Yes. The COLA isn't applied in the calculations except in that section. Unless your computer's clock is running ahead and showing 2026, I suppose.

Is it possible that the future earnings records are configured differently?

If you know of any other sites interested in integrating like this to make the tools both more useful to their users, let me know.

Note these also take into account gender if provided or just average based on cohort tables if gender is not set. However, the gender difference is relatively small compared to the outliers on the health sliders so those are a more noticeable effect.

What do you think? Is this a useful way to increase someone's understanding of the decision space?

Useful beyond just a single recommended filing date based on assumptions (such as actuarial distribution or fixed age)?

You can visually see where the lower earner (Chris in this scenario) often wants to file at 62 years old while the higher earner (Alex in this scenario) wants to wait until a bit later, possibly as late as 70.

The rows and columns represent possible ages of death. The width of each row / column reflects the actuarial probability of dying at that age. While the table shows the optimal filing age for that combination.