If I remember correctly, older engines of AGEOD are up for licensing, so this might be the way to go.
From my project experience I would say that, research, scratch, script etc eat up a mountain of working hours even before the first line of code is written or the first graphics are drawn.
I can only advise you to develop a general idea of what the product should look like and who would play it. As clear and sharp as possible. (Think Twitter, not Tolstoi
) That spares you a lot of frustration throughout the process.