Jehu, maybe your Scythe Chariots were forced (by terrain or buildings) to queue up on approach... This would make the missile units' life so much easier!
Overall, back to the main topic, I also think that Composites are overrated.
My reasons:
I've lost more games to CA than I can remember - immediately!
The games I lost vs. Composites were by attrition - indeed I managed to run away and survive many.
When I'm the one to use Composites, either I am not much harassed before I can get them (well, a bit, but without compromising their production, which, in turn, turn the tide) or I simply fail to get them in time.
Or the position is not good enough to wall properly...
I have seen that you do need a combination of factors that helps you with getting them before you are ... er, crippled.
This means:
Out4blood, Ender, Methos, my cousin and all the fast-wise-guys out there can win with Compies. For slow-me they are just another nail in my (announced) coffin.
My idea is that Compies are ONLY viable for people that have not YET been crippled by an attack when they get them.
The reasons could depend on own player (he's faster, period), on the map (favourable to boom / "power-up"), or on the opponent (did not bother you enough in due time).
But, clearly, you MUST be in reasonable economic and demographic shape to get them.
Now, about getting the enemy.
In a 1x1, I think that CA are more popular not only because they are adept at killing peons, but also because they have two other fine functions, that people do enjoy, for they kind of dispense with forward building and the preliminary (and hard to manage) peon scouting:
CA are great Scouts, able to defend themselves.
CA are as fast as Cav, so, location of Archery Range not that critical.
If you are capitalising on Wood mainly and are really short of gold, Ca look also more attractive.
Anyway, the reasons to prefer CA to Composites look to me as credible enough.
The reasons to opt for Composites do require, IMO, a combination of factors that is hard to achieve for many players (some speed, opponents not that fast, favourable map, etc.).
When those conditions are met, I think that Metho's idea is perhaps the most adequate.
But he, I know my limits.
Your_Slow_Friend
(hehe)
[This message has been edited by Your_Old_Friend (edited 12-16-98).]