A PhD should have a combination of high- and low-risk projects. An advisor who places their students only into high risk projects is gambling with _your_ life and career. If the project succeeds, the PI gets the lion's share of the kudos, and if it fails, you lose years of your life.
Ideally, the projects should be a low-risk project which gets your feet wet in the methodology and is "guaranteed" to produce a paper, combined with a higher risk project that potentially pays off big. Once the first low-risk project works, pick a second low-risk project to replace it; the amount of time you devote to the high risk project can increase once the first low-risk one has worked.
Ideally, the projects should be a low-risk project which gets your feet wet in the methodology and is "guaranteed" to produce a paper, combined with a higher risk project that potentially pays off big. Once the first low-risk project works, pick a second low-risk project to replace it; the amount of time you devote to the high risk project can increase once the first low-risk one has worked.