By M.G. Hill | M G Hill Consulting

The cases that turn rarely do so on new data. They turn on someone asking a different kind of question.

There is a particular kind of silence that falls over a conference room when an expert witness says something no one expected.

Not a dramatic silence. A recalibrating silence — the kind that happens when a frame shifts and everyone in the room feels it before they can articulate why.

I have been in that room many times over thirty years in oil and gas measurement, custody transfer, and pipeline operations. What I have learned is that the cases that turn — the audits that uncover what everyone missed, the depositions that change the trajectory of litigation — rarely turn on new data. They turn on someone asking a different kind of question.

That someone needs two cognitive tools most technical professionals never deliberately develop: analogical reasoning and probabilistic thinking.

This article is about what those tools are, why they matter in the specific world of custody transfer disputes and operational failures, and how an expert who wields them changes the outcome of a case.

---

The Problem With Pure Technical Expertise

Technical expertise is necessary. It is not sufficient.

A measurement engineer who can recite API MPMS Chapter 21 from memory, who knows every nuance of flow computer configuration and meter proving tolerance, is genuinely valuable. But that same engineer, presented with a disputed measurement scenario, will often do what technically trained people always do: look for the data error.

He will look at the meter calibration records. He will check the flow computer configuration. He will examine the pressure and temperature compensation factors. He will run the numbers.

This is correct. This is also incomplete.

Because the question in litigation — and often in a field audit — is never simply what happened. The question is why it happened, who knew what and when, whether the deviation was detectable, and what a reasonable operator should have done differently.

Those are not purely technical questions. They are questions about pattern, probability, and structural analogy. And answering them well requires a different set of tools.

Analogical Reasoning: The Expert as Pattern Recognizer

Analogical reasoning is the ability to recognize that two situations share the same deep structure even when their surface appearance is completely different.

In custody transfer disputes, this capacity is the difference between an expert who describes what happened and an expert who explains why it was predictable.

Consider a common scenario: a midstream operator disputes allocation volumes with a producer, claiming the producer’s fiscal measurement point has been systematically under-reporting. The operator’s expert presents calibration records, meter factors, and deviation analysis. The producer’s expert presents counter-analysis. Both sides have data. Both sides have numbers.

The expert with analogical reasoning asks a different question: Where have I seen this structural pattern before?

And the answer comes: This looks exactly like the systematic drift scenario from a case years prior. Different operator, different meters, different geography — but the same sequence. Intermittent over-range conditions. Uncorrected meter factors carried forward through multiple proving cycles. A flow computer configuration that masked the cumulative error in daily reports while monthly totals diverged.

That structural recognition does three things in a litigation context.

First, it allows the expert to move from description to prediction. If the structure matches, then the deviation didn’t begin when the dispute was filed — it began earlier, at a specific triggering event that the records will confirm if you know where to look.

Second, it reframes the question of operator knowledge. If this pattern is recognizable to any competent measurement professional, then the question “should the operator have known?” has a defensible answer that doesn’t rest solely on what the specific operator claims to have reviewed.

Third, it gives the trier of fact — judge, jury, arbitration panel — a narrative. Not just numbers, but a story with structure. Stories are how human beings process causation. An expert who can say “this is a pattern I have seen before, and here is how it develops” is more persuasive than one who says “the meter factor was outside tolerance.”

Probabilistic Thinking: The Expert as Calibrated Skeptic

Probabilistic thinking is reasoning in degrees of confidence rather than binary certainties.

Most technical disputes are presented in absolutes. The operator claims the measurement was accurate. The producer claims it was not. Each side’s expert states conclusions with the rhetorical confidence of someone who has made a decision rather than rendered a calibrated judgment.

This is intellectually dishonest and strategically weak.

A measurement dispute is not a case of “the meter was right” or “the meter was wrong.” It is a question of: Given the available evidence — calibration history, proving records, operating conditions, configuration documentation — what is the probability that the measurement deviated beyond acceptable tolerance, in which direction, for what duration, and by what magnitude?

An expert who answers probabilistically is more credible, more useful to counsel, and harder to cross-examine than one who states certainties that opposing counsel can chip away.

Here is how probabilistic thinking changes the practice in three specific contexts:

In field audits, the probabilistic thinker doesn’t just flag anomalies — he assesses their significance. A meter factor outside API tolerance on a single proving run is not the same as a systematic drift over twelve consecutive proving cycles. The first may be noise. The second is signal. Distinguishing them requires a calibrated sense of base rates: How often does this kind of deviation occur in normal operations? How often does it indicate something systemic? An auditor who treats every anomaly as equally significant generates noise. One who calibrates significance generates findings that hold up.

