I rarely get the opportunity to play with pirigosu-go, but whenever he is online Klaus is playing with him. Pirigosu-go happens to be a big fan of primeval isles, an island map which is 256x256, the largest tileset with probably 100 islands filled with 8 mineral patches a piece and sometimes a gas geyser. Pirigosu-go is always trying to get FFA'S up and running but the only ones who bother to join are Klaus and I.
From what I've seen of the map from a strategical point of view I think a good unit composition is to go mass air. Land drops will take too long and if your opponent happens to spawn all the way on the other map the other player will have already accumulated 24 units worth of air to stop any possible land drops. Mobility is also important so any capital battleships like battlecruisers or carriers aren't that great and will end up in this hopeless goose chase trying to catch more faster air units that pick off your bases while avoiding the main fleet. You're so far stretched out on this map.
My first impressions of primeval isles was this 4 hour long replay pirigosu-go showed me that he played back in the day with his Terran friend.. it was really epic....and long. People got a billion resources in the bank, maxed out air fleets. Terran went valkyrie, wraith, science vessel, and Battlecruiser. Protoss went arbiter, carrier, scouts, and corsairs. The armies would constantly trade and whoever had the losing army would simply make a unit composition counter to what they had just lost their army to. It's no biggie to lose an army because you have enough resources to rebuild your fleet dozens of times over.
The soft unit counters seem pretty evident and it's hard to get a mix of all the units. Valkyries kill off corsairs and scouts quickly but are really bad against carriers because of the carriers' high armor and the valkyrie's spread out splash damage. Battlecruisers are good against corsairs and carriers but suck vs. scouts. Wraiths are really good vs. carriers and scouts but horrid against corsairs. I remember seeing the Terran transition from valkyrie/battlecruiser because he knew protoss neglected to bring an observer. If Protoss went a corsair centric army, Terran would go battlecruisers mostly with some valkyries. Then Protoss would rebuild their fleet with Scouts. Terran goes valkyries, and the cycle continues. The Terran ended up losing the game.
The replay shown was my first impression of the unit interactions and unit compositions but the more I play the map I feel like omitting capital ships is the better choice in exchange for mobility.
On that same map I played an FFA game with pirigosu-go and Klaus, and a separate 1v1 game with just Klaus. In the 1v1 game I was Terran and Klaus was Protoss. Klaus went DT rush. Luckily for me after going for double CC expand I made engie bay+academy for comsat scanner and a turret ring. i was able to hold it off. It's really strange though that he got a close spawn position on me...
In the FFA game Klaus went for reaver rush. It was so strange because he did no scouting and just went clockwise and the player there was me instead of the 5 other computers pirigosu-go put into the game because no one else bothered to join. I mean, really, close spawning positions on a 256x256 with 8 different places? It was just ridiculous. I made a stargate first. I don't think there's a reason to get stargate so quickly, may as well go for robo facility for shuttle so you can island expand fast. I didn't scout with my corsair though, I was having too much fun killing off a zerg computer's overlords...
Other than Primeval Isles, Pirigosu-go also plays a lot of a map called 999999. It's a normal sized 128x128 map where you get infinity resources. It feels overwhelming and there's a lot of game theory and opportunity that I'm not used to begin thinking about. The only thing about the map that limits you is how fast you can spend your resources. It's overwhelming at first like everything in your mind is on overload but when it comes down to it you have to be selective about what you're committing your apm and attention to because you can't do it all at once.
A lot of my careless mistakes cost me a game even if I had better multitasking and better mechanics just because my opponent, pirigosu-go, who was rusty had better decision making which is something I'm not used to. In a melee game you can make whatever, doesn't matter with good mechanics and wise resource spending against a much worse opponent, you're just going to have more stuff and roll your opponent over. Resource management doesn't matter, and pirigosu did things, smart tactics like send a probe to my base, make a bunch of pylons and TONS OF CANNONS. Nothing limits the number of cannons which is scary to think about. It's like this self sustaining virus in your base. Sometimes he'd make random nexuses all over my base to mess up my building layout.
Something I've realized about the map is building placement is crucial. I kept on macroing slower than my opponents and their opening moves like mass cannon rush or nexus blocking messed up my stuff and I didn't pay much attention to making sure my buildings were neatly placed. Instead, I placed them anywhere because I didn't want to fall behind my opponent macro wise. All of these things threw me off and pirigosu-go would make air units which I've learned really are the best option in PvP infinite money. There's also a ridge that surrounds your base. You can land a probe there and set up cannons. it becomes super annoying and that's what pirigosu did exactly.
