Nine strong El Niños. Nine different winters.
A composite map averages the strong El Niño winters into one tidy pattern that no single winter ever looked like. Here they are one by one, at the detail of NOAA's 344 climate divisions: which third of winters each division landed in, once the long-term trend is removed. The pattern is real, and so are the exceptions.
Why the older winters look so cold
A check on the methodTemperature shows a split: 1965–66, 1972–73 and 1987–88 were cold across much of the country, and 1991–92, 2015–16 and 2023–24 warm almost everywhere. That is not an artefact of the straight-line trend removal. Across all winters, not only El Niño ones, the share of division-winters in the warm third was 5–14% in the 1960s and 13–22% in the 1980s. For the 1990s and 2010s it was 31–49%. That holds whether the trend is removed with a straight line, a quadratic or a 31-year running mean, or not removed at all. The 1960s through the 1980s were a run of cold U.S. winters for reasons beyond ENSO, including the Pacific's decadal swings and the North Atlantic, and the older strong El Niños landed inside it.
So the recent strong El Niños are the better guide to how a strong event plays out over today's background. That background is itself part of the odds, which is why the 2026–27 numbers in the atlas show the trend and the El Niño effect separately.