I think I would also need to see the point at which agile projects are scrapped vs waterfall and how much money is sunk into them by time of scrapping.

My company knows agile will fail more often but also that they fail earlier. So they take on more projects and those seemed to be a bit riskier compared to what they would take on if it were to go by waterfall process.

I am not an agile acolyte, but failure % alone is not convincing. “Fail early, fail often” is a common mantra for a reason.