6 Failure Modes That Kill Plaintiff Expert Witness Depositions in 2026
Verify it yourself — free, no login
See how AI medical-record review links every fact to the exact Bates page that proves it — click any citation and jump straight to the record.
See the 60-second demo →The plaintiff bar lost more medmal cases in 2024 to non-responsive expert witnesses than to bad facts. That's not me — that's defense bar attorneys talking openly at the DRI conference last fall. Your case theory can be airtight, your records perfectly organized, your damages model bulletproof. None of it matters if your expert witness gives ground in deposition.
This is the playbook. Six failure modes that show up in nearly every losing expert deposition I've seen, paired with the cross-examiner technique that exploits each one. Use this as a checklist before you prep your expert.
1. Non-responsiveness
The witness explains instead of answering. Defense counsel asks a yes-or-no question. Your expert delivers a paragraph. Two things happen: (a) the witness sounds evasive on the cold record, and (b) every additional sentence becomes new impeachment material.
The cross technique: "Doctor, my question wasn't whether the patient had risk factors — my question was whether you reviewed the cardiology consultation note dated March 14. Yes or no." Locking the witness into a binary forces either a clean answer or visible evasion. Either outcome favors the defense.
Prep fix: Drill your expert on the difference between "answering the question" and "explaining the answer." The order matters. Answer first, then explain only if invited.
2. Scope creep on credentials
The cross opens with "Doctor, you're not a [specialty X], are you?" Your expert, trying to seem authoritative, says "Well, I'm board-certified in internal medicine and I've spent twenty years..." — and walks straight into the credentials trap. Now the deposition transcript contains the witness's own admission that they're outside their lane.
The cross technique: Lock the witness into broad credentials early, then narrow surgically. "You're not board-certified in [the specific subspecialty at issue], correct?" → "You've never personally performed [the specific procedure], correct?" → "You've never published on [the specific question]." Three questions, three closed doors, before the expert ever touches the case facts.
Prep fix: Your expert needs to know exactly which credentials are in-scope for the opinions they're offering, and where the FRE 702 reliability boundary sits. Walk through their CV line by line and identify every question that ends with "and you've never done X, correct?"
3. Volunteered scope-shrinking admissions
The single most-coached witness behavior in defense prep. The witness, trying to sound reasonable, volunteers something like "Well, of course every physician faces these judgment calls in real time." That sentence — unprompted, off-the-record-prompted — just gave the defense a foundation paragraph for their closing argument.
The cross technique: Defense counsel doesn't even have to ask. They wait. Trial lawyers use deliberate silence after an answer; the witness fills it with extra explanation; the extra explanation contains the concession.
Prep fix: Your expert must learn to feel a silence and resist it. The hardest single thing to drill. The only way to fix it is to put them through a hostile cross with realistic silences built in — until they've felt the urge to fill silence three or four times and chosen not to.
4. Hindsight-bias confessions
The cross sets up: "Doctor, you reviewed all the records before forming your opinion?" → "Yes." → "Including the autopsy report?" → "Yes." → "So you knew exactly what the final diagnosis was when you reviewed the ED chart?" The trap: any expert who answers "yes" to that last question has just admitted that their methodology let them work backward from the known outcome.
The cross technique: Methodological hindsight bias is a documented Daubert vulnerability. Newman-Toker and the diagnostic-error literature have established the canon. A defense expert challenge can argue your expert's methodology is a 702(c) failure — the reasoning works only because the answer was already known.
Prep fix: Your expert needs a documented review protocol that explicitly states they evaluated the presentation evidence without reference to outcome data — and they need to be able to articulate it. "I reviewed the ED chart and formed my standard-of-care opinion before reviewing the autopsy report. The autopsy was used only to verify the actual cause of death, not to reconstruct what should have been seen at presentation."
5. The "I would have done X" trap
Defense asks: "If you had been the treating physician that night, you would have ordered the head CT, correct?" Most experts say yes — they're proud of their judgment. They've just turned a standard-of-care opinion into personal practice testimony. The standard of care isn't "what this expert would have done." It's what a reasonably prudent physician in the same circumstances would have done.
The cross technique: Once the witness confirms "I would have ordered the CT," defense can argue at trial that the expert is holding the defendant to a personal-best standard, not the reasonable-prudence standard.
Prep fix: "Doctor, the question isn't what I would have done — it's what the standard of care requires of a reasonably prudent emergency physician. The standard of care here required documentation of either the HEART score risk stratification or a clearly articulated clinical reasoning for omitting it. That standard wasn't met."
6. Failure to anchor opinion to records
The expert opines that "the standard of care required earlier intervention." Defense asks: "Which specific records support that opinion?" The expert can't cite Bates. The opinion becomes vulnerable — under FRE 702(b), the testimony must be based on sufficient facts or data. An opinion the expert can't trace to specific record evidence is a methodological failure.
The cross technique: "Doctor, please point me to the Bates page where the trigger for earlier intervention should have been recognized." Silence on this question is more damaging than any wrong answer.
Prep fix: Every opinion in the expert report should be anchored to specific Bates-numbered records. Your expert should be able to recite the Bates anchors for their five key opinions in their sleep.
How to drill these six modes before a real deposition
Reading this list is not the prep. The prep is putting your expert through a realistic hostile cross where these six failure modes get tested — repeatedly, with delayed feedback, against records from your actual case.
We built an AI deposition trainer that does exactly this. Upload the case records, pick a witness type, and the AI plays a hostile defense attorney trained on the canon for the relevant specialty. After each turn, the system shows which of these six modes the witness fell into and the specific prep correction.
Try the 2-minute demo — no signup
Sample case (missed SAH on thunderclap headache). Three scripted turns. Click "show coaching" to see the ideal response and the trap each question sets.
Try the 2-minute demo →The full tool runs on your actual case files. The 2-minute demo runs on a sample case so you can feel the loop in 90 seconds before you commit to anything.
Bottom line
Defense bar isn't winning medmal cases on better facts. They're winning on better witness preparation. The plaintiff bar's expert prep has been roughly the same for thirty years — read the case file, talk through the opinions, hope. The defense bar has been buying enterprise witness-prep tools for two years.
If you're a plaintiff medmal attorney spending $4,500-$12,000 to retain an expert witness for trial, the question isn't whether you can afford to prep them better. The question is whether you can afford not to.
— John Mahoney, Founder, MedLegal AI