The one thing that some games are not very good at is the end game. Some games just feel like they didn't take the time to make sure the end game felt right. Other games, however, don't have an issue with the ending, but have an issue with ties. I read a post about a game of Agricola that lasted 3 hours and resulted in a tie. I play for keeps so that tie would have been unacceptable.(Not really but sometimes ties are just no fun.) "Well, we tied. We have to play again." This is why I have a die in the box to settle any ties in Agricola. This just trades one bad thing for another, random chance, but that is a different topic.
Sometimes tie games are just unavoidable. The game designer might list 100 things to break the tie, but you just happen to tie on all 100 events.(I wonder how many dice rolls a gaming group has gone through because they kept tying on the roll to decide who goes first. See coin flip from Rosencrantz and Guildenstern Are Dead.) You happen to play that 1 in a million game that ends with a tie. That is just something we'll have to live with, but if a tie like that happens every other game, then that's no good. There should have been something to settle the tie.
Like the Starting Player conundrum, there is apparently another game that will settle the difference with ties. TieBreaker was created to prevent games ending in a tie. I have yet to try it but it does sound like an interesting way to settle any ties. It should be more of a novelty and for those 1 in a million games that end in a tie. There should have been something in the original game to address the tie.
I'm pretty sure I have some prototype rules that end in a shared victory. I need to go back and make sure there isn't anything that I can do to fix it. For some games, a little math could solve the problem. If there are 14 total points possible in a two player game, change it to an odd number and make it the first player to the halfway mark. The math for this can get a lot harder when adding other scoring methods and players, but it can be done. Is it worth it? Depends on your audience.
I recently just played Gloom for the first time. The way they handle tie breakers is by making it the player's fault. If you don't want a tie, don't end the game. I think this is a great approach. People that don't mind ties will end the game. People that do will make sure it doesn't end like that. Games that have the players determine the end are going to have this as a big solution. Another game like this is Railroad Tycoon. The name changed since I bought it, but it looks like it's the same game with a board that isn't in French. The game won't end unless a player makes it happen. This shifts the blame of a tie from the designer to the player and I like that. Similar to a stalemate in chess, sometimes that's what the player is shooting for if they can't win.
So this ends my little rant, information, exploration about ties in board games. Are they a big deal? To some people. Sometimes it's a tournament and ties will not work. Can they be eliminated? Yes, but it might be a lot of work for an audience that doesn't mind games ending in ties. A fun experiment would be to make a game that ends in a tie 99-100% of the time. It would make being the only winner that time so much better.
Play games. Always.