I've done quite a bit of experimentation with yes/no predictive spreads both in classes and here on the forum (there was a thread several years ago). Some of what's below is from an unposted blog post that I've been working on for quite a while.
For instance, with "Will X. get this job?" I find about 50% accuracy overall with this kind of clear binary prediction. I've discovered, however, that when the answer was wrong readers immediately start justifying themselves to show how the reading (or their intuition) really was correct. I even noticed myself doing that after I had gotten a yes answer when a person asked about getting a job. He didn't get the job, BUT the job had been eliminated entirely, and so there was no job! Could he have been high on the list before the position was eliminated? Maybe. But we'll never know. The fact is, he did not get the job.
Additionally, there are all kinds of Yes/No questions. Not all of them predictive.
• Will I ever find my soul-mate and when?
• Am I going to get the job I applied for?
• Does he love me?
• Should I stay with this person or move out?
• Is this the right person/job/location for me?
• Can I financially afford to get a divorce and maintain my life-style?
• Was I wrong to do that?
• Did someone steal my computer or did I mislay it somewhere?
• Will it rain tomorrow during our picnic?
A variety of responses are asked for: prediction, description, information, choice, advice and opinion (or even moral judgment). Several of the above questions are really more than one question and therefore ambiguity steps in. You have to ask, Is a Yes-No format really the best way to address an issue. Yes/No questions are closed questions calling for a simple either/or answer and nothing more (even if you allow for a "maybe").
I've seen many readers ask the question, throw the cards and then answer a different question entirely. "Will ___ ever find her soul-mate?" Answered with, "I see you've been deeply hurt . . ." Were the cards that showed "pain" really answering "No"? Does the reader just not want to deliver the bad news? Unfortunately many readers don't even notice what they've done; essentially they're seeking to "fix" the anxiety both in themselves and the client.
I like James Ricklef's version where he will only answer with: "Yes, if . . ." "No, if . . ." or "Maybe, if . . ." It provides a much richer context, opens possibilities and invites discussion. Ultimately the person is left with a feeling of potential.
Personally, I use Yes/No questions when it feels right and have seen some great break-throughs when a person has to confront an absolute (whether they accept that absolute or not).
I've also found a SERIES of Yes/No questions can be used to clarify the parameters of an issue, especially when the answers aren't seen as absolutes, but rather as refining the issue or honing in on what is really important. Pendulum work (dousing) often uses this approach.
There is an art to asking Yes/No questions with Tarot and it's worth learning.