I agree with many of the comments above. It appears to me you have two very different groups of players, and they aren't ever likely to mix well. You might have the occasional newbie who becomes a serious player, but he's usually only going to do that when he experiences a lot of success, then he moves up. If he never experiences that success, he never becomes a serious player. If that guy starts with sharks, he'll get eaten before he can play with the big boys. That makes two different games a solution.
Saying that, I'm wondering this. If you are a serious player, and for you a social game would be a big step down and you would be board, is a "starter" game something you want? It's really hard to have a starter game with more than one or two serious players and they will basically eat everyone else's stacks. If starting players never experience winning, they aren't going to enjoy poker for long.
I'll share what I was taught many moons ago about buy-ins. As what I see for cash games doesn't even come close to what I was taught, I'm not sure where it fits, but I'll share it. I have an idea at the end tied to that.
Cash games
For limit poker, how many betting rounds does the average game have? The types of games you play will matter too. Draw with 2 betting rounds will not need as many chips as a game with 4 or more betting rounds. Seven card stud has 5 betting rounds. If you start playing with twists and stingers (switching a card or adding a card), you add betting rounds. I've played a version of a game called Texas Tech or double-barreled shotgun with 8 planned rounds of betting. That's expensive for anything less than 60x the limit and it needs 100x if played often. With limits you also have to consider the maximum number of raises. The higher that number is, the higher the "x factor" needs to be. A starting stack of 40-60 maximum bets gives plenty of room for play with 2-4 betting rounds. 30x might work for only a couple of betting rounds, but less than that is going to increase variance. The less serious the players are, the more important the formulas are because less serious players are likely to see bigger losses, and they aren't prepared for multiple buy-ins.
When you get to the no limit, you don't know the max bet. So you look at the ante or big blind, and multiply the limit formula by 50%.
Here's an example of the lowest level game using this formula. $.01 ante; $.10 limit. $4-6 buy-in depending on games played. If it's $.01 ante; $.25 limit, $10-15 buy-in. That game with 7-card stud or comparable number of betting rounds only should probably be a $15-20 buy-in.
So when you go $.50/$1, this formula will have a buy-in of $400-600 for $10 limit; no limit maybe up to $1000. That's a pretty serious game. Most of what I hear about at that level has buy-ins nowhere near that. To me, that means new players are just tossing their money away.
Tournaments
Tournaments are different. I'd say 3 factors are critical. You can structure a good game from 100BB to 500BB for say 4 hours, and it not be a luck fest. The blind increases and length of rounds can determine that. Freeze out vs. re-buys must be figured in. The big factors between a social and serious game are going to be the buy-in and number of payouts. Tournaments for more serious players have higher buy-ins and fewer payouts percentage-wise.
Knowing what kind of players you have
Players play poker for all kinds of reasons, but I think generally they can be classified along the spectrum of social to serious to professional. The further apart they are on the scale, the less likely they are to mix.
Idea -- The "baseball" system
Professional baseball has the major leagues, AAA, AA, A, and maybe Rookie league. Applied to poker, you run or know of others who will run games at various levels. I'll call them A-Z. As players have the kind of success that makes them want to move up, there is another game for them to fit into. Every player has a "A" level for them, but it isn't the same for all players. Your A game might be another guy's F game. The spectrum in poker is huge.
If you are a major leaguer, and you want to increase the number of major leaguers, those players have to be developed. The guy who might do well in the Rookie game won't do well at the Major level. Going there to early might keep him from ever getting there. Most poker players don't think along the lines of dozens of levels, but along 3-5 levels.
I knew a guy who hosted 4 games. He called his A, B, C, and Fun. Fun players never got invited to A or B games. He never had them more than 1 level above where they were a regular, except for himself. When he had a Fun player who experienced success, and expressed a desire for more, he got invited to the C game. He didn't even know there was an A or B game. If he became a regular at the C game, and wanted more, then he learned of the B game. He told me his system was like a pyramid. The candidates for the A game were few. In fact, he said to fill his A game required a B player or two. He had the most players, by far, at the C level. He said the system allowed someone not doing well at a level to be moved down to keep them in the game. He had enough of a relationship with these players they would talk to him if they either weren't doing well or wanted to move up. He would give a player not doing well a tip or two in hopes they improve a little. He made serious money in the A and B game. He told me he made his living in those 2 games and 2 outside games somewhat similar to his A and B games.
Poker seemed to be all he did, though he claimed to run a small company. It was an interesting system he shared. I don't know how much was true, but I've heard of others doing that.
If you are willing to play at widely varying levels, two very different games should work, but you might need a step or two in between to move players from the bottom to the top games.
It'd likely be chaos.
