The record 2,928 barrels per rig that came from counting 11 rigs

Pale grey planning slide, masthead “Coteau Ridge Gathering · Capacity planning · Williston Basin” on the left and “2022 expansion case · November 2020” on the right, headlined “Each Bakken rig now brings on 2,928 barrels a day of new oil — nearly double March. We are sizing the 2022 gathering expansion for that rig.” A standfirst says the figure is EIA’s new-well oil production per rig for the Bakken region, from the Drilling Productivity Report: the oil that wells in their first full month of production add each day, divided by the number of rigs drilling, the figure EIA’s model uses to turn a rig count into new production; linear axis from zero, monthly, January 2014 to October 2020. The main panel, “Bakken region · new-well oil production per rig”, is a single teal line with a light fill on a zero-based axis running to 3,000 barrels a day per rig, gridded every 500. It starts at 385 in January 2014, climbs to 1,186 in late 2016, levels off around 1,100 to 1,500, reaches 1,502 at a hollow marker labelled “March 2020 · 1,502”, dips to about 760 in the spring of 2020, then shoots almost vertically to a solid marker labelled “Record · 2,928 b/d · Oct 2020”. Three cards run down the right. “The headline number” prints ×1.95 in large type beside “new oil per rig, March to October 2020” and says from 1,502 to 2,928 b/d in seven months, a record, more than seven times the 385 of January 2014: fewer rigs, each one far better. “Planning read-out” says operators expect rigs to return once prices recover, that at 40 rigs, 40 × 2,928 ≈ 117,000 b/d of new oil would arrive every month, and records the decision to size the 2022 expansion for 117,000 b/d of new-well volume and bring the compressor order forward to Q1. “Basis of measurement” says new-well production per rig is oil from wells in their first full month of production divided by rigs active two months earlier, published values rounded to the barrel. A row of green ticks along the foot reads “Published EIA series”, “Zero-based linear axis”, “Monthly, no smoothing by us”, “Every month plotted” and “One region, one metric”. A footnote credits EIA’s Drilling Productivity Report data file and states that the slide is a demonstration drawn by Misleading Charts in the manner of a midstream planning deck: the company, its expansion and its decision are invented, and every figure is from EIA’s final May 2024 release, not the earlier estimates a planner would have seen in November 2020. — Showing the misleading chart

The same EIA series on a cream background, kicked “The same EIA series · with its numerator and denominator” and headlined “The rigs didn’t get twice as good. There were a fifth as many of them, and the new oil fell by more than half.” A standfirst says a per-rig figure is a fraction and in 2020 its bottom fell much faster than its top: the rigs it divides by, counted two months earlier, dropped from 52 to 11 (−79%) and the new oil implied by EIA’s series fell from 78,100 to 32,200 barrels a day (−59%), so the ratio nearly doubled; it quotes EIA that the rise was “directly related to low rig counts” and “not a direct measurement of each active rig’s productivity”, and says a backlog of wells kept coming online after their rigs had gone. Three aligned line charts share a January 2014 to December 2023 time axis, each linear from zero, with two tan bands marking when the rig count was culled, October 2014 to May 2016 and March to August 2020, and a dashed rust line at October 2020, the month on the deck. Panel A, “Rigs drilling in the Bakken region · the denominator”, falls from 191 in October 2014 to 24 in May 2016, hovers around 50 to 60, drops from 51 in March 2020 to 10 that September, and recovers to 41 by January 2023. Panel B, “New oil from new wells implied by EIA’s series, barrels a day · the numerator · computed by us”, peaks near 93,800 in February 2015, falls to about 25,000 in 2016, climbs back to about 78,100 in March 2020, drops to about 26,000 by August 2020 and is 32,200 in October, marked in rust, and about 65,700 in January 2023. Panel C, “New oil per rig = B ÷ rigs two months earlier, the figure the deck plotted”, climbs during and after the first cull to 1,186 in late 2016; in 2020 it dips from 1,502 to about 760 in the spring, then shoots up to a rust-marked 2,928 in October, and slides back to 1,603 by January 2023. Four cards run down the right. “Same fraction, three months” is a table of rigs two months earlier, new oil and per-rig figure: March 2020, 52, 78,088, 1,502; October 2020, 11, 32,207, 2,928 in rust; January 2023, 41, 65,719, 1,603; beneath it, the deck planned for 40 × 2,928 ≈ 117,000 b/d, and by January 2023 EIA’s own series had the figure back at 1,603 per rig, so 41 rigs imply about 65,700, 56% of the plan, with a note that after 2020 the series is EIA’s modelled trend. “Why the ratio rose” gives two reasons: the numerator lags, because completed wells waiting to produce kept reaching their first full month after the rigs that drilled them had gone; and the cut may not be random, because operators short of money tend to keep rigs on their best acreage, high-grading, which EIA’s figures cannot separate out. “What to plot instead” says to plot both halves of the fraction beside the ratio, and to plan per unit by following the same units through time, like same-store sales: oil per well, by the month it was drilled. “The tell” says the per-rig line moves against its own denominator, up during and just after both culls, in 2020 after a spring dip when wells were shut in, while the new oil fell. A footnote credits EIA’s Drilling Productivity Report data file, final release May 2024, notes that EIA publishes the per-rig figure as a smoothed, and from April 2020 modelled, trend applied to the rig count two months earlier, that panel B is that figure multiplied by the lagged rig count, computed by us and implied by EIA’s series rather than its reported well data, cites the August 2020 DPR Supplement for the quotations, and says both drawings are ours and the gathering company on the first is invented. — Showing the misleading chart

