What you were expecting to happen, and what actually happened?
Expected 28 gems destroyes on cast. Actually only 15 gems destroyed.
What are the steps to make it happen again?
Cast with Dragon’s Eye weapon
Do you have any screenshots or video you want to share with us so we can see the problem? Attach them to your post!
https://youtu.be/mbuhSBL4GWQ
time about 00:06
This is going to depend on whether or not the game enforces all 28 destroyed Gems to be unique. If it allows the same Gem slot to be selected a second time, even if already destroyed, then the actual average expected destroy count is in fact:
64*(1-((63/64)^28) ≈ 22.82
15 is on the low end of the distribution curve here but not out of the question. If instead a given slot is only allowed to be destroyed once (which is a reasonable expectation given the wording), then of course you’re being short-changed by a considerable amount.
I had it all typed… Damn you Lyya! Good answer.
OT: Merlion’s cast doesn’t seem to obey this oddity of Random selection. Like Ships Cannon does. More testing needed.
Yes, i understand that game selected duplicate gems. But as you say i was expected that it will be 28 unique gems.
And in case if the game works as intended then description should be changed to something like “destroy up to 28 gems” I think.
I’ve never observed other random gem destroyers behave like this, though. I’ve had situations where I got Marilith boosted up to destroy exactly 64 gems, and she invariably destroyed the entire board several times in a row. It would be odd that they created a new mechanic specifically for this weapon and then didn’t document it.
You’re right, and I just checked the data. It’s a bug, and I know the cause.
The weapon text should in fact be reading: “Destroy [(Magic / 2) + 1] Gems. Transform a random enemy into a Baby Dragon.”
Which would have been immediately noticed if even a single person had tested it before release.
There is no QA in this game. They write code and push it out to servers for us to alpha test to them. Its like they are creating a game for someone else and we are part of the dev cycle to find the bugs.
I can understand if an obscure bug occurs that requires specific conditions that would be understandable to miss in testing, but there is no excuse for this one.
We’re aware of this issue and a text fix should be going out soon.
Is the fix going to make it destroy 25 (or whatever) gems without the ability to duplicate spots (cause honestly, it says destroy X number of gems, if a slot is chosen twice its not actually destroying a gem the second time as the slot is already empty)?
The issue was that the text on the card didn’t match what the spell was actually doing. As @Lyya mentioned, the text should be saying [Magic / 2] + 1, not [Magic] for the number of gems destroyed. So the 15 gems reported by the OP was correct, as 28 magic / 2 = 14 + 1 = 15 total gems destroyed. There should be no doubling up of gem spots.
Thank you for official clarification.
But why test something that 99.9999% of players are never going to use once?
How about when you have a troop down to 7 life and it switches that troop to a baby dragon fully healthy at 53 life. Kind of helps the opposition more than you.
If baby dragon enters at same health I could squash it like a bug. At full health it has a great shot to transform into another dragon which is a whole other issue.
But… why would you want to transform it when it only has 7 life left? 
It’s all situational, really.
If a Draakulius with 7 life left is about to fire next round and decimate my entire party, then I guess I’ll take my chances with a Baby Dragon.
It transforms a random enemy. You don’t get a choice
The way it is Dragonator 3000 is a much better choice for dragon class. I use black manacles though.
Test important things, like can you actually click on merlantis to pick it. How that got missed…
Dragon’s Eye text appears to be fixed.