Game design is hard, much harder than it looks. Treading new ground is really hard, there are pitfalls everywhere and only with skill and experience can you spot them before setting foot in one. Perimeter 2 feels like another practice lesson in the training of a new game designer and he is still further from his goal than he thinks.
The game has learned some lessons compared to its predecessor: Mechanics like the recipe morphing of units (micro intensive and the promised versatility of the system never really appeared with the low population caps allowing no surplus components in your squads) and the turning-the-world-into-a-parking-lot terraforming (terrain? What terrain?) went out of the window but many new issues were introduced.
Since Perimeter 1 wasn't exactly a massive hit I'll describe it quickly: You had a Frame (a kind of movable but usually stationary city and base) from which you can build an energy network out of generators, you build chains of them to reach areas on the map, they generate energy (the one resource in the game that pays for everything) as they cover area and other buildings must be built in the covered area. If your chains get broken you lose control over the unlinked part of the network and the first player to connect to it gets to keep it.
The energy network concept got bastardized badly in the sequel. Instead of a connected network you can just plonk a generator (called geocore or aquacore, depending on your faction) down anywhere and it'll function fine. In fact they start out as mobile units (like C&C MCVs) that then deploy into the respective cores so you have no reason to build near your existing base. Without a central structure to destroy this system makes bases an amorphous structure comparable to cancer, there is no weak point and every part of the base must be torched to prevent it from regrowing.
Even worse, instead of having the covered area dictate your energy income every core produces energy just by existing. You can throw up large clusters of the things and they'll happily provide you with gigantic amounts of energy that make all costs in the game trivial. There is no restriction to the number of cores you can build simultaneously and they cost almost nothing anyway. That means it doesn't matter what everything costs.
The only cost you'll feel at all is buildable area (and your own ability to order new buildings fast enough). One big advertised feature is the land vs water concept for base construction, one faction builds on land, the other in the water and their cores will terraform the nearby area to make it suitable for their bases. This sounds fine in theory but in practice it's a glaring imbalance, since water flows it will quickly cover a wide area for the Harkback (the water faction) to build in while the Exodus (earth) can only build on perfectly flat ground that usually only happens near their cores. VERY near. You'll have a hard time fitting larger structures like factories or the research lab anywhere in an Exodus base. That would be fine if the Harkback didn't get so much space so quickly that they can make clusters of factories on par with their clusters of generators and utterly crush any Exodus base with a gigantic army (factories also double as population cap increasers, one squad per factory) seconds into the game. Even if they don't their experimental super unit is a massively powerful vehicle that fires a beam that instantly vaporizes any target and sweeps across a base, firing constantly while the Exodus counterpart is a cloaking device that doesn't cloak itself and serves no apparent purpose (I think it can keep units cloaked when attacking so they are untargettable and practically invulnerable while it is around but that seems buggy, the AI units will shoot at cloaked units). Reaching the superunits takes almost no buildtime and only costs a lot of energy (which you have so much of that it doesn't matter). Maybe they didn't have testers that were smart enough to figure out that a cluster of generators + rapid teching means you can have these superunits faster than medium regular units...
Speaking of regular units, the design there is another case of sounds-good-on-paper: Every unit can morph instantly between driving on the ground and flying around as a plane. Units and turrets can only target either ground or air units so you can use that to dodge enemy fire. The problem with this is that antiair turrets and units are a tier 2 technology. At T1 your only way of fighting airborne units is airborne units so whoever brings more to the fight wins. While T1 air units are unable to target ground units (only higher tech air can) and that would theoretically be a weakness everything, including pure antiair, can target buildings so these flying units can happily demolish a base without giving a damn about the ground units. Even with antiair in place the rockets fired by that are so retarded in their guidance that they have a hard time shooting air units down. To top it off there is only one unit per tech level (three levels in total) with air and ground forms each. The first two behave pretty much the same in both modes (except preferring different targets) so the morphing won't provide a difference there, only the third level regular units do more than just move faster and over everything when set to fly, they turn into bombers (if you've played Maelstrom, think of the heavy tank of the Ascention, they're exactly like that).
