From

Location

Message

Date Posted

Deer Park, WI

Let the off season begin.

With the "Terminator" complete, we can close the books on 2018 and turn our sights towards next year. Congrats, Mr. Daletas, BTW, for your scale crushing effort in the final hour.

With an incredible 23 pumpkins surpassing the one-ton mark this year, there will be no shortage of exciting options for next year's seed lineups.

The topic of seed genetics has always perplexed me. We've never had a quantitative means to help guide optimized seed decisions. Its always been rooted in gut-feelings and random dart throws.

Wanting to apply some statistical analysis to seed performance, I decided to start crunching some numbers, extracted from the GPC database on this very site. And I found some very interesting conclusions.

My effort has culminated in a 5 page thorough review of seed genetic performance, including charts and graphs that show which seeds appear to produce progeny above the rest...a list of the top 35 performing seeds as calculated by average progeny weight from over 5000 GPC entries. I summarize the evidence strongly suggesting seeds from the largest pumpkins go on to produce larger average fruit. Weird eh?

Anyways, I encourage you to hop on over to the SCGA website where the article is posted in our members-only section. Not a member? For the same price as a case of beer, you get year-long access this as well as a library of original content articles that take a deep dive into the nuance of growing giant pumpkins.

https://www.stcroixgrowers.org/blogs/news/identifying-the-best-giant-pumpkin-seed-a-statistical-approach

Looking forward to engaging in some lively discussion this off-season!

10/21/2018 5:22:11 PM

Plainville, Connecticut 06062

I have to jump in on this one. In my opinion since the 2009 Wallace, 1625 Gantner and 2145 MCMullen seeds entered within growers hands the genetics have changed for the better and many more pumpkins are going heavy to the most recent chart. Information shows that the OTT’S haven’t changed that much in the past three growing seasons but the heavy to charts have. In 2016 there were five pumpkins grown over 470 inches in 2017 there were only two and this year 2018 there were also five grown 470+. The big but is that this year 2018 the top 25-30 pumpkins all went heavy to the most recent chart except one. Also most times but not always your heaviest biggest pumpkins generally go way over the chart especially this year. We even had a pumpkin grown by Karl Haist that broke 2,000 that had a 436 inch over the top, that has to be a record. Information also shows that the 2145 MCMullen either sibbed or selfed from last year hit quite a few big ones this year more so than in past years when selfing seeds were grown. 500 inches over the top has yet to be broke, I believe 496 has been the largest so far. Pumpkins grown over 470 inches are still quite rare, but pumpkins going 10-15 and even 15+ over the chart are becoming much more common, especially with the biggest of the big. The bottom line that I see is that our pumpkins are getting heavier at a much faster rate than actually getting bigger shells.

10/21/2018 7:11:40 PM

South East

There use to be the AGGC for this. It took all the guessing out.

10/22/2018 8:01:26 AM

Germany

Very nice done.
Hope to see the 3000 in the future!!!

10/22/2018 10:11:00 AM

Brownsville, MD

I’m skeptical of this approach. We’re dealing with small sample sizes and a lot of variables in environment and growers skill level. I just don’t see how we can glean any information from these stats that is more valuable than old fashioned seed scouting. If this specific method in the article is projecting the 1781 or 2109 Zywiec as one of the top seeds in the world, that speaks to my point. Those seeds have been planted several times and haven’t backed up the projection. They’ve been good, but they haven’t surpassed other contemporary seeds in the way that the statistics might have forecasted.

10/22/2018 12:58:47 PM

Plymouth, MN

Small sample sizes for most seeds is a serious limitation. That said, the 2145 presents a potentially interesting case study in that it was planted so widely and that many people who grew 2145 were able to get 1 or more additional pumpkins to scale.

Since reading Joe's article and to satisfy my own curiosity RE: how important "premium" genetics is on pumpkin size, I've been looking at 2145 vs the other pumpkins grown by individual growers. I'm hopeful that this can minimize some of the grower-by-grower variation, although there are elements that can't be well controlled (e.g. if the 2145 became the favored plant).

