Tag Archive: cherry picking

(Submitted by  blog reader, loyal Seahawks Season Ticket holder Bill Young)

Like everyone else here in Seattle, I watched the Superbowl and noticed more than a few coincidences.

  • Of course Seattle is home of the 12th man.
  • Superbowl XXLVIII (48)
  • 4 + 8 = 48
  • Seattle scored at 12 seconds into the game.
  • Seattle’s first touchdown was at 12:00 in the second quarter.
  • Seattle scored at 12 seconds into the second half.
  • Russell Wilson passed for 206 yards (206 is Seattle’s area code)
  • Seahawks final score 43 points
  • 4×3 = 12

What are the odds?


Below are the extended notes provided by cognitive psychologist and statistician Barbara Drescher for use in Skepticality Episode 230. Take a look and leave your comments below. Also, please be sure to listen to the podcast  for our own sarcastic and hilarious commentary. Also, visit Barbara’s blog.

“The 12th man” refers to the fans. Teams are only allowed 11 players on the field at a time, but fans can affect the outcome of the game by cheering (it’s a subtle effect), so they call the fans “the 12th man”. I’m not sure how Seattle got to be known as “The Home of the 12th Man”.

The odds of all of those things happening are difficult to calculate, but not really worth examining. It’s a good example of post hoc data-mining and cherry-picking. He looked for things that relate to the number 12 and ignored all of the things that don’t. For example, what are the numbers of the players who scored? What was the score at the half? How many coaches were on the field? How many turnovers were there? Lots of numbers involved in the game don’t fit the idea that “12” is a special number that matches the Seahawk’s marketing campaign about the 12th man.

Irish Roots

(Submitted by reader Bobby Goldstein)

[EDITOR NOTE: Bobby requested that the names and dates of his grandfather’s name be anonymized for this post.]

I recently learned that because I have a couple of grandparents who were born in Ireland, I can get dual citizenship. This is pretty exciting to me, and so I’m doing the research and retrieving documents for folks who were born more than 100 years ago.

I knew my grandpa pretty well, and I knew his birthdate, and what county he was born in. The helpful woman at the consulate suggested I start by contacting the parish churches. So I started emailing parish churches in County Roscommon, and I got a hit. One church DID have a John O’Smith born on 1901-10-11. We emailed back and forth and they sent me a link to the government website where I could, and did, order a birth certificate. After I ordered it, I went back and looked at the emails from the church – they had a different birthdate for him, and I just hadn’t noticed – he was born JANUARY 11, not October 11. Everything else checked out, including BOTH parents’ first names.

I checked with some relatives, and they all thought it was plausible that either Grandpa had gotten his birthdate wrong somewhere along the lines, or that someone had mis-transcribed the birth month.

But then I heard from a different parish church in the same town, and they had a John O’Smith born in 1901 on OCTOBER 11. Also, on the second one, while the parents’ (i.e. my great grandparents’) first names were the same, they had a different birth name for my grandmother.

So I called the records office, and there were 2 people born with the name John O’Smith in 1901, and I’ve got the birth certificate for the wrong one.

Now, John O’Smith is not that rare a name, and while Athlone is not a huge city (population 20,000 now, I have no idea what it was in 1901), it’s not tiny, so that part of it seems like not that big a coincidence. But:

  • Both were born on the 11th of the month
  • They were the ONLY 2 John O’Smiths born that year in the county (I THINK they said the county. Might have been the town)
  • Both of them had a father named Patrick
  • Both of them had a mother named Brigid

I know that in cases like comparing Lincoln and Kennedy, you see so many coincidences because there are so many potential coincidences, and so you can cherry-pick. But, here, I can’t cherry pick. I only know so much about my grandfather’s birth. And yet just about everything (except birth month and mother’s maiden name) matches up.

How about that?

Below are the extended notes provided by Barbara Drescher for use in Skepticality Episode 207. Take a look and leave your comments below.

This isn’t really as much like Lincoln/Kennedy as it is like most of the other name stories that we get. There is certainly some hindsight bias involved (in the Lincoln/Kennedy comparison, we notice the things that match and not the myriad of things which do not), but we should still be impressed with the number of things which are the same. Except that we shouldn’t.
I don’t have a good source of name frequencies in Ireland to quantify this, but being of Irish decent myself and having paid some attention to my own family tree, I can say that these names are indeed extremely common. What’s more, individuals born in the same year are much more likely to share a first name than those born apart because name popularity follows a trend. Some, like [John], Brigid, and Patrick are extremely common and timeless names most likely honoring a family member. Since the individuals share a last name, it is highly likely that they were related somehow, increasing the probability that the name would be shared.