Mostly this for me too, but I want to add one thing: The help command.
Some of my players (the roleplay heavy ones) often like to do interesting actions during combat rather than just saying "I attack X". For example they think of a neat way to distract the enemy instead of directly attacking it. I consider stuff like this as "Help" action which grants advantage on the next attack done on the distracted enemy at the cost of an action.
So in theory the group could replace 4 normal attack rolls with 2 attack rolls on advantage if they want to. But I fully leave that up to their imagination.
They could, but they'd be very silly to do so, mathematically speaking.
4 attacks, assuming even
remotely equal hit bonuses, has almost exactly the same chance of total failure (0 hits) or near-total failure (1 hit), but can have 3 or even 4 attacks land, when 2-with-advantage cannot. 2-with-advantage is certainly more likely to get at least one crit, possibly two...but, due to static damage bonuses (from Dex or Str, possibly magic weapon) two crits is
still lower than 4-attacks-without-advantage. Only by heavily stacking on two people with substantially superior damage dice (e.g. greatsword, since it's 2d6) can you get a better average result. I suppose if you have multiple Paladins that only drop Smites on a crit, but that smacks of intentionally optimizing for it, rather than a couple of people in the party deciding to do something because it's flavorful, not because it's useful.
[sblock=Math, for those who care]Assuming, say, a 60% hit chance, the expected value of four attacks is 4*.55*(average damage roll+static value)+4*.05*(2*average dam roll+static value), to account for crits. Most people are going to use a d6 or d8 weapon, so we'll go with the latter. And most people should have at least +2 to damage from their base stats (assuming a slightly low value to be conservative: it reduces the impact of the static part). So 4*.55*(4.5+2)+4*.05*(9+2) = 16.5 expected damage per round. For two attacks with advantage, an initial 60% hit chance becomes an 84%, with a 9.75% chance to crit, so the formula becomes 2*.7425*(4.5+2)+2*.0975*(9+2) = 11.7975, about a 30% loss in damage. If we compensate by switching two of the attacks to 2d6 (average 7) and a +3 static bonus, it looks like:
4 Without: 2*.55*(4.5+2)+2*.05*(9+2)+2*.55*(7+3)+2*.05*(14+3) = 20.95
2 With: 2*.7425*(7+3)+2*.0975*(14+3) = 18.165
The gap narrows, but only slightly. Only by having a substantial difference in damage values (e.g. Battlemaster maneuver, Paladin Smite, maybe Barb Rage) can you even get it up to parity. I'll also note that, while the 60% hit rate was chosen more or less at random, the effect is pretty solid unless it's almost impossible to hit the target....which, as I understand it, isn't really
supposed to happen in 5e because of "Bounded Accuracy."[/sblock]
Everyone attacking without advantage is, in the main, superior to half the group attacking with it. You just don't get enough bang for your buck. And since it's only the
next attack roll, rather than the next Attack Action, many classes actually lose multiple attacks to do it (particularly Fighters) while only generating 1 roll's worth of advantage.
It's kind of a neat idea, and it has its place, but surely there are better ways to try to help an ally. Finding some way to even
potentially inflict Disadvantage on enemy attacks, for example; even if it only works 50% of the time, that can be enough to make most high-AC characters effectively unhittable,