So far, it looks like the 2145s were ~120-~250 lbs bigger than "the rest" from 2016-2018. Interestingly, the smaller, non-significant difference was in 2016 (~120 lbs), before the 2145 grew the WR. This makes me a bit suspicious that a favored child effect is happening.

I'm looking forward to playing with the data and trying to bin the others into broad categories (size of seed, size of mom or dad, etc) as the off season progresses.

10/22/2018 2:13:04 PM

Whidbey Island

Just looking at the 2145 data for 3 years is interesting. In 2016 a total of 28 were grown over 1000# and made it to the scale. In 2017 there were 64 and this year (so far) there have been 32.

The top 10 in 2016 went an average of 13.8 % heavy, in 2017 the average was 4.6 % heavy and this year the average was 7.6 % heavy.

The average % heavy for all 2145 pumpkins above 1000# in 2016 was 8.6 % heavy, in 2017 3.1 % heavy and this year it was 4.2 % heavy.

The number of pumpkins even to or under the chart were 10.7 % in 2016, 28% in 2017 ands up to 39& (!) in 2018.

10/22/2018 11:52:25 PM

Whidbey Island

That last number should have been 39% in 2018.

So the above data raises some questions.

Was 2016 just an amazingly freakish year and since then things have settled down?

Are the older seeds less vigorous?

Is the fact that more average growers (like me) have jumped in and can't do as good as a job as the big guys so the numbers seem to drop down over all?

If that is the case, why have the averages for the top 10 dropped, if we assume those big growers are still as good, if not better, than ever?

10/22/2018 11:59:44 PM

Lexington, Ky

As the GPC OTT chart "tightens up", percent heavy numbers can't be used reliably when comparing year to year. Wasn't the OTT chart "readjusted" for 2017?

10/23/2018 8:48:43 AM

Germany

I think there is also a big pattern if you look at the weather.
Air moisture and sunhours sould be involved in the thinking about why 2016 more of the 2145 went heavy and so on..

10/23/2018 1:27:01 PM

Germany

*should

10/23/2018 1:27:21 PM

NE Arkansas

The chart was changed last year, So if you want to compare % heavy then you need to use the 2016 chart for all or the 2017 chart for all. Can't go between both charts. Gotta pick one or the other or the data won't be right

10/23/2018 3:11:29 PM

Lexington, Ky

Thanks Jeremy, that's what I was thinking. Thanks for confirming.

10/23/2018 3:28:21 PM

Whidbey Island

Thanks Wexler and jlindley for reminding me of the new OTT chart from 2017. I had thought in my feeble mind that it was changed in 2016, but that is not the case. So that probably answers a lot of the questions. I'll shut up now and plan for next year.

10/23/2018 8:28:16 PM

Willamette Valley, Oregon

Here is the average % chart (using the 2017 GPC Chart for all years) for pumpkins and squash 425" to 450" OTT over time from 2011 to 2018:

Year...N....Average % chart GPC2017 (OTT 425" to 450")
2011...19...-4.2
2012...18...-6.1
2013...18...-2.4
2014...68...-4.2
2015...58...-2.4
2016...61...-1.1
2017...64...-1.6
2018...67...+1.4 (2018 is to-date and provisional)

Doing comparisons in a fixed range of OTT is the fairest way to do it, statistically, so that's what the data above are. (Looking at the weightiest pumpkins is a biased sample of OTTs. Even looking at the top OTTs as a group tends to be bias over time because of slight bias in the chart at the upper ranges; and more fruit end up in the upper range of OTT each year.)

You can see there is a trend for increasing % to chart over time. That rate is about 0.5 to 1 percentage points increase per year on average, but with a high degree of variation from year to year. There was a 4 percentage point jump from 2017 to 2018, which does seem like a big jump for one year.
.
.
Here's the average % to chart of all 2145 offspring over time using 2017GPC chart for all years:

Year...n....GPC17%chart
2016...35...+3.9
2017...72...+2.5
2018...40...+2.3

It's hard to say there was any yearly effect there, but generally those 2145's do throw slightly heavy.

10/24/2018 4:13:04 PM