Of course it __is__ **possible** to obtain a 9T Confetti.
The problem is your breeding pattern.
The best way to obtain a 9T Confetti would be to breed for Pupil Size __first__. And only accept a Confetti fur (any color) showing Small pupil from two Confetti Starters. During the course of producing those Confetti offspring, monitor the non-Confetti boxes you're discarding. If any of those boxes show Small pupil, then the Large Pupil Confetti offspring are allowed to breed with Confetti Starters (first choice) or other Large Pupil Confetti who have Small Pupil non-Confetti siblings. If at all possible, breed only Confetti Starters until you have an established line of Small Pupil Confetti.
At the same time, run the exact same program except choose for Mysterious Eye Shape.
Only when you have established lines for Small and other established lines for Mysterious should you cross-breed to produce Mysterious Small.
Only when you have establiished lines of Mysterious Small should you then breed for other trait values such as non-White Confetti. But whatever your next project is, it __must__show Mysterious/Small eyes.
Technically, when selecting for one trait value, you run the risk of loosing others. But the effect is most noticible with eye shape and pupil size. So, if you get side-tracked and decide to work on that cool Tail for a while, you're likely going to loose any chance of getting Small or Mysterious and having to start over again from fresh Confetti Starters.
Finally, remember that a successful Confetti "line" is actually a "stable". If you just have one male and one female, the odds are in a few months you won't have any. You want several. And you want new versions of the project in the pipe-line. I've not done the math, but (lick finger-feel wind) I'd guess you'll want to be starting a new copy (maybe two or even four) of the project every week, forever.
Earlier I mentioned that I took a look at creating a feedback loop in an attempt to 'fix' the problems with breeding populations and stated that doing so leads to the Mandlebrot Set and Chaos.
Here's a fairly interesting video which shows the effect:
The Feigenbaum Constant (4.669) - Numberphile. And a follow-up:
What's so special about the Mandelbrot Set? - Numberphile.
I forget which (or if it was another) explains that the Mandlebrot Set actually includes the Logistic Map when viewed in the real plane (cross-section).
In the video about the Feigenbaum Constant, there is a value, lambda (λ), which represents the rate of reproduction. (This value has been in the news much, recently, BTW. Who says esoteric math doesn't show up in the Real World?) The speaker talks of "the death zone", stable zones, and what happens when the lambda value gets too large. My estimate is that lamda for Confetti is just a bit smaller than 1 (so, in the "death zone") while the value for non-Confetti cats is well up there, probably around 12 or so, putting it well into exponential growth, if not into a Chaos zone.
(ETA: Oops, did a check and 'my' lambda is a ratio of parents and children, where 'theirs' is the result. So what I mean, in their terms, is that KittyCats feels like it's somewhere above 3.5, doubling or re-doubling and possibly chaotic)
I am so frustrated with Confettis I wish I was seeing Red!
