Playing a fairly standard GW token deck against a RUB control shell. On the draw, game 3.

Selesnya Sanctuary
Nest Invader
Raise the Alarm
Mortarpod
Kavu Primarch
Scatter the Seeds

16 land deck, two bouncelands. Keep or mull?

let's look at your ideal line of play:

draw a forest, play it.
draw X, play sanctuary
draw X, play forest, play Invader
draw X, play raise, play Scatter.

this sets you up to a decent position turn 4. those Xs could improve it but I don't want to count on them. but until turn 4 you're pretty weak, and you can only go off turn 4 if you get that land turn 1. you have 14 non-bounce lands and 34 cards left, giving you about a 41% chance of hitting it. if you miss that, which you have a 59% chance of doing, you're pushed back to being weak until turn 5. then you've got a 42% chance of getting one turn 2, giving you only about a 66% chance of getting out of a crappy position before turn 6. in such a high-power format that's pretty damn risky.

Easy mull, I think.

If you don't draw a land you basically lose 95% of the games you'll play, gotta ship that to mulla mulla island.

I would think about this assuming that I will draw a land either turn one or two. You are already down to six, so you would go to five on mull. The question is, can you afford to play your first spell on turn four? And what are you hoping for if you do mull?

Keep. They should be slow, and you have the ideal hand, minus a basic land. Any land will do, as Sanctuary gives you both your colors.

BTW, if they boarded Shrivels, you probably lose.

I chose to keep. I felt that I'd win a high percentage of games where I drew a basic on turn one, most games where I drew two basics in a row, and a fair amount where I bricked on the first land. I also feel that the deck really needs its cards to voltron into something better in order to achieve its goals, and that the average five would fair worse often enough to make it worth the gamble. I didn't end up drawing another land the whole game, so I couldn't even tell you if starting late would have worked. Drawing a land second turn world have been the most interesting scenario, I think.

I guess it's a keep, because a random 5er is what, 25% to win? This deck has a 75% chance (land in 2 draws) to have a ~65% chance of winning (50-50).