Talk:Awards/@comment-83.44.134.172-20120316125313/@comment-4102115-20120317183019

@ParadiseLover, I rather doubt that the developers ever intended anyone to notice this and I rather doubt they noticed it themselves. It would just be basic programming - the GPU has to layer one bubble over the top of the other? Which one? The one with the higher (or lower) IDentifier. The iOS GPU obviously processes this the reverse way. The developers have to make 3 x Leprechauns and 7 x Decoys (Phase 1) hidden in random "charactors' bubbles". Which ones do they make Leprechauns? (Note, I have not looked at the code at all!) They simply generate 10 random beach-walkers labelled (in laymans terms) 1-10 and given the characters are generated and placed randomly, and on our game-screen we can not see the IDentifier labels of the characters, so they simply make 1, 2 & 3 the Leprechauns and 4-10 the decoys. With assigning a random character (purple woman, dregs guy, suit guy, white woman, red man, etc) to each IDent and placing them all individually randomly on the beach, no one would have stopped to think "hang on, since assigning the Leprechauns to the first three IDents, the GPU layering the bubbles will overlay the lower IDent bubble over the top IDent bubble, so players can determine from that which ones are really Leprechauns".

I would be quite confident of that (but impossible to know of course).

Further to that deduction, I would be willing to bet, in hindsight the (St.) Valentine's Day event was probably done in a similar way. They had a pool of male characters to randomly choose from and a pool of female characters to randomly choose from.

For Phase 1 they would have created 8 characters - #s 1, 2, 3, 4, 5, 6, 7, 8.

And they would have done the program like:

'#1 sex=RAND(male); card=shoe

'#2 sex=RAND(female); card=shoe

'#3 sex=RAND(male); card=sun

'#4 sex=RAND(female); card=sun

'#5 sex=RAND(male); card=star

'#6 sex=RAND(female); card=star

'#7 sex=RAND(male); card=strawberry

'#8 sex=RAND(female); card=strawberry

or similar. If so, it would have taken time ranking which pair was "overall on top" etc, but I'm willing to bet that you could have worked out each pair just from the way the Android rendered its graphics