How are you testing your new set drops?If your sets are high-level items, you need a high-level character probably playing Hell difficulty to get monsters that can drop the high-level sets. Some items are ridiculously scarce even in their normal form (e.g. vortex shields and coronas), which is why certain set items for realm characters cost $$$$ on e-bay (cough).
One test mentioned often in these forums is the Quill Rat test. Go into TreasureClassEx.txt, and find the lines for Quill Rats. Replace one line with entries for the items in one set. Use the three letter codes for the base items, if your set contains a long bow, put "lbw" in the item# column and "1" in Prob# column just right of it---same for each other item. On the left there are columns Pick/Unique/Set/Rare/Magic; put 1024 in Set and a negative number in Picks of the items in your set---if you have 3 items in the set, put -3 in Picks. Again, if you are trying for high-level items, change a Quill Rat (H) line...and you might even have to define Hell Quill Rats as L90+ monsters too. Just remember to make note of your changes so you can back them out after testing. The Quill Rat test worked fine for me. (Aside: this is a common test that maybe deserves its own paper.)
If there is more than one set item of the same type (e.g. rings and amulets), the game will calculate the odds based on the item rarity as compared to all other items of the same code. Does anybody have a formula for this? Does position in the set table count also?
[This message has been edited by Myhrginoc (edited 13 January 2002).]