Inspired by your earlier posts, I tried to do something similar - but met with little success.
It was the volume of work that was the real problem though, I can see why you've heretofore limited yourself to NZ Racing. I pursued it for just over 2 months of Aus/NZ racing and identified waaaay too many possible stable runners. Of course this number would vary depending on the number of sires/broodmare sires you judge worthy of following: my list eventually comprised around 30 of each, judged by winner to runner, and blacktype winner to runner statistics. To pare the list further I also looked at each sire's stats for the last couple of seasons (to see their quality wasn't dropping off) and their record for young horses (to ensure I was getting a fair chance of early maturers). A qualifier had to have both their sire and broodmare sires on the two lists.
Once the lists were compiled, finding qualifiers was automatic. But then the manual process began.... I wanted their dosage index above 20 (using the list of Aus/NZ sires), a dam or granddam that had at least placed in blacktype races, at least 1 chef in its last couple of generations....all hints you either dropped in your own threads, or that I had picked up elsewhere.
In the end, it left me with a stable of 36 debutantes. I abandoned the effort due to not getting immediate results (
) but reviewed it recently to see if there was any long-term success to be found. Limiting each runner to a max of 5 bets before they hit their first win....it came out slightly less than even.
I had noted Cross stats (i.e. results from previous breeding between the same sire and brooodmare sire, if any) and there was little correlation. Actually, the worse, the better, as far as debut victories went. The less chefs in the last couple of generations, the better, was another unexpected result (so ruling out runners with 0 chefs may have been a bad move).
7 of the 36 won their debut (without any further investigation re trial form, or their competitors). But there did seem to be a clear sweet spot at a certain range of DIs or CDs (and you'd imagine that would be the case, as most horses debut in sprint races - and you'd assume those best suited to sprints to perform better; if this methodology reflects reality). A particular DI range contained 6 of the 7 winners and 3 misses - 2 of which were seconds.
Sounds awesome, right? Not really... There must have been some good trial form, or a "name" trainer attached to what were likely premium purchases per paper pedigree (or alot of people doing handicapping in much the same way as I was) as the odds in this range were quite low: $1.80, $2.00, $2.20, $3.70, $3.80 and $6.50. The 2 seconds were also short; the miss was at $13.
Basically, if a stable runner didn't win their first race, you were doing your best to make your money back. After its first run, the rest of the market was clued in regarding its ability. Given the amount of time spent identifying prospects, my particular method wasn't worth the effort. Certainly, as CosMos has demonstrated, there are ways to have some success at pedigree handicapping, but it ain't for me....
...not unless there's a source that I can scrape dosage profiles from (that include Aus/NZ chefs; I struggle to see the use otherwise... Apparently Black Caviar's "official" dosage indicates she was a 1650m runner, it's not until the local chefs are added to the mix that her Brilliance comes to the fore).