One data point is an anecdote.

Vikings QB Kirk Cousins misfires on a late-game drive. Or his protection breaks down. Maybe both? The Vikings needed just a field goal to win — like, for instance, last year against Tennessee — but instead they lose.

Fans are left fuming. Cousins isn't clutch, they yell. The O-line stinks, they bellow.

Those of us more analytically inclined say whoa, whoa, whoa. We need a larger sample size. Just because something happened once doesn't make it true in general.

So we run the numbers and ... sometimes the anecdotes and eye test — those one-off reactions — turn out to match up with hundreds of data points.

That appears to be the case with some interesting (and sobering) numbers about Cousins, past Vikings quarterbacks and the offensive line in late-half situations that I came across recently — and discussed in some detail on Thursday's Daily Delivery podcast.

If you don't see the podcast player, tap here to listen.

Ben Baldwin, who writes about football and data for The Athletic, ran the numbers for "expected points added" by NFL quarterbacks over the last two years, isolating on snaps taken during the last four minutes of the first half and second half.

That was his way of trying to gauge which QBs (and particularly Cousins) are performing well in situations where their team needs to score — often obvious passing situations if a team is trailing or trying to score just before the half with limited time.

Cousins ranks No. 29 of 32 QBs (minimum 150 snaps) in that span. Many top QBs are in the top 10.

But where it gets even more interesting is when we add years. Baldwin ran the same four-minute data for the last 10 seasons (minimum 400 snaps). There were 39 qualified QBs.

The top seven quarterbacks on the 10-year list — in order: Patrick Mahomes, Russell Wilson, Ben Roethlisberger, Tom Brady, Aaron Rodgers, Drew Brees and Peyton Manning — have combined to win 14 Super Bowls. All of them have won at least one.

Cousins was No. 35. But ... Case Keenum was No. 30, Sam Bradford was No. 34 and Teddy Bridgewater was No. 36.

Those are the four primary QBs of the last seven seasons under Mike Zimmer. All of them have taken a lot of snaps in the last 10 years with other teams, so the data becomes harder to sort, but it does lead to some interesting questions:

Are the Vikings just doing a poor job of choosing quarterbacks — at least those who can make a big play when it matters most?

Is there something scheme-wise or philosophy-wise that has followed multiple coordinators and Zimmer through those years?

Is it an indictment of the Vikings' offensive line, which has been subpar most years in that seven-year span, since protection in obvious passing situations is critical?

How much can be attributed to any one factor is hard to determine. But that doesn't make things any less frustrating to watch.

No, your eyes aren't deceiving you. Vikings QBs, likely in concert with their offensive line, have not been clutch.