>>248891
>>248898
I am not telling you to use Kokona. I am saying that I neglected healing my team for the entire fight and my 3* C.Hare survived for about ~1:45. When I say neglected, I mean there was literally almost no healing for my team. Ako NS has a 45s cooldown so in a <2:00 fight, she only heals twice. For the speedrun to work, I need Kokona NS to heal the lantern in order to heal through the debuff with her EX. At most, only the very first Kokona heal went to C.Hare and by the time I finished healing the lantern my C.Hare died.
My point is that in your case, if you were to mimic a speedrun strat but spend an extra one or two heals on your team instead of the lantern, you can maybe keep your team alive but still be fast enough to improve your score. I.Mari can still work for this if she's sufficiently invested. But I don't know how much you were healing your team in the run you posted. The idea still stands that if you want to push for a faster score, you need to compromise on how much you are healing your team which will add mald into the equation. As an example, if you healed your team 3 times during the run, try skipping a heal and see if you can still mald a better score like that. Or, as I said in an earlier post, try running maldy students like C.Hare for a better score. You won't know if you don't test it out in a mock.
If you're improving for the sake of Platinum, just know that clearing Insane for this TA will probably be tight. NA, a slacker region, had ~3600 out of 5000 slots filled with Torment in the previous Hiero TA (except it was D.Hina's terrain). EU, a tryhard region, had ~1900 out of 2000. Despite the existence of Kisaki now, I expect smaller numbers this time because it's the wrong terrain for D.Hina. Maybe something like ~3000 in NA and ~1800 in EU. If your Insane speedrun isn't going to be close to a 2 minute clear, you're not gonna have much luck securing Platinum. You would probably be better off saving your mental by chilling in Gold or, if your roster allows for it, going for Torment.