Yeah, that's a general assumption I'm willing to afford to any game being released on consoles through the major publishers for sure. Higher management probably makes sure that their products are balanced enough to be completed in theory (by means of the QA teams). I mean, I can't think offhand of any game released since the 3d era began that I couldn't complete/beat based on internal design cues. There have been games I couldn't beat cuz I hated them, weren't fun, were buggy, etc. But not because I couldn't progress to the end due to vague directions or goals.
I suppose the abstract equation for finishing a game with these assumptions and without outside help would look something like this:
Y = ( ( (A * E) + (P * 1.5) ) - ( (R * S) + L) ) / 10
P = patience index (1 - 5)
L = mean length of game index (1 - 5)
A = intrinsic ability index (1 - 5)
Y= likelihood you finish a game (%)
E = past gaming experience index (1 - 5)
R = degree of game's actor difficulty index (AI) (1 - 5)
S = degree of game's non-actor difficulty index (puzzles, progression design, etc) (1 - 5)
where Y < 0 = 0% and Y > 1 = 100%
That is if you could quantify this stuff, and add a bunch of other variables.