Courts do not require every expert to put a precise number on every opinion. But under Rule 702 and Daubert, they often do require more than a conclusion when the subject matter is measurable and the discipline normally supports estimation. The real issue is not “math for math’s sake.” It is whether the opinion is reliable, tied to the facts, and presented with enough methodology to help the factfinder.
The core question
A useful way to frame the issue is this: if a qualified expert in this field could reasonably estimate the point at issue, a court may view a purely qualitative opinion as incomplete, speculative, or ipse dixit.
That does not always mean a single figure. Depending on the context, courts may accept:
- A reasonable range
- Scenario analysis
- Percentages of allocation
- Dose reconstruction
- Statistical sampling
- Sensitivity testing
- Probability estimates supported by accepted methods
Just as important, some opinions remain admissible in qualitative form where precision is not possible or the field does not operate through exact quantification.
When quantification is most often required
Certain categories draw closer scrutiny because they naturally involve measurable questions.
Damages
Damages opinions are a common target. Courts are more likely to expect calculations when an expert addresses:
- Lost profits
- Reasonable royalty
- Diminution in value
- Future economic loss
- Business interruption
- Cost of repair or remediation
A damages expert usually needs to show the data used, assumptions made, and calculations performed. If key assumptions drive the number, courts often expect some explanation of how those assumptions were selected and whether alternative scenarios were tested.
A conclusory statement that a business “lost significant value” or that a royalty rate is “reasonable” may not survive challenge if the expert could have modeled the issue and chose not to.
Causation, exposure, and dose
In toxic tort, product liability, and some medical causation disputes, quantification questions often arise around:
- Exposure level
- Dose
- Duration and frequency
- Relative contribution of multiple causes
- Probability of future harm
Courts vary by jurisdiction and subject matter, but many distinguish between general causation and specific causation. An expert may be allowed to testify qualitatively that an agent is capable of causing a condition, yet be limited or excluded on specific causation if the opinion does not address exposure, dose, or alternative causes with sufficient rigor.
Differential diagnosis or etiology can still be accepted, but it is more vulnerable where measurable exposure evidence is feasible and the expert does not engage with it.
Apportionment and allocation
When an expert seeks to divide responsibility among multiple causes, actors, or periods of loss, courts often expect some defensible method of allocation.
That may take the form of:
- Percentage apportionment
- Time-based allocation
- Market-share or benchmark comparisons
- Statistical or actuarial modeling
If the expert assigns responsibility without explaining how the allocation was derived, the opinion can look arbitrary even if the overall theory is sound.
When qualitative testimony may be enough
Courts do not always demand numbers. Qualitative opinions may be acceptable when:
- The discipline is inherently observational or clinical
- The available data do not permit a responsible estimate
- The expert clearly explains the methodological limits
- The opinion does not pretend to offer precision it cannot support
- The testimony is framed narrowly and tied to accepted professional practice
In those situations, over-precision can be its own problem. Courts are wary of false accuracy just as they are wary of unsupported generalities.
The safer course is often to quantify where reliable methods permit it and to explain uncertainty candidly where they do not.
What “enough math” looks like
The best reports do not confuse quantification with certainty. They show the work and define the bounds.
A defensible quantified opinion often includes:
- The question being measured
- The methodology used
- The underlying data sources
- Key assumptions
- Any benchmarks or comparators
- Calculations or modeling steps
- Validation or cross-checking
- Sensitivity analysis or alternative scenarios
- Limits, error sources, or uncertainty
In practice, a range is often stronger than an unsupported single figure. A bounded estimate can show methodological discipline without overstating precision.
Common failure points
Courts are especially critical when experts:
- Offer conclusions without calculations where calculations are feasible
- Use assumptions without tying them to record evidence
- Ignore contrary data or alternative causes
- Cherry-pick benchmarks
- Fail to disclose the method in the Rule 26 report
- Present a number without explaining how it was derived
- Use a qualitative label to avoid a hard measurement problem
Those failures can lead to full exclusion, limitation of the opinion, or an order confining the expert to narrower issues such as general causation only.
Practical takeaway
The question is rarely whether every expert must quantify. It is whether the opinion calls for quantification to be reliable and useful in that case. If the issue is measurable, and accepted methods exist to estimate it, courts often expect more than a bottom-line conclusion.
For litigators, that means building expert reports that show method, data, assumptions, and uncertainty from the outset. In close cases, a well-supported range or scenario analysis is often far better than either false precision or no quantification at all.
