Here's my take on it. I SotA's basic premise, it was stated that it would be 5 episodes in the same game world, running in the same engine. So, let's liken this model to a house. In this house, I have a basic floorplan I intend to establish, one that contains within it all that I need to live and grow for the very near future. BUT, because I KNOW it I need to expand in the future (in this case, four times), I have to plan for that. So, I draw up some plans that will contain all the basics, yet include in the design the framework I need to expand in the directions needed in the future. I don't know what walls I will need to blow out, or in which direction I will go, but I secure the land necessary to encompass my anticipated growth, lay a foundation that will support most if all the needed expansion, and set core systems so that when future loads are added on, they will either easily meet the demand, or can be easily modified to do so. While I am doing this, I need to run this all through simulations so that I can be as certain as I can reasonably expect to be that all of the above is true, and all my systems perform the way I need them to right now, and will continue to perform as future loads increase. This wasn't done with SotA. My personal design choice: Level ranges per episode. I would have decided upon long range level set for the entire game, say levels 1-400, and the system used to gate this, such as the experience pool system used here in SotA. Then, set a level range I expect each episode to be accomplished within, and design the core elements of that episode around this decision. This would have set the stage for deciding upon progression elements, and given me a framework around which I could design not only the game world itself, but the story elements, and how the story would flow within that framework. Example: Epsiode Design decision: a character should need to progress no higher than AL 80 to complete Episode 1, and 80 AL's for each episode thereafter. Consideration 1: completing the episode should take about 40 hours of gameplay Consideration 2: Length of time between launch of episodes Result: Difficulty and experience gain should be such that the storyline is challenging all the way up through 80 AL's, with with other scenes and adventuring opportunities available to provide progression beyond what is provided past completion of the current episode's story, but not so much that achieving anything beyond 20 AL's higher could be achieved without extreme amounts of grinding prior to release of the next episode. This allows players that have achieved the pinnacle of power in the current episode to begin the next episode without being so extremely overpowered that they blow through the next episode in much less time than intended. With the above stated, then you go about designing the game with it in mind, which will govern decisions on everything from spawn times and creature strength, to skill power and experience gain, and give you a way to control the flow of the story so that the player progresses in a fashion that is consistent with that of a SP game, but also give them room to grow a little while they wait for the content the next episode will provide. Each episode could also introduce new abilities per skill, new skills, new creatures that have different resistances, etc. It doesn't need to have everything on rails, but there was a definite lack of direction and thus a feeling of progression I experienced while playing this first episode, that was exacerbated by all the other flaws and made the whole thing more acutely painful than it would have been otherwise.