Next Moons

The decks aims to go very fast, use NEXT ability, Estelle Moon to accelerate your game and outpace the runner, use MCAAP to fast advance an Elective Upgrade and win from there.

On the operation front one should do a lot further testing. Maybe Biotic Labor is necessary, so cutting Beta Test or Restore might be an option. Beta Test offers some heavy punishment against the runner trying to keep up and Restore can help you get Moon or MCAAP back, so it serves as an out for when your FA plan is contested, much like Biotic would.

Now on the crux of the deck. Obviously we want as many assets as possible to spam out, but also want as many ICE as possible to use the NEXT ability to its fullest potential. Correspondingly we run only 8 agendas. The deck runs pretty low on economy, hence Pop-Up is a good choice.
Regarding the NEXT ability, here is some math for you. The deck currently runs 17 pieces of ICE. The Hypergeometric distribution tells us the following probabilities for drawing x ICE in our starting hand of 5.

x | P(X=x)
0 | 0.1056
1 | 0.3206
2 | 0.3537
3 | 0.1769
4 | 0.03994
5 | 0.003245

So thats about a 10% chance for a complete whiff, which isn't too bad and a 18% chance of living the dream.
Mulligan is hard and heavily influences your winning odds, especially post jackson. But with decks like this it obviously has even higher impact.
Keeping a hand in the more likely scenarios of getting 1 or 2 ICE (mean value is 1.73 ICE) heavily depends upon what the other cards are. Generally you want Moon > MCAAP > other assets > agendas. So maybe consider keeping a 1 ICE hand, if you don't have any agendas but a Moon/MCAAP in it. I tried to make a list for myself, when to keep/mull:

• keep
  1 ICE, 0 agenda, moon
  2 ICE, 0 agenda
  3 ICE, 1 agenda
  more than 3 ICE
• mull
  0 ICE
  1 ICE, >=1 agenda
  2 ICE, >1 agenda

Not sure where to put (3 ICE, 2 agenda) and (2 ICE, 1 agenda) hands, really depends on the kind of ICE.

We could also look at what happens if we tweak the ICE suite to enhance our NEXT odds. In a 18 ICE deck for example, we can get 3 ICE hands nearly 20% of the time and drop the 0 ICE chance to 8%:

x | P(X=x)
0 | 0.0891
1 | 0.297
2 | 0.3607
3 | 0.199
4 | 0.04975
5 | 0.004493

On the other hand removing ICE down to 16 leads to the following probabilities:

x | P(X=x)
0 | 0.1245
1 | 0.3433
2 | 0.3433
3 | 0.1551
4 | 0.0315
5 | 0.002291

One can see, that increasing ICE by one roughly gives us +2% odds for 3 ICE and -2% for 0 ICE hands. But you can test yourself with HypergeometricDistribution[5,n,49] on Wolfram Alpha, where n is number of ICE ;)