My own take on the odds of life originating by chance alone is to look at the scales of things -
1.3 billion cubic kilometres of liquid water (on Earth ie one planet)
= 1,300,000,000,000,000,000,000,000 ml
About 1,000,000 bacteria per ml live in sea water, so if the chemical precursors for those are present in primordial sea water we get enough to make...
= 1,300,000,000,000,000,000,000,000,000,000 bacteria's worth.
Give it 500 million years of chemical reactions that happen at much faster than 1 per second per ml rates
I'll be very conservative and say only 1 reaction per second... rather than the millions or more I suspect would be the case -
= 15,750,000,000,000,000 seconds x 1,300,000,000,000,000,000,000,000,000,000 bacteria's worth
= 20,475,000,000,000,000,000,000,000,000,000,000,000,000,000,000 opportunities for random chemistry + selection make the appropriate complex chemistry for earliest ‘simple’ life.
Now this isn't intended to be definitive by any means - add or subtract a few zeros if that makes you happier. It is just an attempt to see how "very unlikely" fits with extremely large numbers of opportunities for "unlikely" to happen.
(Note, I posted this elsewhere and repost it here. I had some version of this in "Abiogenesis - science or faith?" thread, but Search is not bringing that up, nor does it appear any more in my contents pages. Don't know why.)