Give Them Grass

Nov 17, 2026 · Where the method comes from

The sweet peas and the yellow flag

The idea behind the needle is about 140 years old, and it started with a man mailing seeds to his friends.

In the 1870s Francis Galton, a Victorian gentleman with money, curiosity and no need of a job, wanted to know how much of a parent turns up in a child. People are awkward to experiment on, so he began with sweet peas. He sorted seeds by size, packed them up and sent them to friends around Britain, from Nairn in the north of Scotland to Cornwall in the south, who planted them and sent back the next generation.

The big seeds had big offspring, but not as big. The small seeds had small offspring, but not as small. Each generation leaned back toward the middle.

Then he tried people. He offered £500 in prizes for family records and collected the heights of 205 sets of parents and more than 900 of their grown children. He multiplied every woman’s height by 1.08 so that mothers and fathers could share a scale. The pattern was the same. Tall parents had tall children, but only about two-thirds as far above average as they were themselves. Parents three inches taller than most had children about two inches taller than most.

In 1886 he published it under the title “Regression towards Mediocrity in Hereditary Stature,” and the word regression has been attached to the method ever since.

An aside, because it would be odd to leave it out. Galton is also the father of eugenics. He coined the word in 1883, and the idea crossed the Atlantic and did some of its worst work in the United States. For those who want to go down that rabbit hole, start with Buck v. Bell and bring a strong stomach.

The half of his work that deserved to last is the plain one: decide what is normal, then measure how far the thing in front of you sits from it.

What it has to do with football

Two things, and the first is the method itself. Galton’s normal was the average height. Ours is what a typical crew calls. For every play this season we estimate the chance that a judgment flag comes out, given the down, the distance, the score and the clock, and what that flag would be worth. Over a typical game it adds up to about three judgment flags on each team. The needle is the difference between that and what was called.

This is also how the needle gets at the calls that are never made. A hold that goes uncalled leaves no record. But a game that should have produced four flags on one side and produced one has a gap in it, and a gap can be measured.

The second thing is the warning the sweet peas carry. An extreme result is partly luck, and luck does not repeat. A team that was hammered by the flags one Saturday will usually meet a more ordinary afternoon the next, without anyone having done anything about it. It is the same reason the rookie of the year tends to have a quieter second season.

So one game proves very little, and the site says so every time. “About 1 game in 25” is our way of telling you how often luck alone produces a night like that.

What it cannot do

It cannot see a play. If you watched an official miss a hold that happened right in front of him, you know something the model never will, and you are entitled to keep knowing it.

What the model knows is different. Over ten games, or a season, or a crew’s whole year, it can say whether the totals look like the totals a normal crew produces. Your eyes judge plays. The needle judges totals. When they disagree about a single game, both can be right. When they agree over a season, it is worth paying attention.

Sources: Francis Galton, “Regression towards Mediocrity in Hereditary Stature”, Journal of the Anthropological Institute, 1886. James A. Hanley, “‘Transmuting’ Women into Men: Galton’s Family Data on Human Stature”.

See the needle for your team · More from the blog