You can't predict round by round whether the EXP RMs will proc, but it is known they have a 20% probability of proccing and doubling the EXP for that char for the fight. The more runs you do, the more likely the actual result will be closer to 20%, thus saying those characters will gain 20% more EXP long term is a valid assumption.
The thing is, the difference overall between a party of 3 with EXP RMs and a party of 5 with 3 EXP RMs is really minor, so minor its really not even worth stressing over. As the party of 3 is only ~7.1% more EXP efficient (under the assumption of a 20% proc rate) than the party of 5, the party of 3 would get around 15 chars from 50-80 in the time it would take the party of 5 method to get around 14 chars from 50-80.
I don't even use the RMs and I have every named character at least 65, with 74 characters at level 80. Yes, I know I'd probably have another few chars at 80 if i did, but when refreshing 100+ times on an orbfest, I opt for safe and lazy auto-battle setups.
Short answer rata: adding in RM procs will always favor the smallest possible party.
Long answer: say we get 120k exp for a fight.
1 person with exp rm will average 120% over time (144,000 exp).
2 people with exp rm will average 112% over time (134,400 exp).
This trend continues the more RMs you add into the mix. RNGesus aside, the lower the # of people in the party, the more efficient it is.
Incorrect, a party of 3 all with EXP RMs is equally as efficient as a party of 1 with EXP RM in terms of total EXP per fight. The efficiency goes down if you add party members who are not equipped with an EXP RM.
As for the full math, assuming you own 3 EXP RMs and equip them where possible:
Party of 1, if base EXP is 120k, long term average is 120k * 1.2 = 144k per battle
Party of 2, each member would average 60k * 1.2 = 72k each * 2 members = 144k per battle
Party of 3, each member averages 40k * 1.2 = 48k each * 3 members = 144k per battle.
Party of 4, each member receives a base of 30k EXP, 3 of them with RM (30k * 1.2) * 3 + 30k = 138k total per battle (-4.3% efficiency)
Party of 5, each member receives a base of 24k EXP, 3 with RM and 2 without (24k * 1.2) * 3 + (24k * 2) = 134.4k total per battle (-7.1% efficiency)
This holds true until we receive Luneth's level 99 EXP RM which procs double EXP at a 40% rate, at which point using that in a party of 1 becomes the most EXP efficient.
Five lets me take thunderstrokers and fruitlessly farm RMs faster, so by golly I am going to take five.
Definitely not pulling on Tactics banner lol. Hoarding mythrils again for now
Everyone ready for Lucky pull disappointment #6?
We have another lucky draw coming?
For Celerity 5* characters. Nightmare dungeon is tonight.
Ayyyyyyy; i'm a throw mytirls at this one.
Is there anything I'm supposed to be saving up for? I haven't ever burned mythril for refreshes in the orbfests(joke's certainly going to be on me when I get all my sweet full 3* 11x pulls). I'm a day 1 player with multiple trinity pieces, shout, and PCecil's BSB but a small part of me wants to burn up some mythril on both of the shitty beginner banners. I feel like I'm going to miss out if I don't grab Tyrfing(or something) from the crap banner and sorta want to see what I can get from the new Tyro banner.
Tyrfing isn't really even "that" good anymore now that we can easily cap attack with shout and another minor ATK buff, now Luneth's SSB and BSB, oh boy, they're pure sex, Luneth's BSB especially, multi-hit wind with 100% stun and grants him en-wind and one of his commands is 4x wind hits with stun, its beyond awesome for bosses weak to stun and wind.
I'll admit he hasn't had a place in my party in forever.
I will pull for sure. 11 blanks hooooooooooo let's do it!