A popular focus in the polling for the presidential race right now is to ask about the various potential general election match-ups: What if it was Clinton versus Trump? Or Sanders versus Cruz? Clinton versus Rubio? Sanders versus Bush? As often as we hear that kind of data, professional pollsters will tell you that these kinds of questions, at this stage at least, have virtually no predictive power. The problem is that the question is based on a hypothetical “What if…” scenario. People have enough trouble self-reporting their attitudes in the relatively static context of current conditions. Add in the hypothetical scenario and you are also adding in a number of other unknowns: For example, what kind of amazing turnaround (or spectacular collapse in the other candidates) would have catapulted Bush to the GOP nomination? If the Democrats end up with Sanders, will that be because he effectively courted and converted the establishment wing, or will it be because his progressive wing overpowered the establishment? Our preference in any of these match-ups would be shaped by the events leading up to it. Or, more simply, the hypothetical match-up question posits a future situation that we just have not had a chance to think about or to get used to. Respondents can and will give an answer, but the answer isn’t terribly reliable.

The problems with hypothetical questions apply to litigation as well. Particularly when taking depositions, attorneys will want to get a witness to weigh in on a hypothetical scenario. The physician-deponent, for example, might be asked what they would do if a patient presented with symptoms A, B, and C. That can be an attractive question for the attorney because it secures agreement at a very general level, and that general answer can then be mapped against the more specific facts of the case. The attorney can also use the hypothetical question in order to get around an argument at another level: They can set aside for the moment the physician’s belief that the patient in question lacked symptom B, for example. The hypothetical questioning style can be tricky for the witness, though. Prepared to talk about the facts, witnesses can instead find themselves in a world removed from those facts, and being led along by just providing the seemingly obvious or easy answers can lead to trouble. This post will share four ideas on handling the hypothetical question.

No witness benefits from a scripted approach on any question or type of question. But all witnesses benefit from having a tool box of sorts to draw from when faced with these kinds of questions. Not every item in the toolbox will apply to every question, but some general familiarity with each will keep the witness in good shape, able to answer honestly without being played. Here are four tools:

1. Acknowledge the Hypothetical

The record can sometimes get fuzzy on the question of whether we are talking hypothetically or talking about the actual facts. That confusion is sometimes unintentional, and sometimes a product of the attorney’s attempts to use the hypothetical to make a point on the actual facts. For the witness, it helps to keep the context clear by beginning the answer with, “In that hypothetical situation…” or “That was not the case, but if it was…” Explicitly referencing the hypothetical nature of the question also serves as a reminder-to-self, for the witness, a way to keep the limits of the question in mind.

2. Call Out the Missing Information

The hypothetical is a partial world with some details included and other details unmentioned. Knowing that the patient “had a headache,” would tell you one thing, but not the persistence, the severity, or the myriad other symptoms that would accompany that patient’s presentation. A first line of defense against hypotheticals is to point out that only part of the picture is filled in at this point. Useful phrases to rely on would be, “It depends,” or “Not necessarily.” Don’t give these responses too lightly thought, as you could expect an “It depends on what?” as the attorney’s next question. So you can also answer, “I would need to know X, Y, or Z.” If that necessary context isn’t provided, then the witness can very reasonably answer, “I could not answer that without speculating.”

3. If You Can, Answer It As a Hypothetical

The witness won’t always be able to give a fair or complete answer to a hypothetical question (and when they cannot, they should say so). But when they can answer a hypothetical on its own terms, they should go ahead and do so. Disclaimers, like “Generally” or “Based only on that limited description,” can be added in order to avoid the overbroad concession. But within that hypothetical box, if a known and reasonable answer is evident, then avoiding it isn’t typically worth creating the impression of an evasive witness.

4. Pivot Back to the Facts

Even when the question is hypothetical, it sometimes helps to pivot your answer back to the non-hypothetical facts of your case. For example, if the attorney asked about “A patient who makes repeated complaints of severe neck pain,” the physician witness could reply that “Those complaints would obviously be taken into account, but in this case, the charts and the admitting questionnaire show no such complaints.” Responding in that way requires some anticipation of the opposing counsel’s purpose. It requires understanding that the only reason counsel would ask about the hypothetical is in order to draw a connection to the real. And if the attorney plans to go there, then it can be effective for a smart witness to just go there first.

One way of looking at it is that in a testimonial context, no question is ever really hypothetical: Its only relevance is to make a point on the actual facts. So the hypothetical question is another setting where it makes sense for the witness to be sensitive to opposing counsel’s purpose and to understand the tactics.