Midrange: Central-Value Estimation, Outlier Sensitivity, and Petroleum Data Analysis in the WCSB
The midrange is a simple measure of central tendency defined as the arithmetic mean of the lowest and highest values in a data set: midrange equals the minimum plus the maximum, all divided by two. It is the midpoint of the data's range, and because it uses only the two most extreme observations it is the cheapest central-value statistic to compute by hand. In oilfield and reservoir work the midrange shows up wherever an engineer or geologist wants a quick central estimate from a small set of readings without running a full average, for example bracketing a range of core porosity values, gas analyses, drilling-rate samples, or daily price marks. Its great weakness, and the reason statisticians treat it with caution, is that it is maximally sensitive to outliers: a single anomalous reading at either end of the data set pulls the midrange directly toward that extreme, because the extremes are the only inputs it considers. This contrasts sharply with the median, which ignores extreme values entirely, and with the arithmetic mean, which dilutes the influence of any one outlier across all observations. For a clean, roughly symmetric data set the midrange, mean, and median converge on nearly the same number, and the midrange becomes a fast and acceptable proxy. For a skewed data set, or one with a measurement spike, the midrange can mislead badly. In the Western Canadian Sedimentary Basin this matters in practical ways. A petrophysicist logging the Cardium at Pembina might see core porosities ranging from 8 to 18 percent across a cored interval; the midrange of 13 percent is a fast bracket, but if one washed-out sample reads 28 percent the midrange jumps to 18 percent and overstates the true storage capacity, where a median would have held near 13 percent. Similarly, a drilling engineer comparing rate-of-penetration samples in the Montney, or a market desk averaging the high and low of a day's WTI or AECO marks, gets a defensible quick number from the midrange only when the data is well-behaved. The midrange is also the natural centre point of a uniform distribution and is used as a quick estimator of the population mean for data known to be uniformly distributed, a property exploited in some quality-control and instrument-calibration contexts. Because reservoir and production data are frequently log-normal or skewed rather than uniform, the responsible practice in WCSB technical work is to report the midrange only alongside the median and mean, never as a standalone central value, so the reader can judge whether the data's extremes are trustworthy. Used that way, the midrange is a legitimate and informative first look; used alone on skewed petroleum data, it is one of the easiest statistics to fool.
Key Takeaways
- Definition and Formula: The midrange is the mean of the smallest and largest values, computed as (minimum plus maximum) divided by two. It is the geometric midpoint of the data's range and the single fastest measure of central tendency to calculate, requiring only the two most extreme observations rather than the full data set.
- Maximally Outlier-Sensitive: Because it depends only on the extremes, the midrange is the least robust common central statistic. One anomalous high or low reading shifts it directly toward that value, unlike the median, which ignores extremes, or the mean, which spreads an outlier's influence across all points.
- Best on Symmetric Data: For clean, roughly symmetric or uniformly distributed data the midrange, mean, and median nearly coincide, making the midrange a valid fast proxy. It is also the natural mean estimator for a uniform distribution, useful in some instrument-calibration and quality-control settings.
- Risk on Skewed Petroleum Data: Reservoir porosity, permeability, and production-rate data are often log-normal or skewed, so a single washed-out core plug or a measurement spike can make the midrange overstate or understate the true central value, leading to errors in storage-capacity or deliverability estimates.
- Report It in Context: Sound WCSB practice presents the midrange only beside the median and mean, never as a standalone figure. The spread between the three immediately signals whether the data's extremes are reliable and whether the midrange can be trusted for that particular sample.
Midrange Versus Median and Mean in Core Analysis
Consider eight Cardium core porosity readings at Pembina: 9, 11, 12, 13, 13, 14, 16, and 17 percent. The minimum is 9 and the maximum is 17, so the midrange is 13 percent, matching the median of 13 and the mean of 13.1. The three statistics agree because the data is well-behaved, and the midrange is a perfectly good quick estimate. Now replace the 17 with a washed-out 29 percent reading: the midrange jumps to 19 percent while the median holds at 13 and the mean rises only to 12.6. That single contaminated plug has pulled the midrange six points off the true centre, a difference that, applied across a 12 m net-pay interval, would materially overstate original gas or oil in place.
Where the Midrange Is Still Useful
Despite its fragility the midrange earns its place in fast field calculations. Daily price desks quote a midpoint of the high and low for WTI, WCS, or AECO benchmarks as a quick session reference. Instrument technicians calibrating a pressure gauge or a gas chromatograph against known high and low standards use the midrange as the expected mean of a uniform response. Drilling supervisors bracketing rate-of-penetration across a homogeneous formation can read the midrange in seconds. In every case the justification is the same: the data is either symmetric, uniform, or being used only as a rough bracket, so the midrange's outlier sensitivity does not bite.
Fast Facts
The midrange is the oldest and simplest of the central-tendency statistics, predating formal statistics entirely; sailors and surveyors averaged a high and low reading long before the mean was formalised. In modern statistics it survives mainly as a teaching tool and as the optimal mean estimator for the uniform distribution, where it is provably more efficient than the sample mean. For nearly every real petroleum data set, which is skewed rather than uniform, that efficiency vanishes and the median becomes the safer choice.
Related Terms
The midrange sits within the family of central-tendency measures used throughout reservoir and production analysis. The median is its robust counterpart, immune to the extreme values that distort the midrange, while the arithmetic mean balances every observation and is the most common average in volumetric work. Understanding how all three respond to outliers connects directly to standard deviation, which quantifies the spread that makes the midrange unreliable in the first place.
Real-World WCSB Scenario: Bracketing Gas Analyses Across a Montney Pad
An ARC Resources reservoir engineer reviewing inlet gas analyses from six Montney wells on a single pad near Dawson Creek sees CO2 readings of 1.1, 1.2, 1.3, 1.2, 1.4, and 1.2 mole percent. The midrange of 1.25 percent, the mean of 1.23, and the median of 1.2 all agree, so the engineer confidently uses the midrange as a quick design basis for the amine unit sizing, saving a full statistical workup for the formal facility application under AER Directive 056.
A week later a seventh well returns an anomalous 3.8 percent CO2 from a different fault block. The midrange leaps to 2.45 percent, double the true central value, which would have oversized the treating plant by a wide margin at a cost of hundreds of thousands of CAD. The engineer instead reports the median and flags the high reading for re-sampling, a textbook illustration of why the midrange must never travel alone.