Depending on the length and area of the voyage and other factors (enemies, storms, mistakes, etc.) it could vary

I have several statistics for the 16th and 17th century for ship loses on one of the longest routes of the time: from Europe to India.

Here is the table for Portuguese ships from 1500 to 1600: Table as we can see in the roughly first 80 years loses were around 5% (31 out of 620) while in the next 30ish years they skyrocketed to almost 19% (35 out of 186). The reasons for the increase are mostly due to changes in the Portuguese ship and route choices due to various factors (wars and threat of piracy being some more important ones) which lead to overloading the increasingly bigger and harder to maneuver ships, later than ideal start of voyages, and avodiance of stops and certain places infested with pirates, all of which turned out to increase losses. Here is another table with losses data broken down to half-decades. It's slightly harder to follow which is why I recommend the first.

Another table tries to break down the reasons of losses into minor categories. Here we see that at least significant number of losses occurs because of poor navigation, which usually means near the coast and not on the open ocean. Storms also often had more terrible impact near the storm then in the open, but of course it was subjected to circumstances .

We can also contrast the Portuguese losses to Dutch East India Company VOC in the 17th century, in this table. Here we see their losses being between 2.7% and 4.5% with more losses occurring on the outbound voyages (which is also true for the Portuguese losses)

All in all, for the 16th and 17th century we see the losses on the longest and most difficult voyages were usually 3-5% so 1 in 33 or 1 in 20 with possibility of higher percentage of losses - 1 in 5 - if multiple bad choices/conditions align. On shorter voyages I suspect the losses would be smaller but as I don't have data I can't say anything