The A.M.A.N. Trove BC

Your bravery was rewarded in kind...

I had seven loud thuds in six orbs today (2 Mars 4 Venus, one of which was a recharge freebie). Unfortunately apart from a pulse weapon it was all crappy shards.

Nice. My understanding is this is the most amount of Loud Thuds that I've seen in a screenshot. If there is a possibility for 4, I'm not sure how I would break it down. Maybe?

1.) LT on first 3.
2.) LT on the floor (BB)
3.) Rainbow chest at the end.
4.) Terminal coffee gives LT.

I guess this might be a case for Venus orbs, where the Terminal Coffer is a Thud or Loud Thuds. (Just an educated guess)

4 Loud thuds in a run is certainly possible, happened to me last year - all 4 in a row: 7th, 8th, 9th, and Terminal coffer. Apologies for the bad screenshot not showing the first, also not sure if this was a Mars or Venus, but I'm pretty sure it was Mars since those were the only orbs I used at the time.

Sorry for the necro: I have two mules with 300 vouchers saved up on each. I was thinking of running both of them on different orbs; one mars, one venus, and seeing what the spread is.

Do you think this is something worth doing? Is the sample size big enough? Or am I wasting my time?

Likewise should I aim for ILT or just LT?

Well I think you would need to be careful how you tracked it.

I'd probably just stop at first ITG personally, for maximum data probably try and always open 9 chests, but you'll fail a lot.

I'd probably track it like: orbs used, average chest opened per orb, #thuds, #loud thuds, #ITG, #fails, #successes

300 vouchers would be around 18-19 orbs of each type for each char, not sure if that's a meaningful sample size, assuming you tried to open 9 chests and an even distribution i'd expect you'd probably get the mimic first try ~2, and ~2 full opens. If you stop after first ITG or 2 LT or something you'd have better success but less data.

If we community sourced the data (kinda like how Treasure Hunter was done on the original wiki before guildwork tracked it) probably get better results but, its still messy and would have some selection bias to it.

??

Mars = 6 vouchers
300/6 = 50 Mars Orbs (x 2,500,000G = 125 million gil on Asura market)

Venus = 10 vouchers
300/10 = 30 Venus orbs (x 5,000,000G = 150,000,000 gil on Asura Market)

So 80 orbs total. I don't know what "spread" you're trying to deduce, it would be almost impossible to record it properly and glean much from your data alone due to sample size/fails (i.e. mimic on first or second try gives you next to no meaningful data). Idk what "worth it" means, without knowing what your goal is. Do you want specific items, gil, or to "prove" data? If it's the latter, seems like a waste IMO.

OR! RISK IT ALL BABY UWU

Not sure who you're replying to, but if you're testing probably want a similar sample size of each:
300 = 18x10 + 20x6 or 19x10 + 18.3x6

If someone wants to do linear algebra:
6m + 10v = 300
m - v = 0

Of course need whole numbers and it said "about".

I wouldn't do 50 of one and 30 of the other, those are different sample qualities.

I don't know how you assign gil value to the orbs. (i've made very little money from trove)

I should have been more clear.

For a long time the accepted way to spend vouchers was on mars orbs because the upgrade to venus wasn't worth it in terms of probabilities to get good items/gil items. They have since changed it so that venus now has a higher probability of getting ILTs. My question was, given that I have a set number of vouchers to spend(300), is there value in testing the hypothesis that venus is now better than mars in overall efficiency of vouchers.

I'm asking if my experiment is worth doing, or if I should just throw all in on mars or venus and not do the other one.

Venus is already known to be better (now)

If you want to find out how much better, that's not enough