Two Outside Experts Disagree and You're in the Middle: Decision Tree

Why this matters

Sometimes you are not one of the two people arguing, you are the one called in to settle it. A home inspector says one thing, a manufacturer rep says another, or two prior contractors left opposite diagnoses, and the customer wants you to pick the winner. Picking a side to be agreeable is the trap. If you rubber-stamp the wrong one, you now own it, and the failure lands on your name. This tree keeps you from being the tiebreaker who guessed, and turns you into the one who actually settled it.

Start here: is the disputed system a live hazard

Before you referee anything, check whether the thing they are arguing about is dangerous right now. If the dispute involves gas or combustion, live electrical, a pool or spa bonding and grounding question, a pressure vessel, or compromised structure, the safety call comes before the technical verdict.

• If the system is unsafe to operate, make it safe, isolate and tag it, and document that, regardless of which expert is right about the cause. A disagreement between two other people does not lower your duty to the person standing in front of you.
• If it is safe, move to settling the technical question below.

Do not pick a side, re-derive it yourself

The whole reason you were called is that two opinions cancel out. Adding your vote to one of them does not settle anything, it just makes it two against one. Ignore both conclusions and work the fault from the evidence as if you were the first tech on the job.

• Read the symptoms fresh, without adopting either narrative.
• Take your own measurements, and compare them to spec yourself rather than trusting either report's numbers.
• Reach your own conclusion first, then see which expert it matches, if either. Occasionally both were wrong, and only a fresh look finds that.

Find where the two experts actually diverge

Two experts rarely disagree about everything. Locate the exact fork.

• Do they disagree on the reading, or on what the same reading means? A data disagreement is settled by re-measuring. An interpretation disagreement is settled by spec or by a demonstrated test.
• Did they inspect the same things? Very often one accessed a component the other read from outside, and the "disagreement" is really a difference in what each one looked at.
• Name the single question that separates them, because that is the only thing you actually have to answer.

Name the one test that would settle it

A good tiebreaker does not out-argue anyone. They identify the measurement or teardown that makes the answer undeniable and then run it.

1. Define the deciding test: "If we read this here, one of these two positions is confirmed and the other is not."
2. Run it in front of the customer where possible, so the verdict is shared, not asserted. See related: The Evidence That Settles a Three-Way Technical Dispute.
3. Preserve the evidence the test produces, so your call holds up if either expert contests it later.

When your fresh diagnosis matches neither

Be ready for it. If your independent read lands somewhere neither expert called, say so plainly and back it with the evidence, rather than forcing your finding to fit the closer of the two. "Here's what I actually measured, and it points to a third cause" is the most valuable thing a tiebreaker can deliver, even though it satisfies no one in the moment.

Deliver the verdict without indicting anyone

• State your finding as evidence, not as a ruling against the losing expert. "The reading points here" beats "the first guy was wrong."
• Give the other experts an honest out. "It's a reasonable read from the outside; the cause only shows on teardown" keeps the peace and keeps you the professional. See related: Staying Professional When Another Expert Is Plainly Wrong.
• Put your finding in writing for the customer, with the deciding test and its result, so the settled answer survives the next person who reopens it.

The recap

1. Safety first: make the disputed system safe before you referee it.
2. Do not vote; re-derive the diagnosis from your own measurements.
3. Locate the exact fork between the two experts.
4. Run the one test that settles it, in the open, and preserve the result.
5. Deliver it as evidence, leave the other experts a face-saving out, and document.

References

• Trade-standard practice for independent failure analysis
• Manufacturer documentation on diagnostic tolerances and specified conditions
• See related: The Evidence That Settles a Three-Way Technical Dispute; The Three-Way Diagnostic Disagreement