Can facial coding predict how a real jury will react to my witness?

It won't predict outcomes, but it flags where a witness loses credibility with a general audience. If panelists show doubt or discomfort at the same question every time, that's a coaching point worth fixing before trial.

Do I need a full mock trial to run this test?

No. A single mock cross-examination session, recorded on video, is enough. You're testing the witness's demeanor under specific questions, not the whole case narrative.

How is this different from testing deposition clip selection?

Clip selection testing happens after testimony exists, when you're choosing which segments to play at trial. This workflow happens before that, during prep, so you can fix credibility problems before they're ever on the record.

What if my witness is coachable but nervous on camera?

Facial coding separates nervousness from perceived dishonesty. Panelists often read nervous witnesses as sympathetic, not untrustworthy, which can change how you coach them.

How to test mock cross-examination footage before your witness takes the stand

Witness prep is guesswork without data

You run the mock cross. Your witness answers fine. Everyone in the room nods.

Then deposition day comes and the witness folds under real pressure, or worse, comes across as evasive on tape in front of a jury.

The problem isn't the prep session. It's that you and your team are judging credibility with the same brains that are only slightly better than a coin flip at spotting deception.

A well-known meta-analysis by Bond and DePaulo found that humans detect deception at roughly 54% accuracy, barely above chance. Lawyers aren't exempt from that number.

That's the gap facial coding fills. Instead of relying on gut reads from people in the room who already know the case, you show the mock cross footage to opt-in panelists who don't know your witness, your client, or your theory of the case. Their faces do the talking, second by second.

Try it: Schedule an EmotionTrac demo and see second-by-second emotion tracking in action. Or visit Audience for more information.