In regulatory matters, probabilistic reasoning is the antidote to two common failure modes. The first is the operator who treats any measurement within nominal tolerance as unproblematic, ignoring the cumulative probability that a series of individually acceptable deviations has produced a systematic bias. The second is the regulator or plaintiff’s expert who treats any out-of-tolerance reading as proof of deliberate manipulation, ignoring the base rate of measurement variance in normal operations. Both errors misrepresent reality. Both are correctable with probabilistic discipline.

In litigation preparation, the probabilistic thinker helps counsel understand not just what the evidence shows, but how strong that evidence actually is. There is a significant difference between “the records show the meter factor was outside tolerance on three of the last eight proving runs” and “the records show a statistically significant drift pattern inconsistent with normal meter behavior and consistent with an uncorrected calibration error.” The first is a fact. The second is a probabilistic judgment that tells counsel how hard to push and what additional discovery is worth pursuing.

Where the Two Tools Intersect

The most powerful expert deployments combine both tools simultaneously.

Consider a scenario that appears regularly in midstream operations disputes: a pipeline operator faces a claim that allocation errors over a multi-year period resulted in significant underpayment to a producer. The operator’s position is that the measurement system was properly maintained, regularly proved, and operating within API tolerance throughout the disputed period.

The expert applying analogical reasoning recognizes a structural pattern: the dispute period coincides with a change in flow computer firmware applied across multiple measurement points simultaneously. He has seen this before — not with this operator, not with this firmware version, but with the structural pattern of a system-wide configuration change that affects measurement in ways that don’t appear in individual meter proving records.

The expert applying probabilistic thinking then asks: Given this structural pattern, what is the probability that the firmware change introduced a systematic calculation error that would be invisible to standard proving protocols but visible in aggregate volume reconciliation over time?

The answer, based on base rates from similar scenarios, is: not trivial. Not certain. Not proven by this observation alone. But sufficient to direct discovery toward specific records — the firmware change documentation, the pre- and post-change volume reconciliation data, the flow computer configuration backups — that will either confirm or disconfirm the hypothesis.

This is how the work should be done. Not by arriving at the answer before examining the evidence, but by knowing which questions to ask and which records to demand, because the expert has reasoned structurally and probabilistically about what the evidence should show if the hypothesis is correct.

What This Means for Attorneys and Claims Professionals

If you work in energy litigation or insurance claims involving oil and gas operations, the practical implication is direct:

The technical expert who can only tell you what the numbers say is a commodity. Valuable, but a commodity. The expert who can also tell you what the numbers mean, what pattern they represent, how confident we should be in that interpretation, and what additional evidence would confirm or defeat that interpretation — that expert changes the outcome of cases.

When evaluating expert candidates for measurement disputes, ask two questions beyond the standard credential review:

Can you show me a case where you recognized a pattern in the measurement records that the other side’s expert missed? The answer will tell you whether the candidate reasons analogically or only technically.

When you form an expert opinion, how do you express your degree of confidence, and what evidence would change your conclusion? The answer will tell you whether the candidate thinks probabilistically or in absolutes.

The best measurement expert is not the one with the most impressive resume. It is the one who has spent thirty years building a pattern library and the cognitive discipline to deploy it with calibrated precision under adversarial pressure.

Conclusion: The Premise Auditor

I have written elsewhere about the role of the expert witness as a premise auditor — someone whose primary function is not to confirm the client’s position but to rigorously examine the assumptions on which the measurement dispute rests.

Analogical reasoning and probabilistic thinking are the core instruments of that audit.

The analogical reasoner asks: Have I seen this structure before, and what did it mean when I did?

The probabilistic thinker asks: How confident should I actually be, and what would change that confidence?

Together, they produce something that neither technical expertise nor advocacy can manufacture: an honest, calibrated, structurally informed account of what the measurement evidence actually shows — and what it doesn’t.

In a disputed custody transfer, that is the most valuable thing in the room.

---

ABOUT THE AUTHORM.G. Hill is an independent consultant with 30 years of specialized expertise in oil and gas measurement, custody transfer, pipeline operations, and regulatory compliance. He serves as a technical advisor and expert witness for litigation attorneys and insurance professionals in energy sector disputes. M G Hill Consulting is based in Oklahoma City, Oklahoma.

If this article was useful to you, consider sharing it with colleagues in energy law, midstream operations, or claims management. For consulting inquiries, contact M G Hill Consulting directly.