Max Exposure of Atkins Diet

Hi all, I am trying to calculate the Max Exposure for a Slot Game like Atkins Diet.

Let Max Exposure be the Max Winnings for Max Bet.

Initially I was running billions of simulations recording the max win and calulating its propability from the number of Hits per simulation plays. However, this takes considerable time and CPU power and at the end the results very from normal to exotic, so I cannot trust the simulation.

I decided to start analyzing the problem and use an algorithm. My approach was:

I am looping on Reels to find RNG stops that provide max Reel View winnings for Max Paying Symbol (eg. Wild) and Scatter which I need it for entering Free Spins.

For every Max Paying Prize rule combination of Scatter and Max Paying Symbol, I run this loop to calculate Max Exposure for Max Bet:
{
Max Exposure for Max Bet = Max Base Game Winnings + Max Free Spin Winnings =>

Max Exposure for Max Bet = Max Base Game Scatter Winnings + (Initial Free Spin Winnings + Additional Free Spin Winnings)

I decided to break Initial and Additional free spins in 1st round + remaining rounds:

For: Initial Free Spin Winnings = 1st Initial Free Spin Winnings + Remaining Initial Free Spin Winnings

For: 1st Free Spin Round Winnings = Max Pay Rule Prize x Free Spins Multiplier = 5 in a Row Max Paying Symbol x Free Spins Multiplier

For: Remaining Initial Free Spin Winnings = (Initial Free Spins - 1) x Max Pay Rule Prize x Free Spins Multiplier (*** but it seems obsurd ***)

For: Additional Free Spin Winnings = Max Scatter Winnings during Free Spins x Free Spins Multiplier + Remaining Additional Free Spins Winnings

For: Remaining Additional Free Spins Winnings = (Additional Free Spins - 1) x Max Pay Rule Prize x Free Spins Multiplier (*** but it seems obsurd ***)
}

I undestand an algorithm like this one would calculate the Max Exposure but it would be ... rather exotic, since it uses Max Pays in every case and does not limit winnings by applying probabilities. Also the algorithm assumes only 1 set of additional free spins, but obviously there could be more.

I feel I should involve propabilities in Free Spins to get it right, so I have calucalted with help from other posts the following:

Scatter 3,4,5 Probabilities = p1
Initial Free Spins Probabilities for 3,4,5 scatters = n1p1
Additional Free Spins Probabilities = n2np2
Expected Free Spins For Scatters in View = n1 / [ 1 - Σ(n2p2) ]
Expected Free Spins per Base Game = Σ(n1p1) / [1 - Σ(n2n2) ]
Average Free Spins when Free Spins are Triggered = { Σ(n1p1) / [1 - Σ(n2p2) ] } / Σ(p1)


But I am stuck, I have no idea how to proceed.

Anyone?