Another issue are the off-map support powers. You get those by digging up crystals with an excavation tool that acts pretty much like a bucket (you can theoretically dig trenches and build barricades with that but by the time you finish one the enemy has engraved his initials into the burning debries that once made up your base). The mechanic kinda works in the campaign (you level your powers up and stuff) but it's tedious and in multiplayer it's plain stupid as there are no tiers for the abilities, you can dig up one crystal and already fire a nuclear missile or a meteor shower at the enemy (yeah it takes 10000 energy, as I said energy is so plentiful you can afford that easily). I've finished a campaign mission by simply nuking the enemy base from orbit about 30 seconds into the game.
Speaking of the campaign, that part is actually decent and I suspect it's the only part the developer actually focussed on. Like Perimeter 1 the enemy AI is active during the campaign, it won't just idly sit there while you build up your base and army or attack at predefined times, it will expand and attack quickly, tech up, rebuild lost buildings ASAP and generally be a total pain in the rear. It's not particularly smart (it won't try to pull the same core-cluster-factory-cluster approach a human would) but it does recognize when there's no antiair and set its units to fly (or avoid that if there's more antiair than ground defense) which effectively means you'll have 1-2 light flying squads in your base in the first minute of any mission. It's counterable (by spamming more air) but it may be a shock for people who aren't used to playing RTSes fast and would expect some rush free time to go base building. Overall you're better off by attacking early instead of being defensive and letting the AI build up, another general RTS lesson. If the singleplayer mode of an RTS is meant to be a preparation for multiplayer then Perimeter 2's campaign does a good job there. Too bad multiplayer is so incredibly imbalanced that you really won't want to play it. Even mirror matches aren't necessarily fair with maps being apparently laid out with zero thought put into the terrain's effect on the base building so different start points can make or break your game.
As a topping on the mess that is multiplayer the game doesn't even have a server browser for online play! It's back to entering IP addresses directly. Considering the game is sold on Steam only from what I can tell it really shouldn't have been impossible to integrate Steam's server browser into the game.
So should you buy it? No. Really, just no. It's an interesting exercise in game design that I hope taught the designer a lot about theory and practice but it's not a good game by any means. It might be worth it if you enjoy seeing just how badly broken a game can be or just want to see the silliness in action (I knew what I was getting into, I didn't think any of the previous games of the developer were real quality anyway, just interesting for being unconventional). Well, on the forums there's already a list of complaints compiled by the users and the developer promised a patch but it'd probably still be a fairly uninteresting game even with all those fixes added.
The game has learned some lessons compared to its predecessor: Mechanics like the recipe morphing of units (micro intensive and the promised versatility of the system never really appeared with the low population caps allowing no surplus components in your squads) and the turning-the-world-into-a-parking-lot terraforming (terrain? What terrain?) went out of the window but many new issues were introduced.
Since Perimeter 1 wasn't exactly a massive hit I'll describe it quickly: You had a Frame (a kind of movable but usually stationary city and base) from which you can build an energy network out of generators, you build chains of them to reach areas on the map, they generate energy (the one resource in the game that pays for everything) as they cover area and other buildings must be built in the covered area. If your chains get broken you lose control over the unlinked part of the network and the first player to connect to it gets to keep it.
The energy network concept got bastardized badly in the sequel. Instead of a connected network you can just plonk a generator (called geocore or aquacore, depending on your faction) down anywhere and it'll function fine. In fact they start out as mobile units (like C&C MCVs) that then deploy into the respective cores so you have no reason to build near your existing base. Without a central structure to destroy this system makes bases an amorphous structure comparable to cancer, there is no weak point and every part of the base must be torched to prevent it from regrowing.
Even worse, instead of having the covered area dictate your energy income every core produces energy just by existing. You can throw up large clusters of the things and they'll happily provide you with gigantic amounts of energy that make all costs in the game trivial. There is no restriction to the number of cores you can build simultaneously and they cost almost nothing anyway. That means it doesn't matter what everything costs.
