The perplexing properties of paralysis.

The only thing I really know about Paralyze is that it's stupidly random, and probably affected by MND because everyone else says so.

I'm yet to see any decent testing on it, so I'll have a shot myself.


Our subjects are going to be the cute little onion-heads resident to Sarutabaruta, Tiny Mandragora.


I've chosen these because they are a consistant level (Level -1 according to the wiki).

Other misc assumptions:
-As 'level -1' mobs, they will all have the same MND value.
-Potency of paralyze is determined by a Player MND vs Mob MND check.
-A level 75 player should have sufficient skill to floor spell resist rates (base resist rates are unknown, but assumed to be very, very low if they even exist). As such, the innate magic accuracy bonus from level 2 Paralyze II should not be a problem.

The pre-test
I intend to do 20 trials for each of six conditions:
1) Paralyze 1 cast in a) No relevant equipment b) Full MND equip c) Full skill equip
2) Paralyze 2 cast in a) No relevant equipment b) Full MND equip c) Full skill equip

Equipment sets will be posted when I get around to conditions b) and c).

I expect Paralyze 2 to be better than Paralyze in some way, either duration or minimum/maximum proc rate. Full skill is theorised not to have an effect, as resist rates should not be an issue against level -1 monsters.

MND equipment could still have an effect despite the level of the target monster. Since Slow/Slow II cap at 75 dMND (difference in MND between Target and Caster), if Paralyze operates in the same fashion then I won't be at the cap with 62 base MND. I don't know the MND of target mobs, and the only test for MND I am aware of won't work on these because they're so low level.

For condition a), testing was conducted in East Sarutabaruta on Tiny Mandragora by a Mithra 75RDM/37BLM. Paralyze was cast, from within melee range, and then the number of total attacks and number of paralyzed attacks were recorded, as well as the duration of the spell on each trial. Mandragora were then killed and a new one was found to repeat the next trial (make sure no one is leveling there!). Game day was noted, but was not controlled for.


Attack rounds and paralyzed attacks were recorded by hand on a notepad, the duration of each spell was recorded using the Windower plugin 'Timestamp'. Duration was recorded from the message:
"Player casts Paralyze.
The Tiny Mandragora is paralyzed."
until the message:
"The Tiny Mandragora is no longer paralyzed."

I've done 1a) and 2a). I need to put them in excel or something so I can put them on here, not looked at them properly yet, still not sure of a good way to analyze it.


Some Results
These have just been typed up into Excel and ordered by number of attack rounds and duration


Paralyze 1 vs Paralyze 2 in no equipment (Mithra RDM/BLM)

For comparison's sake:
- pps is a measure of 'Procs per second' (Duration/Proc), rounded to 2 decimal places.
- a:p is the Attack:Proc ratio (Attacks/Proc), rounded to 2 decimal places.

Paralyze 2 has a higher pps, however there is minimal difference, so with such a small sample it's unlikely to be significant (not that I can do any tests for significance, the best I can do with excel is find correlations).

As Para 2 also has a better a:p ratio than Para 1 (3.05 vs 4.2 respectively) and a higher total duration and average duration, this suggests that Para 2 is at least marginally more effective when fighting level -1 mobs. I have no useful implications yet, sorry!

I'll complete 1b)c) and 2b)c), then move on to mid-level mobs if I haven't lost interest by then.



UPDAET: 1b) and 2b) testing complete!

Procedure was the same as before, this time the Mithra 75RDM/37BLM was equipped with the following:



Will write up the details later. But for now here's a juicy little observation:


There doesn't appear to be a proc rate cap, at least for Para 2. It can proc at any time, any number of times for it's entire duration. Whether this means the effectiveness (or potency) of a Paralyze spell are calculated at the time of it casting, or player MND is checked upon each of the mobs' attacks whilst the spell is in effect, I don't know. It's a bit of a mad hypothesis, and it's not the focus of this investigation so I won't dwell upon it.

27/03/09 Right, sorry it took me so long to write this up, I've been lazy.


Paralyze 1 vs Paralyze 2 in Full MND equipment.

I didn't order these results, but it doesn't matter for what we're looking at anyway.

Para1 appears to have not benefited at all from the increased MND in condition b) (Duration, proc rates, pps or a:p hardly differ at all) suggesting that Para1 is not improved at all from increasing dMND, however due to the low level of the subjects in this investigation, it is more likey that Para1 has a low dMND cap, which is easily achieved on the target mobs. The dMND cap (according to this testing) for Para1 is most likely <60, but I can't be certain of an exact number until I find a way to establish the MND value of Tiny Mandragora.

Paralyze 2 appears to have benefited greatly from the increase in MND (duration, procs and a:p ratio have all improved noticeably). Suggesting that Para2 has a higher dMND cap than Para1 (or even no cap).

In order to certify that there is a cap (as opposed to no cap) it would be possible to do so by replicating this test with alot more MND than I used (160~ should be enough to see a difference if there is any), and seeing if there is a significant difference in performance compared to my 136 MND condition. If performance remains the same, then obviously dMND does have a cap on Para2.