Be the first to star this exhibit

A planning slide we drew plots EIA’s new-well oil production per rig for the Bakken: 1,502 barrels a day in March 2020, a record 2,928 in October. The slide reads it as each rig getting twice as good and sizes a pipeline expansion for 40 such rigs. But the figure is new-well oil divided by the rig count, and the rigs it divided by fell from 52 to 11 while the new oil implied by EIA’s series fell from about 78,100 to 32,200 barrels a day. The ratio rose because the denominator was culled. By January 2023 EIA’s own series had the figure back at 1,603 per rig, so the 41 rigs then working imply about 65,700 barrels a day, 56% of the plan.

U.S. Energy Information Administration, Drilling Productivity Report data file (dpr-data.xlsx, final release May 2024), Bakken region: monthly rig count and new-well oil production per rig, the latter applied by EIA to the rig count two months earlier; new-well oil production reconstructed by us as per-rig figure × lagged rig count. Quotations from EIA, “Rig counts fall but new-well production per rig rise as new-well production persists”, Drilling Productivity Report Supplement, August 2020 (https://www.eia.gov/petroleum/drilling/pdf/August_2020_Supplement.pdf), 2024Added October 31, 2026

01The claim

EIA’s Drilling Productivity Report tracks how much new oil each rig in the Bakken brings on: the oil that wells in their first full month of production add each day, divided by the rigs drilling. In March 2020 it was 1,502 barrels a day per rig. By October it was 2,928, a record and nearly double. The figure is the one EIA’s own model uses to turn rigs into production, so a pipeline company planning for the recovery takes it at face value: when operators bring 40 rigs back, 40 × 2,928 ≈ 117,000 barrels a day of new oil should arrive every month, and the gathering expansion gets sized for it.

02The trick

Every figure on the slide is EIA’s, and the arithmetic is right. What goes wrong is reading a fraction as a property of each rig. New-well oil per rig is new-well oil divided by the rig count two months earlier, and in 2020 the bottom of that fraction fell much faster than the top. The rigs fell from 52 (counted in January, for March) to 11 (August, for October), a 79% cut. New oil from new wells fell too: the volume EIA’s series implies went from about 78,100 to 32,200 barrels a day, down only 59%. Divide the second by the first and the ratio nearly doubles without any rig drilling a better well. EIA said so at the time: the rise was “directly related to low rig counts causing a disproportional rise in new-well production per rig and is not a direct measurement of each active rig’s productivity.” Two things kept the top of the fraction up while the bottom fell. Wells drilled before the cut, and a backlog of completed wells waiting to produce, kept reaching their first full month after the rigs that drilled them had been laid down. And operators may also have kept their remaining rigs on their best acreage, which is called high-grading; EIA’s figures cannot separate that out. Neither effect carries over to the rigs that come back. The 2016 bust had shown the same pattern: the ratio rose from 468 in November 2014 to 1,186 in November 2016 as the rigs it divided by, counted two months earlier, fell from 194 to 28. The path is not a clean mirror, though. In spring 2020 the ratio first dipped to about 760, when wells were shut in and completions stopped, before it surged. (Both drawings are ours; the gathering company and its decision are invented. Every figure is from EIA’s final May 2024 data release, or computed from it in the case of new-well oil, so it is not what a planner would have seen in November 2020.)

03The fix

Draw the fraction apart. The redraw stacks three aligned panels on the same 2014–2024 time axis, all zero-based: rigs, new oil from new wells, and their ratio, with the two culls shaded. The per-rig line rises during and just after the months its own denominator collapses, in 2015–16 and in 2020 (after a brief dip that spring, when wells were shut in), while new oil falls in both. Then it runs the deck’s plan forward on EIA’s own series. By January 2023 that series, a modelled trend after 2020 rather than a fresh measurement, had the per-rig figure back at 1,603. With the 41 rigs then counted, that implies about 65,700 barrels a day, 56% of the 117,000 the slide planned for 40. The general rule for any per-unit figure, whether per rig, per store, per employee or per acre: plot the numerator and the denominator next to the ratio, and be suspicious of a record set during a cut. If you need a per-unit number to plan with, follow the same units through time, the way retailers quote same-store sales. Here that means oil per well, grouped by the month the well was drilled. A ratio measured on whatever was left after a cut does not describe the units that come back.