I have a number of questions for consideration...
The first, question, of course, if whether you want this to be at all practical. It's almost certainly just a thought experiment anyway, so is it a thought experiment about What Would Be Really Cool For Me Personally, or is it a thought experiment on How To Make The World Better By Giving It The Perfect Mecha Game?
The second question is what sort of a community you want to have. Clearly, The Best Mecha Game Ever would *have* to have multiplayer. Preferably, it would have an option for large numbers, to allow for entire battlefields of mecha going after one another in realtime. Taking *that* to the Coolest Possible Level, in turn, would require some reasonably sophisticated tools of communication between allies, to let people who are good at strategy and cooperation and building complementary mecha to really shine. This all means that somewhere, somehow, you're *going* to sprout a community, if you're at all successful, and that, in turn, means that a large portion of the gaming experience is going to depend directly on who exactly is signing up over the internet to play with you, and what attitude they're bringing to the entire thing. Fundamentally, the community is shaped by the game, and the community shapes the game experience in turn. What sorts of gamers are you rgoing to encourage? What sorts of attitudes are you going to encourage? Who is this game *for*, really?
The third questions exactly how deep in the weeds you want to go. It was initially the second question, but you really need to answer the previous question first before you get here. How much numbercrunching do you want people to have to do to build/optimize their mecha? Do you want it to be the sort of thing that can be done simply over the course of five to ten minutes, the sort of thing that takes half an hour and a book/design program, or the sort of thing where "I design mecha in my head" is a potentially awe-inspiring boast? Computers *allow* for a great deal of complexity. Given that, do you want it? If you make it so that an endgame mecha takes five to ten minutes to put together, max, and winds up all chibi and cute with a couple of equally cute superpowers and your choice of a primary fire and an alt fire weapon, then you'll get a fair number of casual players, but most of the more serious players will lose interest pretty quick. (Not all of them. Everyone needs a break sometimes - but most of them.) If you make it the sort of thing that takes at least half an hour of serious thought and a reasonable head for numbers/strategy to make moderately competitive, then you will keep a lot more of the more serious players, and some of them will happily burn hours in front of your design software/books, logging in primarily to test and refine their designs. On the other hand, getting the casual players in at that point will require that you make the learning curve nice and gradual, particularly at the beginning, and most of them won't do well in competitive play even if they do make it through the single-player missions. If you make it the sort of thing that takes three or so hours of strenuous numbercrunching and an encyclopedic knowledge of the rules and available gear - well, you'll be losing a fair number of even the "serious" gamers, and your community will turn somewhat insular. The sorts of people who actually enjoy stuff like that over the long run tend to be moderately obsessive introverts with the feeling that achievement requires suffering, and the matching feeling that if something must be suffered for to be had, that's an indicator that it's worthwhile. They tend to make very specific sorts of social structures, and are almost fanatically loyal to the games they devote themselves to. Now, of course, it's not just one dimension. There are all sorts of tweaks - from optional extra complexity that people can open up (there's never any reason why you'd have to muck with the gear ratios on your tank treads, but there are some nifty tricks that you can run with if you *do* - particularly if you happen to know what sort of terrain you're likely to be fighting on in advance) to starting slow (everyone starts with the same chassis, and it's going to be a while before you can afford to swap out your body, legs, or generator. Look! your life just got simpler.)
Oh - and since you're building the cosmology from scratch, I'd suggest some excuse for fitting in a mecha-based AI/helper-person, with a moderate range of choices in voice. At its simplest, this would be a simple AI with a choice of six voices, and a set of swappable modules that have moderate in-game effects (you have three AI slots, slotting this module will improve your accuracy with missiles, etc.) At its most complex, this would involve a variety of different AI/copilot options, with different modifiers and effects, that developed over time as they gained experience alongside you (If you go with the structure of Living Crystal that you dug out of that crashed alien ship, you have to avoid vibration-based weapons, but your ECM and ECCM go way up. If you managed to save the Old Scientist's Niece, she'll tune up your engines and help with repairs, along with whatever else she does. Also, she's cute. And so on). This is partially because many of the people who would wind up playing this game are male and at least somewhat low on affection, and the option to have a friendly, identifiably female voice running around with them on the battlefield would noticeably improve the game for them (and give the choice to eject a lot more emotional impact - unless they'd paid for the appropriate extra modules). This is partially because almost everybody feels better and gets more emotionally engaged when they have someone rooting for them and helping them specifically. This is partially because just including a persistent, user-selectable voice gives all sorts of ways for people to feel cooler. The guy who chooses the Boris Karloff voice is going to have a noticeably different experience than the one who chooses the Blood For The Blood God voice, which, in turn, will have a noticeably different experience than the guy who picked the Sultry voice (and to be fair, if we're going to have the Sultry Woman voice, we should have a Sexy Man voice as well) but they're all going to have a *better* experience, because it'll help them get into the feel of the character that they want.
