I'm not an astronomer, though I do work with them. My impression from talking to them is that there is no consensus around whether or not habitable planets are common or rare.
Our principal detection method for finding planets excels at finding gas giants with masses similar to their host star. This is not a great tool for finding habitable planets. Direct imaging of planets is not feasible at an intergalactic scale; I'm not sure it's even really possible for stars further out than a few hundred lightyears. This greatly limits our sample size. The less you know about your distribution, the less accurate your simulation is going to be.
A priori arguments from simulations have the problem of either producing an absurd cornucopia of habitable planets or making too few. Steven Dole's _Habitable Planets for Man_[1] discusses the odds and convincingly argues for a large number of habitable planets, but it's pretty out of date. His ACCRETE simulation is great at producing solar systems that resemble our own, but it is not well-respected by astronomers. The time I spoke with an astronomer about this, he said that current simulations have trouble getting from rocks to planets. ACCRETE is a pretty crude simulation; to Dole and his coauthors the interesting thing about it was that such a crude instrument could produce such legitimate-looking results.
One of ACCRETE's problems is that there are a number of parameters, and if you change them slightly you get very different solar systems. I would think a good simulation would have a small number of these parameters but they usually give you enough rope to hang yourself. In any event, to choose the right settings you must have either physics dictating them or accurate numbers from samples, and we still have a very small set of examples.
As an aside, the sheer volume of stars makes a strong argument that even if the odds are vanishingly small, there ought to still be a mind-numbing of habitable planets.
Our principal detection method for finding planets excels at finding gas giants with masses similar to their host star. This is not a great tool for finding habitable planets. Direct imaging of planets is not feasible at an intergalactic scale; I'm not sure it's even really possible for stars further out than a few hundred lightyears. This greatly limits our sample size. The less you know about your distribution, the less accurate your simulation is going to be.
A priori arguments from simulations have the problem of either producing an absurd cornucopia of habitable planets or making too few. Steven Dole's _Habitable Planets for Man_[1] discusses the odds and convincingly argues for a large number of habitable planets, but it's pretty out of date. His ACCRETE simulation is great at producing solar systems that resemble our own, but it is not well-respected by astronomers. The time I spoke with an astronomer about this, he said that current simulations have trouble getting from rocks to planets. ACCRETE is a pretty crude simulation; to Dole and his coauthors the interesting thing about it was that such a crude instrument could produce such legitimate-looking results.
One of ACCRETE's problems is that there are a number of parameters, and if you change them slightly you get very different solar systems. I would think a good simulation would have a small number of these parameters but they usually give you enough rope to hang yourself. In any event, to choose the right settings you must have either physics dictating them or accurate numbers from samples, and we still have a very small set of examples.
As an aside, the sheer volume of stars makes a strong argument that even if the odds are vanishingly small, there ought to still be a mind-numbing of habitable planets.
[1]: https://www.amazon.com/Habitable-Planets-Man-Stephen-Dole/dp...