From

Location

Message

Date Posted

Deer Park, WI

8Buck posted the following topic a few threads down:

http://www.bigpumpkins.com/MsgBoard/ViewThread.asp?b=3&p=537888

It got my wheels turning on how we can step away from speculation and subjective seed selection and move towards a more refined, objective means of continuing our highly selective, accidental (oxymoron alert) breeding process.

If you'd like to pry the WR out of Beni's hands, or claim the title of state record, or simply contribute to the upward evolution of our seed pool, then a more purposeful, mathematical seed selection process may be in order.

In the thread cited above, I layed out an argument that the only objective means we have for selecting top genetic potential is through a calculation called "cross weight average (CWA)". In its simplest form, one maximizes the average weight of the parents of your prospective seeds.

We can further refine this process by adding the weight of the pumpkin that produced the seed as a 3rd variable.

Weight, there's more. A 4th variable, as proposed by Captain 97 in that thread, suggests that one adds the weight of the pumpkin produced by the pollinator. That's brilliant. I've labeled this process 4-variable cross weight average (4vCWA).

4/7/2015 6:23:03 PM

Deer Park, WI

What we're left with is a pretty nifty way to benchmark seed genetic potential against each other based on their family histories.

By no means does this allow us to select that champion show hog from the runt in a litter. Rather, it increases the odds that you are working with the highest quality genetics available, based on numerics.

Another metaphor...by using 4vCWA and identifying the seeds with the highest average value, its like drafting a fantasy football team from the NFL vs your local high school team. Your likelihood of scoring an all-star draft pick is greatly enhanced by choosing from the highest caliber pool of players.

A limitation in quickly calculating 4vCWA is that we dont always have easy access to the weight of the pumpkin produced by the pollinator. This requires a shift in our conventional seed nomenclature system (a discussion for a different thread). However, Im very interested in a list of the top 10 seeds based on their 4vCWA. Anyone willing to put the research in to find that out?

Im also interested in your thoughts on what I propose here. Thanks for reading!

4/7/2015 6:23:22 PM

Colorado

So Joe that puts the 2323.7 meier and 2002 at the top of that list.Part of picking seeds fun is the pollinator stories that one hears. You know "it was growing 70 lbs a day" or "it was on a 2500# pace before it blew".1882 seed gehweiller pollinator sounds kinda like that.

4/8/2015 12:53:16 PM

Shingle Springs, CA

Joe, I like your approach for seeds produced in the current/prior year for which no data is yet available on pumpkins grown from these seeds. This approach clearly recognizes the very likely potential of the 2323.7-Meir, 2102-Meir, 2096.6-Meir, 258-Hawkley, 2036-Glasier, 1928-Globus, 1916-Barron and 1873-Steil seeds(among others).

However, once we have data for pumpkins that have been grown from candidate seeds, my opinion is that there might be value in looking at the weights of these "children" and then evaluating the parent seed on this basis.

4/8/2015 4:26:29 PM

Shingle Springs, CA

I have looked at the data for the past eight years many different ways attempting to get a reasonably useful view of the seeds based on their children's weights. The one that seems to work the best is to look at only those seeds which had at least 5 pumpkins grown (this seems to introduce a certain level of repeatability into the result set) and to only look at pumpkins which weighed at least 850 pounds (this was done to limit the pumpkins analyzed to those which were likely grown under conditions that gave them a shot at competitive weights - in other words exclude from the result set those pumpkins which did not reach a competitive weight recognizing that weather, patch size, grower experience or other factors may have seriously limited the weight that the seed was able to achieve).

Using this approach the top three seeds were (the top 30 are on my webs site www.shewmaker.com under the "Pumpkins" tab):

1) 1744.5-Fulk-13 (220.3-Debacco-13 x 1789-Wallace-11) with 11 such pumpkins grown averaging 1484.77 pounds (1951 was the largest). The average OTT for the 11 was 411.45 and the average %heavy was 2.09%.

2) 2009-Wallace-12 with a 1458.96 pound average for 64 pumpkins and

3) 1623.5-Wallace-12 with a 1422.48 pound average for 23 pumpkins.

19) was the 220.3-DeBacco-13 with a 1299.10 pound average for 5 pumpkins and demonstrates how this approach does not penalize seeds which came from pumpkins grown for seed and not competition.

I certainly expected the 2009-Wallace to make the list and be near the top. I was rather surprised by the high placing of the 1744.5-Fulk, but it illustrates the value of all of these different seed ranking approaches (yours, mine and everyone else's) - we are looking for seeds that haven't yet reached the status of "obvious".

Thanks for starting the threat - it is thought provoking.

4/8/2015 4:26:59 PM

Austin MN

That's why I am growing a 1495 the top 10 average is over 1500+ I'd say 2009 but I don't have one unless some one wants to donate one lol ;)

4/8/2015 11:48:01 PM