And while we're talking about cosmology... this one doesn't matter but so much, but... Normal, straight-up hardtech mecha? Mecha with a few bits of bizarre and poorly understood alien tech thrown in for spice? Mecha and magic? Biomecha?
For that matter, you want to consider what you want to encourage. Just like the dogfighting vs Age Of Sail discussions over in fenspace, the shape of the battlefield and the overall gaming experience will be hugely affected by What Works. If you can pick up a few levels and then start clearing the field of mecha twice your size and eight times your cost by blowing their heads off at extreme range, then the entire game turns rather quickly into an elaborate stealth/detect/sniper situation (shaped further by how much stealth is actually available). If The Things That Work are all within relatively short-range engagement envelopes, and melee weapons are particularly effective, then it turns into an oversized variant fo the beat-em-up with some interesting twists. Personally, my suggestion would be to try to make as many different techniques as possible effective but not overwhelming - because different techniques will naturally do better with different temperments and different sets of *player* skills, which will mean that players will tend to have to design for themselves rather than pulling The Design Of The Gods off of a strategy website, which in turn means that design skill will be moredirectly useful and important to the player. It also means that the battlefield will be full of people using wacky strategies against each other (It's that moment when the Ninja Knifer manages to pretend to be unaware *just* long enough to draw down the Mammoth From Above, slipping out of the way just in time - and look! he was standing on hi-test mines! He must have a Combat Engineer friend. The Mammoth gives a solemn Farewell to Legs.)
The first, question, of course, if whether you want this to be at all practical. It's almost certainly just a thought experiment anyway, so is it a thought experiment about What Would Be Really Cool For Me Personally, or is it a thought experiment on How To Make The World Better By Giving It The Perfect Mecha Game?
The second question is what sort of a community you want to have. Clearly, The Best Mecha Game Ever would *have* to have multiplayer. Preferably, it would have an option for large numbers, to allow for entire battlefields of mecha going after one another in realtime. Taking *that* to the Coolest Possible Level, in turn, would require some reasonably sophisticated tools of communication between allies, to let people who are good at strategy and cooperation and building complementary mecha to really shine. This all means that somewhere, somehow, you're *going* to sprout a community, if you're at all successful, and that, in turn, means that a large portion of the gaming experience is going to depend directly on who exactly is signing up over the internet to play with you, and what attitude they're bringing to the entire thing. Fundamentally, the community is shaped by the game, and the community shapes the game experience in turn. What sorts of gamers are you rgoing to encourage? What sorts of attitudes are you going to encourage? Who is this game *for*, really?
