WARREN, Minn. — It sometimes feels like you could use a friend with a PhD to help you sort your soybean, corn and wheat seed varieties for how they’ll work on your farm.

If you’re any more cooped up with COVID-19 social distancing, you might connect with Garth Kruger of Warren in northwest Minnesota. He’s a farmer who also separately operates Evaluation Group LLC. The company started its data and statistical modeling to evaluate the effectiveness of health and other programs but in the past few years has extended it to compare seed varieties, from northern Minnesota to North Dakota, Montana and Wisconsin.

Nearly 50 people subscribe to a website Kruger has established, which compares crop varieties in northern Minnesota as well as points to the east and west.

Farmers pay $100 per year for yield data in various visual forms if they submit at least three seed tags and brand names to contribute for a data base for all participants to view. The fee is $150 for people who don’t care to submit seed tags.

Kruger so far has compared 5,000 data points from 100 site plot locations on soybeans alone, with almost 600 brands, allowing subscribers to quickly see which varieties rise to the top. Each point includes data location, performance compared to the mean average among peers, yield, maturity and brand.

Among his clients is Kent Rivard, 32, of Argyle, Minn. Rivard raises sugar beets, soybeans and wheat. Rivard has been using Kruger's service since hearing about the concept at the Prairie Grains conference — annually put on in Grand Forks, N.D., by seven crop and farm groups in North Dakota and Minnesota — a year and a half ago.

“I think it’s well worth it,” Rivard says.

For Rivard, the analysis is especially valuable for soybean varieties, as a way to see “what varieties double-up among seed companies.”

Psych trained

Kruger, 46, graduated high school and went on to North Dakota State University, graduating in 1996 with a degree in psychology and a minor in statistics. He went on to Mansfield University, a small school in Pennsylvania, to pick up a master’s in rural, community and clinical psychology. After a brief stint with the Michigan Public Health Institute, he came home in 2002 to farm with his father, Jerry. They raised wheat, soybeans and sunflowers, and later corn.

At the same time, Kruger worked full-time 30 miles away at the University of North Dakota Center for Rural Health at UND’s School of Medicine and Health Sciences. While there, he met Richelle Tangen of Hawley, Minn. They married in 2005 and now have four children.

UND is also where he got his doctorate degree in research methodologies.

In 2015, Kruger started using his data skills to evaluate crop seeds. He contacted Hans Kandel, an NDSU Extension agronomist, who mailed data from NDSU research trials.

“I took anything that was a third-party trial — non-company trials,” he says. “I want complete neutrality, as much as possible, just because.”

Kruger requires a minimum of two years of data and five plot locations on corn and 10 locations on soybeans.

“That’s kind of a starting point, but if I had 50 locations over five or six years, that’s even better," he says.

Collecting tags

Kruger says the key to the program is collecting variety tags from farmers. Most farmers know that several companies may be selling the “same bean” but charging different prices.

The program is a way to avoid having to submit your sensitive farm data to large corporations who might not have farmers’ best interest in mind. For soybeans, Kruger focuses on varieties with a 0.001 maturity to a 0.6 maturity or less than 90-day maturity corn. Those might be planted as far south as Ada, Minn., in Norman County, to the west in Bismarck, N.D., or Sidney, Mont., as well as Wisconsin to the east. He starts putting them in field-size test plots of 5- or 10-acres, or strips.

The historical rule of thumb is that industry-leading products have yield plot winning percentages 60% to 65% of the time.

“I want to be in the top third every year,” he says.

Curious insights

Kruger's farm has problems with iron-deficiency chlorosis, or IDC, and — now — soybean cyst nematodes. NDSU from 2012 to 2019 had 454 brands with two or more years of IDC data. He can plot a range of IDC results on a “piano chart” and see the best ratings on the left, with the best being “two and a half standard deviations” better than its peers versus the mean average. Most of Kruger's soils are in the 8.3 to 8.7 pH (alkaline) soils, which are susceptible to IDC. He learned over multiple years that he can only safely buy beans that are “1.8 standard deviations better” than the mean.

Some data he’s collected defies explanation. One variety number in one official test was even suspicious — labeled “10,9,8,7,6,5,4,3,2,1.” The law requires using real registered variety numbers in seed labels.

“Maybe a company doesn’t want buyers to know that their variety is the same that another company is selling,” he says.

In one particular case — A1025934 — is sold by nine companies.

“Based on volume and time of year, whatever, you’re always going to get a different price. But once you know this is the one I want, you can kind of figure out who has it," Kruger explains.

Most nematode-resistant soybean genes are in the PI88788 breeding line.

“You’ve got to be able to rotate nematode-resistant varieties,” he says. “If you are planting the same exact variety from two different companies, and you think you are planting a different one, you’re setting yourself up for big trouble.”

Kruger doesn’t post any price comparisons on his site. Price can depend on factors such as volume and time of year.

“Everyone’s got their own price,” he says, smiling.

Kruger thinks farmers are “parting with their hard-won data” without compensation. In some cases, they give the data away as a membership. Sometimes they pay for the privilege.

“I call it 'stealing’ your data, ‘borrowing’ your data, doing something with your data,” he says.

There is an inordinate amount of venture capital dollars chasing new software platforms.

“They’ve got so much money out there that they’re trying to find a way that can build a platform, so they can collect as much data as they can get and then make it into something people perceive to be useful," he says.