly roach. This is probably the most practical and the most seen. Again I also recommend 2-4 ultras (I dont ever see this enough though) to break FFs and screw up colossus splash. You won't have gas to go heavy ultra and have leftover for broodlords anyway... plus Ultras wont dps this army well anyway unless you get a crazy surround or want to risk Ultras in OL drops.
You lose your Zerg ground army, but you can pretty much kill all colossus and get away with nearly all of your corruptors. Then you just have to rebuild

roaches and maybe make a few broodlords. Colossus take too long to rebuild and the Protoss is now under the gun.

One more way to deal with this army might be killing observers and burrow. Popping unburrowed in a non-line pattern is bad for splash. Engage with rest of army a second before unburrow and stalkers cant focus fire. I didn't test this, but it theorycraftily makes sense.

Finally, infestors have to have some serious potential. Out of time to test them.



Some of the WORST zerg compositions against colossus+stalker in the tests were

All roach. They just line up perfectly for splash. You want VARIETY in your army to make splash target on tiny sections of your army!!!!
8 to 10 ultras + roach. Ultras just spaz out too much in big numbers. (Sometimes Ultras in numbers above 12 did OK when I micro'd the shit out of Zerg though. Wasn't convincing though and had a strange tipping point. Would require more testing)
Heavy hydra. Obvious.


Disclaimer: Real tournament matches mean more than unit testing, obviously, because getting into the situation needs to be possible. However I stacked the fight slightly in the Protoss favor on purpose to make sure it was winnable by Zerg.

Obviously you often lost most of your ground as Z, but if you take out all colossus and half the stalkers I consider it a Z "win" since you can rebuild roaches and possibly roll P. The more banelings you have though, the easier it is to kill the whole P army. All Test battles were done in no choke + no flank options (5 and a half colossuses could fit across). This seemed neutral enough that players would engage on either side without huge disadvantages. So in real games with similar battles more choke is better for P and more flank options better for Z. No creep was used in all tests.

I did not test sentry heavy builds!!! But if they go heavy sentry, I'd recommend more reliance on banes in overlords or ultras in overlords to drop on FFs, burrow ultras (their own control grp), let the banelings and lings in.

Most tests were done with a little above or below 100 food armies (80 to 130 food per army). Protoss ratio was usually 6-10 colossus, the rest mostly stalkers (2 to 6 sentries mixed in with full energy). Tried putting in a couple of immortals here and there... only seemed to help against heavy heavy roach or ultra armies... surprise. Zerg ratios differed, see above.

Did attack moves and also did micro on one or both sides. Results were not heavily reliant on micro. Micro helped Ultra + roach + corruptor the most I think. Heavy Banelings pretty much required no Z micro except OL drops. Sometimes I would burrown move roaches closer or further from fight or to allow Ultras to get better positions. Corruption seemed to help if you went corruptor heavy... get the kills and get out quicker with your air.

Played on slowest speed so I could micro both sides at the same time when I wanted.

Upgrades:
All units had their special abilities upgraded (bane speed, blink, colo range, etc.). I did more tests with + 2 range (ground) attack + 1 (ground) armor and +1 air attack on both sides than anything else. More armor would favor Zerg a little bit I believe. Here again banelings have the luxury of being just as effective even with lots of upgrades on both sides.

Finally, imbalance can sometimes be caught in the unit tester. Remember siege tanks before the damage nerf? No army could stand up to them... even mass immortals. Unit tests have their place to find a composition you want. It's up to the players to make it possible.