![[Welcome to Sensei's Library!]](images/stone-hello.png "Sensei's Library")
Tournament performance produces winners. It is not necessarily the job of a tiebreaker to say who is the better player between 2 people, it is the job to say who performed better during the event. One extra game does not say which is stronger for certain, indeed 10 or 100 games might not establish a clear difference in strength between 2 players. Therefore to some degree a playoff is a lottery, should SOS predict the outcome of a lottery?
If you think that play-off is a good tie breaker, then yes. If you think that play-off is a bad tiebreaker, then you disagree with proposal #1, and it is irrelevant.
The thing is that a tournament is usually a fixed number of rounds. If you have time for a play-off that's fine. If you don't, are you absolutely looking to see who would win a play off? Maybe Player B always loses to Player A due to style differences. However Player B is much better at beating other people. I do not say that a play-off is a bad tie breaker, merely that it is not the sole purpose of a tiebreaker to look at who would win between just two people. A tournament is generally a competition involving many people. A tiebreaker must consider general performance. You must also factor in things like tournament stamina, maybe C starts strong but wears out towards the end. Maybe D panics too much in big games.
I suppose that really I think that a tiebreaker and a play-off are almost two distinct things.
The thought is that is one tie breaker is good, then if another tie breaker gives the same or similar result is also good.
A play-off need not be between two players, it might also be amongst a larger group (eg 4 players playing two rounds knockout, or three players playing round robin, or whatever). A play-off may thus not even give a result. Between two players a best-of-100 match might be a very fair tie breaker, even though it might end in a 50-50 draw. (and SOS for example can similarly result in a tie).
It is true that there might be factors like stamina that might influence a tie breaking play-off or extra rounds. Another tie breaker may or may not be able to predict those influences. (eg: I think that in many scenarios Direct Comparison on average would predict that the loser of the DC tiebreaker has less stamina than the winner).