"place" means nothing. If there are five consoles on the market, but the sales of the #5 console are 18 million units, and the sales of the #1 console are 25 million units, and each machine ends up with a final attach rate of 6 games, is #5 *really* all that worse off when compared to #1?
Last generation, if either Xbox or GameCube hadn't existed, I think the other one would have failed. They kept each other alive by saying "see, the competition's viable, and I have the same market share." Xbox was the more numerically successful of the two, but GameCube made more money. When you combine the two market shares . it gives you something within reasonable reach of Sony, meaning the two combined couldn't be ignored. But one alone could have been. See DreamCast.
What matters in the long run is final numbers, in comparisson to the competition. All of the competition. If you've got a competitor with numbers that look similar to yours, you're viable. If you've got a competitor with numbers three times as big as yours, you're lagging behind. But you're only a failure if the only one that applies is #2. Saturn was a failure because its numbers were too far below N64 numbers, and PS was trouncing them both. N64 was viable, but not a success, because of its early performance and it's relatively high software attach ratio.
You can't really make concrete rules for things like this.