The third questions exactly how deep in the weeds you want to go. It was initially the second question, but you really need to answer the previous question first before you get here. How much numbercrunching do you want people to have to do to build/optimize their mecha? Do you want it to be the sort of thing that can be done simply over the course of five to ten minutes, the sort of thing that takes half an hour and a book/design program, or the sort of thing where "I design mecha in my head" is a potentially awe-inspiring boast? Computers *allow* for a great deal of complexity. Given that, do you want it? If you make it so that an endgame mecha takes five to ten minutes to put together, max, and winds up all chibi and cute with a couple of equally cute superpowers and your choice of a primary fire and an alt fire weapon, then you'll get a fair number of casual players, but most of the more serious players will lose interest pretty quick. (Not all of them. Everyone needs a break sometimes - but most of them.) If you make it the sort of thing that takes at least half an hour of serious thought and a reasonable head for numbers/strategy to make moderately competitive, then you will keep a lot more of the more serious players, and some of them will happily burn hours in front of your design software/books, logging in primarily to test and refine their designs. On the other hand, getting the casual players in at that point will require that you make the learning curve nice and gradual, particularly at the beginning, and most of them won't do well in competitive play even if they do make it through the single-player missions. If you make it the sort of thing that takes three or so hours of strenuous numbercrunching and an encyclopedic knowledge of the rules and available gear - well, you'll be losing a fair number of even the "serious" gamers, and your community will turn somewhat insular. The sorts of people who actually enjoy stuff like that over the long run tend to be moderately obsessive introverts with the feeling that achievement requires suffering, and the matching feeling that if something must be suffered for to be had, that's an indicator that it's worthwhile. They tend to make very specific sorts of social structures, and are almost fanatically loyal to the games they devote themselves to. Now, of course, it's not just one dimension. There are all sorts of tweaks - from optional extra complexity that people can open up (there's never any reason why you'd have to muck with the gear ratios on your tank treads, but there are some nifty tricks that you can run with if you *do* - particularly if you happen to know what sort of terrain you're likely to be fighting on in advance) to starting slow (everyone starts with the same chassis, and it's going to be a while before you can afford to swap out your body, legs, or generator. Look! your life just got simpler.)
Oh - and since you're building the cosmology from scratch, I'd suggest some excuse for fitting in a mecha-based AI/helper-person, with a moderate range of choices in voice. At its simplest, this would be a simple AI with a choice of six voices, and a set of swappable modules that have moderate in-game effects (you have three AI slots, slotting this module will improve your accuracy with missiles, etc.) At its most complex, this would involve a variety of different AI/copilot options, with different modifiers and effects, that developed over time as they gained experience alongside you (If you go with the structure of Living Crystal that you dug out of that crashed alien ship, you have to avoid vibration-based weapons, but your ECM and ECCM go way up. If you managed to save the Old Scientist's Niece, she'll tune up your engines and help with repairs, along with whatever else she does. Also, she's cute. And so on). This is partially because many of the people who would wind up playing this game are male and at least somewhat low on affection, and the option to have a friendly, identifiably female voice running around with them on the battlefield would noticeably improve the game for them (and give the choice to eject a lot more emotional impact - unless they'd paid for the appropriate extra modules). This is partially because almost everybody feels better and gets more emotionally engaged when they have someone rooting for them and helping them specifically. This is partially because just including a persistent, user-selectable voice gives all sorts of ways for people to feel cooler. The guy who chooses the Boris Karloff voice is going to have a noticeably different experience than the one who chooses the Blood For The Blood God voice, which, in turn, will have a noticeably different experience than the guy who picked the Sultry voice (and to be fair, if we're going to have the Sultry Woman voice, we should have a Sexy Man voice as well) but they're all going to have a *better* experience, because it'll help them get into the feel of the character that they want.
And while we're talking about cosmology... this one doesn't matter but so much, but... Normal, straight-up hardtech mecha? Mecha with a few bits of bizarre and poorly understood alien tech thrown in for spice? Mecha and magic? Biomecha?
For that matter, you want to consider what you want to encourage. Just like the dogfighting vs Age Of Sail discussions over in fenspace, the shape of the battlefield and the overall gaming experience will be hugely affected by What Works. If you can pick up a few levels and then start clearing the field of mecha twice your size and eight times your cost by blowing their heads off at extreme range, then the entire game turns rather quickly into an elaborate stealth/detect/sniper situation (shaped further by how much stealth is actually available). If The Things That Work are all within relatively short-range engagement envelopes, and melee weapons are particularly effective, then it turns into an oversized variant fo the beat-em-up with some interesting twists. Personally, my suggestion would be to try to make as many different techniques as possible effective but not overwhelming - because different techniques will naturally do better with different temperments and different sets of *player* skills, which will mean that players will tend to have to design for themselves rather than pulling The Design Of The Gods off of a strategy website, which in turn means that design skill will be moredirectly useful and important to the player. It also means that the battlefield will be full of people using wacky strategies against each other (It's that moment when the Ninja Knifer manages to pretend to be unaware *just* long enough to draw down the Mammoth From Above, slipping out of the way just in time - and look! he was standing on hi-test mines! He must have a Combat Engineer friend. The Mammoth gives a solemn Farewell to Legs.)