Publishing papers and getting good results are likely the most important factor of you getting successful Ph.D. defense. However, one more thing I would say, that is equally important, is to keep your supervisor "happy" and with "healthy" relationship between you and your supervisor. If you keep making your supervisor unhappy, there is a 99% chance your supervisor will simply ignore you in the long run nor trying to help you graduate. So yes, you are on your own.
Here, rather than telling you how to make your supervisor happy, I would write instead of the opposite so it is easier to see the picture.
1. Sprouting some random "brilliant" ideas without first reading papers: you are a fresh 4.0 GPA confident graduate. Now you have some very wonderful brilliant ideas in your mind and thus you decided to share with your supervisor to show how smart are you. Most likely, the response of your professor is "did you read this X paper" or he/she simply make a "doubtful" face (trying not to offend you). Your supervisor is probably thinking "he/she is wasting my time....".
2. Propose ideas without any graphs/papers/prototypes: maybe you read a lot of papers. Now you decide to go to your supervisor and propose to him/her some brilliant ideas. Yet, you did not bring any supporting materials with you. Even your supervisor may be an expert in that area, it is very difficult for the supervisor to judge whether the topic is really good. His/her response is likely "Urr........hmm.........ok.....sounds good". But after you actually implement the idea, your supervisor may tell you that it is a bad idea. You are shocked.
3. Don't take notes: you think you got a 1TB SSD drive in your brain and thus you don't have to take notes when your supervisor gives you some valuable comment. The next day, you forget his/her comments and went to your supervisor and ask again. Tomorrow, your supervisor will buy you a notebook or show you his/her brilliant notebooks (implying you NEED to take notes!).
4. Miss meetings: you often miss regular meetings with your supervisor. First time ok, second time irritated, the third time...no, no...there is no third time....you are out.
5. Ask many questions: you love to ask a lot of question to your supervisor on "how to do X", "which algorithm should I use", "where to publish". Your supervisor will go home and think "why this student cannot even do these simple things" or "why this students don't even try to learn" and likely make the impression that you are not worthy of getting a Ph.D.
6. Repeat the same mistakes: your supervisor commented some mistakes on your paper draft. Yet, you keep repeating the same mistakes. First time ok, second time irritated, the third time.....yeah....your supervisor probably gonna wait for 3 months before he/she will check your paper again.
7. Call your supervisor: you have some question so you decided to simply call your supervisor. First time ok, second time irritated, the third time, your supervisor will change his phone number or he/she will post some announcements during his lectures. Your supervisor assumes you know he/she is VERY busy so stop calling if not urgent. Email is usually the best in academia.
8. No results/papers after a year: your supervisor is very kind to you in the first 3 months, giving you a lot of freedom and encouragement. Thus you thought that you can do your research slowly with your supervisor helping you. After one year, you have yet to have any good results nor any signs of possible publications. Your supervisor will email you to call you in and start asking you "how are you doing?" with a smiley-evil face.
9. Don't know how to write emails: long emails, unclear subject line, no opening/closings, unclear what you want from him/her. These are all the signs for the supervisor to suspect that you probably are not worthy of getting a Ph.D.
10. Love to argue: your supervisor gives you some comments. You decided that it's not right or you are trying to defend yourself. First time ok, second time irritated, third time.....yeah, your supervisor will stop giving you any comments and leave everything to your own fate. (Tips: Never argue with your supervisor. He/she is almost 99% always right or at least partially right. Not to mention that it is not worth for you to argue anyway.)
The TLDR is that your supervisor is also human. He/she is nice but to only a certain limit. Not to mention that almost all professors in the world believe that a Ph.D. is not only about knowledge, but also of one's character.
Note: If you want to know how to make your supervisor happy, instead, read Superman Ph.D. student.