TL;DR-version:
Enriched Elunium, Enriched Oridecon, Bubblegum all work by "double roll" method.
Expanded version:
That is, they do not increase the rate of success of refining or drop rate. They give the game a 2nd chance of a random dice roll to decide if the item drops or not, similar to the "double wing effect" of a status inducing armor card.
Remember how Garm card in wings is not 100% freeze? Garm card in wings produce a 0.5+0.5*0.5=0.75 chance of freezing (that is, 75%). After the first 50% of failing to freeze, the game rolls again for another 50% to see if the target will really freeze or not.
Bubblegum Mystery
Likewise, bubblegum does the same. Best example is variant shoes as it's the most likely target people will spend bubblegums on.
Chance of variant is 60%. With bubblegum, you get a 2nd roll. 0.6+0.4*0.6=0.84, that is 84% chance of variant.
Bubblegum, contrary to what item description say, is not a flat doubling of chance. The drop rate of the server will not skyrocket to be a 6x with bubblegum.
Enriched Ores
Don't think that the double roll statistical work need to be explained again.
This is a direct quote from the comments at the beginning of the enriched ore refiner script:
//===== Description: =========================================
//= [Aegis Conversion]
//= Refiner that uses Enriched ores to increase upgrade success.
//= After a conversation with Doddler, it's been established that
//= the advanced refiner works similar the the "Bubble Gum" item.
//= The success percentage is not "increased" however, if it fails
//= You get a second try. This tries twice at the same time,
//= effectively giving you a re-roll on your attempt.
//= - Dialog is only partly official to iRO.
//= - Uses the iRO position for this NPC.
This is written by L0ne_W0lf, a very credible contributor to eA.
Notice how this also reinforce the idea that bubblegum works as a double roll as well.
Still not convinced? Actual script itself:
Script Wrote:if( getequippercentrefinery(.@part) > rand(100) || getequippercentrefinery(.@part) > rand(100) )
{
mes "[Suhnbi]";
mes "Clink! Clank! Clunk!";
SuccessRefItem .@part;
next;
Emotion e_no1;
mes "[Suhnbi]";
mes "Here you are! It's done.";
mes "It's been a while since I've made such a fine weapon. You must be happy because it has become stronger!";
close;
}
else
{
mes "[Suhnbi]";
mes "Clink! Clank! Clunk!";
FailedRefItem .@part;
next;
if (rand(5) == 1)
Emotion e_cash;
else
Emotion e_omg;
mes "[Suhnbi]";
mes "Cough!!!!";
next;
mes "[Suhnbi]";
mes "Cough...Cough..";
mes "What a shame...";
mes "Your equipment broke during the refining process. I had told you earlier this might happen!";
close;
}
Notice that inside the opening IF statement in this main part of the script, the success rate of refine is being challenged by a random-100 roll twice.
"but iROwiki said..."
Who cares? Script by eA determines how a NPC in the server runs.
So, will heRO "fix" the "incorrect" enriched ore rate?
Actually, irowiki isn't too sure where in the world did the refine success table chances come from. Check this link out:
http://irowiki.org/wiki/Talk:Refinement_System . So if irowiki isn't even sure where those numbers come from, how can the current formula be "wrong" and why will it need "fixing" then?