Against Klaus, who was by far the more sounder mechanically was a tough match. I kept on losing a lot but after a few games I learned from my mistakes. It was a UMS map and all of the locations are designated, not randomized. We played so many zvp games and I always lost them but had I realized the spawning locations were always the same I could've done a 4 probe cannon rush. A lot of zerg's weakness like Terrran is that it's limited by individual drones to make hatcheries so there's a huge timing window to cannon rush before the first three drones finish their hatcheries, and even then 10 lings won't do much against 20 cannons. My cannon rush was weaker because I only was able to get out one probe if I was lucky and sometimes I wouldn't even scout the zerg in the right location because I sent my probes anywhere.
Zerg's production method when it gets up and running is impossible to deal with. You can select 12 larvae, make 12 Ultralisks, make 100 more ultralisks, it's too much. With Protoss it gets more repetitive and it's all about cost efficiency with toss units. Queuing gateways, and time consuming things like energy units, morphing archons are huge hindrances. Eventually zerg just mass produces stuff and it doesn't even matter. The Zerg just has more stuff and so, so many ultras. If you try to go mass reaver, Zerg can go mass Guardian/Devourer which is surprisingly really powerful. protoss stuff takes soooo long to produce.
Eventually when Klaus did tell me spawning locations were always the same I did beat him with my 4 probe cannon rush vs. his zerg. There was notihng he could do. He attempted to leave 4 drones unmorphed to pick off the probes but I kept running the probes away to make pylons, cannons, minimizing hits overall.
We played some zvz games afterwards and I started catching onto the nature of the game. It wasn't like BGH where mineral mining is normal but minerals are virtually undepletable because in BGH you have to do a one solid knockout punch, destroy your opponent's infrastructure and what not because your opponents are gonna have a huge bank. You gotta keep pulling in the pressure and it works great if you have more bases to reinforce. In the 999999 money map it's like you have to do a bunch of combo knockout punches. If you happen to create one sort of hole in their defense or make them lose their footing, you have to slowly widen that vulnerability. It'll take probably 15 minutes of constant pressure but when it's all said and done there'll be no breathing space. There's a lot of suicide bombing to kill off key infrastructure, and it's fine because units are expendable, more so than in fastest or BGH in a way.
PvP is a huge stalemate, not much room to breach either player's defense. I tried going mass ground units vs. my opponent. I went for a counterplay against Klaus' cannon rush which I successfully defended by making pylons in position to make cannons for later. Then I mass cannoned his choke because he kept on doing a nexus block to prevent the same thing from happening to him. He went scouts though and I had a ton of zealots. Things got iffy but I managed to recover. I had the entire map, had all the space to build infrastructure but I could not break him vice versa. I had this idea of using arbiters to have an endless cycle of stasising his carriers but scouts are so resilient and again, suicide bombing units is too good.
ZvZ is different. Something that blew me away was that as soon as your spawning pool is finished, you start your ling speed. As soon as each of your hatcheries finishes, you can morph them into Lairs. It doesn't come to mind at first but it was so....strange. Woah. Mutalisks seem good. Lurker stalemates move the battles to the air. Klaus won 3 games of ZvZ. The third game he won he kept doing ling suicide bombs to pick off my spire. It seemed like ZvZ was a muta vs. muta battle and whoever got to mutas first was gonna win. I lost cuz my spire was sooo late. You'd do some ling pressure, squeeze drones in to make more hatcheries, and make a spire. My lings kept getting out of position and I realized I could make a defensive wall around my spire and afford to let my hatcheries take some hits from the lings. I was also a lot attentive to his suicide bombing and positioned units accordingly.
The fourth game was a different story I thought that I could make only mass lings and try and hit a timing window where his Mutalisks would still be producing and I'd be at home making more hatcheries. He'd be busy using his mutalisks to defend rather than go on the offensive and I could possibly snipe his buildings off. Unfortunately, that was not the case. The rush distance is pretty large and when you have infinity money mutalisks are such a good bang for your buck unit. 100/100 for each larvae is better than 50 minerals. I made a Hydra den out of panic. My Hydralisks did manage to defend somehow miraculously. Then I kept streaming on the pressure with hydras, just a constant flood. Klaus teched to Guardians and Ultras. The ultras however got eaten alive and the Guardians were a huge time investment using up supply and somewhat ineffective against my hydras that defended at home. I got the victory!
Something I learned that day from Klaus was that when you set up your hatcheries, you don't want them as close together as possible or else it becomes incredibly hard to select larvae and just get 12 of them selected. I encountered this problem in the first ZvZ game between Klaus and I that went on for 15+ minutes. It's better to leave Hatcheries side by side and leave another row of hatcheries with one hex of space between the two rows. Klaus was right and I saw a tremendous improvement.
That was my night of starcraft. Thanks for reading.