There is a particular difficulty in your question as posed since it contemplates a soldier’s odds of surviving the entirety of the war. That can’t really be answered as such because the variables of the individual soldiers, units, battles, etc. preclude abstracting survival “odds” to anything very concrete. Some battles had astounding casualty rates (which includes killed and injured). Some soldiers were not active combatants, or were posted away from most of the danger. (World War I exhibited the phenomenon of “fronts” that were not exceedingly dynamic or porous as they would be later; you could take a leave and be comfortable once you were out of artillery range.)
Averaged over the period from August 1914 through November 1918, and not including deaths in prisoner of war camps, historians Stéphane Audoin-Rouzeau and Annette Becker determined in their work “Understanding the Great War” an approximate rate of attrition at 900 and 1300 soldiers per day for the French and Germans, respectively. These are very rough figures, as are most from the period. Pulling a quick reference from the Encyclopedia Brittanica, the total forces under arms at the outset of war in August 1914 stood at about 1.3 million for the French and 1.9 million for the Germans.
If we focus on the French military to keep it largely limited to the Western front, and also assuming that this person was already enlisted at the outbreak of the war (so we can simplify the rate of change in personnel with mass mobilization), and further assuming the necessary fictions that rates of attrition were equally spread across all French combatants under arms at the outset of war—and lastly assuming my napkin math applied to this scenario is correct—France could expect that half of those initial combatants (if none were wounded, only either active or killed) would be killed by August 1916, and all by August 1918.
Of course, these final figures are such abstractions that they sound ridiculous due to the assumptions necessary to reduce a lived-in life of a solider to an abstracted chance of survival. In the end, the figures that make up this thought experiment can be somewhat illuminating, but I don’t think it’s possible to answer this question in both a numerical and satisfactory way.
Taking peak French forces of about 8,800,00 and about 1,360,000 deaths (again, rough numbers from Encyclopedia Britannica) we can say that between one in six and one in seven persons who served in the French army during World War I would ultimately die in the conflict. These are just slightly better odds than Russian roulette.
My answer to a similar question might be of interest: link