No, my description is correct, though I left out "assuming the null hypothesis is true."

The equal probability you are describing is for the p-value itself being more extreme, not the result.