We have formulas for a reason, I suppose. And few series have been as formulaic as Command & Conquer. But a formula is sensitive. Change the right things and you get 2006’s Command & Conquer 3, one of the best real time strategy games of the year. Change the wrong things and you end up with Red Alert 3. It has the same lack of subtlety and rapid pace—plus it looks better. But it comes off as an average and unnecessary entry in the series.

On the plus side, the production values and art direction are first rate. The Red Alert games are cotton candy color compared to the darkness of the Tiberium series, and the colored accents on the Soviets contrast nicely with the Voltron themed Rising Sun, the new Japanese nation. EA LA has perfected unit design; given how easy it would be for a developer to simply throw in a bunch of tanks, they have give each soldier and vehicle a distinct role and make each immediately recognizable in a mass.

You can even tell at a glance which units are in their “alternate” modes, and many units have alternate modes. Japanese anti-infantry mechs can turn into jet planes, every tank seems to have an option to use a special ray to immobilize enemies, and Soviet conscripts can carry Molotov cocktails to clear buildings. It is impossible to quickly select your units based on their current status; a double click, for example, would grab every conscript whether they are wielding guns or flaming gas. This leads to some fancy footwork with grouping, but you could manage it if things didn’t move so fast.

Much has been made of Red Alert 3’s commitment to naval power, and it’s not a bad compromise. Some of it is clearly a cheat – so many units are amphibious, including the Allied heavy destroyer, that the sea becomes merely an extension of the land. This means less micromanaging of transports, though the Soviet tanks need to be airlifted, but it also weakens the distinction between sea and land power on some maps.

But the unit design aside, the game is an exercise in frustration in skirmish mode. Units are overly fussy and overly fragile. This is largely due to some miscalculations about normal reaction times. If there is a single design rule for real time strategy games, it is that once you get a warning you should have options, especially in the opening stages. If I send four or five infantry out to scout, then my first notice that there are, in fact, enemies should not be “Units Under Attack” followed by jumping to four or five corpses. Throw in the overpowered commandoes, each of whom can destroy a series of buildings or vehicles before you even know they are there and you have frustration piled on frustration.