>>278702
Well if you want to use the roll-to-45 strategy for a single sequence and then use any leftover pity/guarantee for another limited char...
<Roll to 45 and get sequence, roll next character without guarantee and no pity
>average cost: 22.5 * 1.5 (factoring in 50/50) * 0.9 (for coral) = 30.375
>average cost of the next character: 55.5 (mean cost of any 5-star) * 1.5 (factoring in 50/50) * 0.9 (coral refund) = 74.925
>probability: 17.7%
>probability-weighted cost: (30.375 + 74.925) * .177 = 18.64 rolls
<roll to 45 and lose 50/50, buy sequence, roll next character with guarantee and some pity
>average cost: 45 + (45 * 0.9) = 85.5
>average pity for next character: 22.5
>average cost of the next character: 47.2 (what the pull calculator says factoring in that pity) * 0.9 (coral refund) = 42.48 rolls
>probability: 12.6%
>probability-weighted cost: (85.5 + 47.2) * .126 = 16.7 rolls
<roll to 45 and get no 5*, buy sequence, roll next character with no guarantee and 45 pity
>average cost: 45 + (45 * 0.9) = 85.5
>pity for next character: 45
>average cost of the next character: 60.3 (what the pull calculator says factoring in that pity) * 0.9 (coral refund) = 54.27
>probability: 99.2 ^ 45 = 69.7%
>probability-weighted cost = (85.5 + 60.3) * .697 = 101.6 rolls
Final cost using that strategy to get 1 sequence and 1 more character after:
137 rolls
Meanwhile, the cost of immediately buying a sequence and then rolling another character:
>45 for sequence
>74.925 for next character
119.925 rolls
Seems better to just buy the sequence. Somebody should check my work though, I think the pull calculator I used skewed too high, though not enough to make the roll strategy better.