methylethyl, in the comments at Joseph Moore’s place:
Whenever I see study titles, or headlines, that involve “models”– I mentally add “In Legend of Zelda” or something equivalent to every conclusion. It helps put it in perspective. So, you know “model estimates covid spread by vaccination rate in Legend of Zelda” or “Climate model predicts 3-foot sea level rise by 2050 in Kingdom of Hyrule” or “Model predicts 10% rise in heart attack deaths with 5% increase in calorie consumption in The SIMS” Because any time you’re working with a model, you’re in video-game land, not the real world: video-game land is simplified, has far fewer variables, by definition cannot have unexpected events or outcomes, etc.