Final Fantasy Brave Exvius Wiki

Talk:Akademeia

Status resistance study[edit source]

Because all of the bosses have 70% paralysis resistance and same element resistances and I was planning to smash ELT a few hundred times with 2-3 Sevens, I decided to track successful stuns using Seven's LB and Insult to Injury, as well as seeing how it interacts with Shock Whip which can also paralyze. I am publishing my preliminary data because it's completely baffling in the hope that others will help gather data as well to expand the data set. One strange observation I made early on in this study was that elemental resistance correlated with success rates! Here's the data:

preliminary data summary
subset stats comments
using Insult to Injury:
..with Shock Whip 20.1% (53/264) 50% weakness and +15% stun
..with non-Shock Whip 16.8% (47/279) includes the next three subsets
....with Vernard  1.9% (1/52) 50% resist... but incredibly low
....with Coral Sword 25.5% (12/47) same weakness as shock whip but much higher success?
....with non-elementals 17.9% (28/156)
Using Seven's Limit Burst: Note: Did not track which weapon used for all LB
..at 4-star 13.7% (7/51) 80% LB stun worse than 50% stun with non-elementals?
..at 5-star 27.7% (13/47)
..at 6-star 48.4% (15/31) much higher than expected

Other remarks:

  • Shock Whip success rates were extremely streaky. The trailing average of 64 attempts had a success rate that ranged as low as 6.25% to as high as 35.94%... since 64 is bigger than any of the LB and Vernard/Coral Sword sub-samples, it shows just how huge the margin of error is.
  • Though weapon type was not tracked for LBs, I estimate about a third of the 6-star LB samples were with my Vernard friend. While this is purely anecdotal, the Vernard LBs seemed to stun at the same rate as non-elemental/Coral Sword LBs.
  • All samples were collected with care to avoid confirmation bias. The first 24 samples of non-Shock Whip did not track whether an elemental weapon was used. To avoid confirmation bias, I did not retroactively modify or discard any data (except a few cases here and there were I was sure I made a typo).
  • All LBs were executed on Vajra. Insult to Injury executions did not distinguise between Vajra and other event machina. No data was gathered on Imperial Soldiers.

I welcome any comment... and please, if you're willing, try to gather some data. I will publish additional data after the event. —CodeHydro (talk) 11:53, 24 January 2017 (EST)

Well the event is over. Was hoping for a little feedback or data but here's the updated table as promised:
data summary
subset stats comments
using Insult to Injury:
..with Shock Whip 18.1% (76/420) 50% weakness and +15% stun
..with non-Shock Whip 15.6% (65/416) includes the next three subsets
....with Vernard  5.6% (4/71) 50% resist
....with Coral Sword 23.1% (12/52) same weakness as shock whip but higher success?
....with non-elementals 16.0% (43/269)
Using Seven's Limit Burst: Note: Did not track which weapon used for all LB
..at 4-star 20.8% (15/72) much closer to expected than in preliminary
..at 5-star 31.0% (22/71) Over half of samples involved Shock Whip
..at 6-star 50.9% (29/57) much higher than expected

The 19 additional Vernard uses had a success rate of 15.8%, which is close to the stats for non-elemental weapons which suggests the elemental-resistance effect may have been a fluke of chance, but the possibility isn't entirely eliminated either. What is really interesting is how the 6-star limit burst kept up with the 50/50 success rate on the 70% resistance bosses and the combined success rate of all limit bursts was still above 30% with a sample size of 200.—CodeHydro (talk) 11:35, 27 January 2017 (EST)