MaxPax takes a bunch of my advice. Somewhat disturbingly, as there's no way anyone he knows reads my blog.
https://www.youtube.com/watch?v=NwIdZL4jtG0
Anyway, result: 3-1s Clem. Could definitely have won the first game by taking the rest. Mistakes: built colossus. (Got viking'd, the way they always do.) Built disruptors. (They all died and killed nothing. Often before even getting to shoot.) 0-armoured the carriers. Had plenty of time to at least get 1, probably 2, but no. Did get prism and obs speed, but only after forgetting those two units exist.
Game 2 built colossus but later, as a surprise. Smelled weakness and used them as a killing blow instead of a staple. Turns out that's a fine use for them. Used templar with (dun dun dun!) a warp prism to guard against EMP.
Game 3 all about chrono boosting those upgrades. Turns out prism+templar is really, really good. Boost out storm so it arrives before terran is ready. If Clem hadn't quit Max would have been at 2/0 before Clem started 2/0.
Game 4. I forgot to mention using shades for scouting. Use them like a skirmisher line. Max didn't forget, though. Didn't forget to keep his upgrades ahead of terran either.
Should pre-emptively rebuild his scouting observer. Once you plant an obs in Clem's base it should be treated as dead. Also, like, use sentry hallucinations for that.
Used storm, at first instead of colossus, then to zone out for the colossus. You can use the unit, but it has to be very delicately timed. A quad-colossus push has to wipe out at least one base, if not win the game outright. If you're not sure, then don't build them.
I don't get why micromanaging the prism/templar is feasible but babysitting the disruptors is, apparently, impossible. Regardless, that seems to be the way of it, so don't build disruptors unless it's for some very specific purpose.
The game 1 tempests would probably have been fine as an assassin strike team. Dart in, blow up one viking, retreat to the field base, which is disruptors. If Clem moves in with marines, fire a purification nova. Keep the fight tempest vs. viking. Have an obs for spotting, with shades or maybe hallucination as backup. Spent disruptors should get evac'ed one way or another, either through a prism or just right-clicking the main. Nova has a long cooldown but it's not longer than building a new disruptor.
Exception: if you have ten disruptors, you can have a nova on the field at all times. Fork strategy: they either have to walk into the nova river, or let the tempests/carriers fire with impunity.
This result brought to you by epistemic training. Lift heavy intellectual things. I can also coach hockey. You can only get good at a domain, but the hack is there's a domain-of-domains. The meta-domain.
P.S. Scouting. Instead of planting an obs inside a pro's base, who always scans and kills it, plant a probe outside the base to watch for the move-out, then siege an obs in the middle of the map to see which assault path they take. Protoss should be even more difficult to ambush than zerg. Use the obs like building your own watchtower. If you want to see inside their base, use the garrison sentry for hallucination. Treat your base sentry like a building which trains hallucinations. Maybe even intentionally seal it in so you can't accidentally f2 it.
If you have a DT shrine, put a DT on hold position at their next expansion. Either go for the cancel, or let them land it and hit the workers when they transfer - don't block the building. Putting a burrowed zergling in the way is cheeky rather than good. E.g. dart in with the flux'd void rays to hit the cancel. This is when you siege an obs inside enemy lines: when it's ensuring the voids can flee in time. As zerg, wait for the transfer then hit it with your mutalisk wolf pack, zergling runby, or baneling bombing run. If the burrowed zergling is offset, then they don't suspect and you can repeat the process.
Out of the blue question, have you ever tried Europa Universalis IV? Not really RTS so much as Grand Strategy, but solid 9/10 game.
ReplyDeleteGrand strategy games are better than RTSs. However, no I've never felt drawn that series in particular. It does look like a 9/10er.
ReplyDeleteYou can't run a few numbers and casually gin up a broken plan in games like EU. They have too much depth for that.
Though come to think Starcraft has one similar feature, which is how individual engagements work. This means the most important feature of your army is the ability to manoeuvre. (Which colossus lack.) Slightly changing the angle of your attack can drastically change how the fight plays out, and predicting in advance what changes and how much is wildly impractical.
I tried getting into starcraft, it just wasn't for me. The only RTS I've really had fun with were the Total War series. Though Manor Lords is looking good, and the fact that it was made by just one dude and it's already putting the major studios to shame is hilarious.
ReplyDeleteStarcraft is badly designed IMO. Usually at least 10% and often upwards of 30% of each match is absolutely identical. Starting a game of SC is boring as fuck and, especially on the ladder, after the boring part it will suddenly end. "Oh whoops, I should have done a different boring thing than the boring thing I in fact did."
ReplyDeleteThere are certain benefits to drawing out this starting phase, but none of them are worth the while. Like yeah you can do all these cheeses or stack up a ton of minor build-order efficiencies, at the cost of making any normal game an exercise in tedium. There are infinite tiny build order variations, but you can't adjust on the fly or choose them reactively. It all has to be tediously practiced in advance.
Oh, bonus round: if the game goes long enough that the setup phase is an acceptably low proportion of the total playtime, it gets into another degenerate state. Security is affordable even in SC2, so neither side wants to attack. (That's what "balanced" means - you can always successfully deter.) They just mine until they run out, then attack out of lack of anything better to do. Two-player EVE online.
That and it's NOT an RTS. It's a real-time tactics game. "Skill" at the game is almost totally dominated by speed, not chosen strategies. That and engagement luck, as engagement happens too fast for humans to meaningfully react.
Being real-time basically makes it stupid, which is why Mechabellum is a drastically superior design.
ReplyDeleteIf real-time is the focus, it should have no mining, basically. Let them both requisition a fortress and an army, then they clash, game over. Going from 6 to 12 starting workers was a good adjustment, but it need more like 90 starting workers and 20,000 starting minerals. Or just allocate presets before the game starts.
company of heroes 1 has no mining and most units come as a set and it's fantastic.
ReplyDeletehttps://www.youtube.com/@MathiasCzR01/videos
coincidentally no one plays coh1.
aoe2 is somewhat interesting to watch, though this guy's way of commenting definitely helps a lot. game is slow enough that adjustments can actually be made.
https://www.youtube.com/@T90Official/videos