The only cost you'll feel at all is buildable area (and your own ability to order new buildings fast enough). One big advertised feature is the land vs water concept for base construction, one faction builds on land, the other in the water and their cores will terraform the nearby area to make it suitable for their bases. This sounds fine in theory but in practice it's a glaring imbalance, since water flows it will quickly cover a wide area for the Harkback (the water faction) to build in while the Exodus (earth) can only build on perfectly flat ground that usually only happens near their cores. VERY near. You'll have a hard time fitting larger structures like factories or the research lab anywhere in an Exodus base. That would be fine if the Harkback didn't get so much space so quickly that they can make clusters of factories on par with their clusters of generators and utterly crush any Exodus base with a gigantic army (factories also double as population cap increasers, one squad per factory) seconds into the game. Even if they don't their experimental super unit is a massively powerful vehicle that fires a beam that instantly vaporizes any target and sweeps across a base, firing constantly while the Exodus counterpart is a cloaking device that doesn't cloak itself and serves no apparent purpose (I think it can keep units cloaked when attacking so they are untargettable and practically invulnerable while it is around but that seems buggy, the AI units will shoot at cloaked units). Reaching the superunits takes almost no buildtime and only costs a lot of energy (which you have so much of that it doesn't matter). Maybe they didn't have testers that were smart enough to figure out that a cluster of generators + rapid teching means you can have these superunits faster than medium regular units...
Speaking of regular units, the design there is another case of sounds-good-on-paper: Every unit can morph instantly between driving on the ground and flying around as a plane. Units and turrets can only target either ground or air units so you can use that to dodge enemy fire. The problem with this is that antiair turrets and units are a tier 2 technology. At T1 your only way of fighting airborne units is airborne units so whoever brings more to the fight wins. While T1 air units are unable to target ground units (only higher tech air can) and that would theoretically be a weakness everything, including pure antiair, can target buildings so these flying units can happily demolish a base without giving a damn about the ground units. Even with antiair in place the rockets fired by that are so retarded in their guidance that they have a hard time shooting air units down. To top it off there is only one unit per tech level (three levels in total) with air and ground forms each. The first two behave pretty much the same in both modes (except preferring different targets) so the morphing won't provide a difference there, only the third level regular units do more than just move faster and over everything when set to fly, they turn into bombers (if you've played Maelstrom, think of the heavy tank of the Ascention, they're exactly like that).
Another issue are the off-map support powers. You get those by digging up crystals with an excavation tool that acts pretty much like a bucket (you can theoretically dig trenches and build barricades with that but by the time you finish one the enemy has engraved his initials into the burning debries that once made up your base). The mechanic kinda works in the campaign (you level your powers up and stuff) but it's tedious and in multiplayer it's plain stupid as there are no tiers for the abilities, you can dig up one crystal and already fire a nuclear missile or a meteor shower at the enemy (yeah it takes 10000 energy, as I said energy is so plentiful you can afford that easily). I've finished a campaign mission by simply nuking the enemy base from orbit about 30 seconds into the game.
Speaking of the campaign, that part is actually decent and I suspect it's the only part the developer actually focussed on. Like Perimeter 1 the enemy AI is active during the campaign, it won't just idly sit there while you build up your base and army or attack at predefined times, it will expand and attack quickly, tech up, rebuild lost buildings ASAP and generally be a total pain in the rear. It's not particularly smart (it won't try to pull the same core-cluster-factory-cluster approach a human would) but it does recognize when there's no antiair and set its units to fly (or avoid that if there's more antiair than ground defense) which effectively means you'll have 1-2 light flying squads in your base in the first minute of any mission. It's counterable (by spamming more air) but it may be a shock for people who aren't used to playing RTSes fast and would expect some rush free time to go base building. Overall you're better off by attacking early instead of being defensive and letting the AI build up, another general RTS lesson. If the singleplayer mode of an RTS is meant to be a preparation for multiplayer then Perimeter 2's campaign does a good job there. Too bad multiplayer is so incredibly imbalanced that you really won't want to play it. Even mirror matches aren't necessarily fair with maps being apparently laid out with zero thought put into the terrain's effect on the base building so different start points can make or break your game.
As a topping on the mess that is multiplayer the game doesn't even have a server browser for online play! It's back to entering IP addresses directly. Considering the game is sold on Steam only from what I can tell it really shouldn't have been impossible to integrate Steam's server browser into the game.
So should you buy it? No. Really, just no. It's an interesting exercise in game design that I hope taught the designer a lot about theory and practice but it's not a good game by any means. It might be worth it if you enjoy seeing just how badly broken a game can be or just want to see the silliness in action (I knew what I was getting into, I didn't think any of the previous games of the developer were real quality anyway, just interesting for being unconventional). Well, on the forums there's already a list of complaints compiled by the users and the developer promised a patch but it'd probably still be a fairly uninteresting game even with all those fixes added.