As Emily Guerin reported, North Dakota releases oil production stats every month. And when they do, they release a snapshot: What was production this month? How much did it increase or decrease from last month?
In this context, it’s easy to forget that real world data – whether it’s oil production, or births, or temperature, or Facebook users – isn’t smooth and pretty. It’s jarring and ugly. It fluctuates.
Here’s what North Dakota’s Bakken formation monthly oil production looks like going back to 2008:
In the past few years, oil production in the Bakken has ballooned from 36,000 barrels a day at the beginning of 2008 to more than a million barrels a day in 2014. Data Source: North Dakota Department of Mineral Resources. View the data.
In this view, it’s pretty easy to spot the trend, right? Oil production in the Bakken has been going up and up and up.
But often, analysts focus on how much monthly production changed from the previous month. You can’t really blame them. As humans – and especially as journalists – we’re always wondering: What’s new??
Here’s what the month-to-month change in production looks like going back to 2008:
Month to month, oil production in the Bakken rises and falls. But the overall trend has been one of rapid growth. For example, in February 2014, monthly production fell by 2 million barrels. The next month, production jumped more than 3 million barrels. Data Source: North Dakota Department of Mineral Resources. View the data.
Another metric oil industry watchers often discuss is the monthly percent growth. That is, what’s the proportional change in production from the month before. Here’s what that looks like over the same time period:
The monthly percent change in oil production is a relative metric. For example, in February 2014 monthly production of Bakken oil dropped 8% from the previous month. But February’s production, nearly 25 million barrels, was higher than every month before June 2013. Data Source: North Dakota Department of Mineral Resources. View the data.
Looking at month-to-month percent change, it’s hard to pull out any trends. What’s more, a one percent increase in production one month doesn’t mean the same thing as a one percent increase in production the next month because the base values will be different.
If we focus on the second and third graphs – which look like cardiograms and only show us the monthly noise – it’s easy to lose sight of the long view, which is captured in the first graph. To help make sense of the jarring ugliness of real-world data, it’s helpful to look at long-term trends, where meaning emerges from the chaos.
