OP is a question I've been pondering for a while, particularly the difference between opmtimal 1v1 and team play. There's already been some great responses, and I'm far from good at teams, but I've noticed some things :)
#1 - composition
More players expanding faster + more noobs=more porc. Lot's of opposing porc=dunt make so many raider. While raiders still have a very important role in teams, they're not the default like in 1v1. Up-scaling your composition to keep it relevant is beyond important. If someone's using monospam against a player in the late game they're just delivering metal.
#2 - noobs in the meta
Good things to take into account is which noobs porc, which noobs over eco, and which noobs know how to composition. This is far more important than it may seem if you're serious about getting elons.
- Leave porcers alone and go smash his team-mate. Outnumbering people is fun.
- For over ecoers, you can basically use a small raid to effectively kill them, then immediately transition into aggression against their communism weakened allies.
- If someone can't composition, bring your cons along because you're gonne feed on them all through early to midgame, then use your inflated army to defend while you manufacture the silver bullet for their stronger players.
- Also, if someone doesn't know how to end-game, you can get a DDM up and then leave lane to gank :))))
#3 - map awareness and punishing over extension
A lot of noobers get scrabbly when I push them out of the center starting position. Essentially, I like a mobile playstyle which is conducive to assisting allies. Many games can be won by assisting your ally to defend against attacks that are overextended, or by doubling up on hammering someone's risky forward position. Noobs favour slow units that are strong locally but weak at global impact, so naturally you want them to take a 'lane' to the side and just push.
I think that team games aren't even close to their final meta as far as ganking is concerned. I see early game combined commander rushing being ridiculously good if any clan manages the coordination. Early doubling/tripling etc. defines the metagame for so many team RTS formats, and the mechanics in ZK seem to lend themselves to it even more (since reclaim, since map chokes are rare, and since you're not required to take workers off resources in order to forward build turrets).
#4 - silver bullet
Not to stoke the sfire, but he's probably one of the best at this role. When he's done pissing about trying to get headhunter, he will contribute nothing for ages then win the game by ramming his win con down your throat. Pluks should all get a mention here.
#5 - late game macro and board control
That thing where you never stop growing. It's really important. *Unlike some other games, ZK doesn't have a population limit. This means the influence of strong players can continue to grow in a game unfettered by anything but communism.
#6 - recognising opposing team weakness at a meta level.
Be 3v3. You're against one godly good player and two nooobers, while your team consists of two moderately good players and one noober. Their one godly good player can't start both air and ground: if you put a moderately good player on both you're guaranteed to win whichever one he doesn't go for. If he's air he can't expand, if he'd ground air is uncontested and he will be forced to fend off ground forces and air forces simultaneously. Pick on him, kill his units over other units. Wall off his team mates with porc and just make sure he doesn't make it to late game. Knowing how to capitalise on the weaknesses of a team with high skill discrepancy is elo farm.
#7 - leadership
Having your team do what they're good at, or fostering a teamwork environment, or even just calling the shots. I tend to go hands off in this respect because it's not a style that interests me - but it's definitely legit.
#8 - Clear win conditions
Sometimes your team is losing and there's not many ways to win the game. Recognising what will win the game in this scenario turns losses into wins. Good examples are stuff like 'slinging' off weaker players by all-inning their base/army/forward and then being prepared to hoover it up immediately. The risk of committing to both a huge attack and the workers to boost your momentum is big, but if you don't have a chance otherwise then you've got to recognise it as an option that is more legitimate than the ordinarily superior "safe" plays. It'll look stupid if it fails, but you were going to lose anyway. Another one I see is recognising that if you're at sea and your land team is failing, you need to not only win sea, but win it with something that can then immediately crush land afterwards. It's no good taking the sea with mass typhoon if your entire army is rendered redundant immediately and you then have to build up in order to affect the game once more. Switch to amph or hover and you can usually backstab your land opponent before they're prepared for you while they're busy trashing your allies, you are then set to win the game since
their army won't be able to touch your sea base.
#9 - attrition
It's true, attrition is more important in teams if you're expected to carry. It's my greatest failing as a teams player, I just don't like having my units for more than a few minutes... Army's metal worth is essentially multiplied by skill, it's better for lower elo players to sacrifice their units though this is an ugly reality for many (and it's not fun to always take the hit). It's also better for the team if reclaim is ceded to the better players as well (though only if it's convenient to do so). A 2k elo glaive is better than a 1.4k elo glaive. Predicting your allies excess and being prepared to put it to good use is also handy.