On February 17, 2016, Nate Silver and FiveThirtyEight released a table projecting where Bernie Sanders should stand in opinion polling in all 50 states should he be tied with Hillary Clinton nationally. The data reveals the states that Bernie Sanders must perform well in in order to succeed in taking the Democratic nomination for president.

To begin analyzing the data, we first parsed the existing table into a more machine-readable JSON format that looked roughly like this:

[
	{
		"date": "February 1, 2016 00:00:00 GMT-0800",
		"states": [
			{
				"state": "Iowa",
				"delegates": 52,
				"difference": 0.06,
				"sanders": 0.53,
				"clinton": 0.47
			}
		]
	}, ...
]			

We then computed the expected delegate share per candidate by simply multiplied the percentage support for each candidate by the number of delegates. Note that this relies on a very basic proportional representation system and assumes a statewide viability. We also made the assumption that the national lead or tie will remain constant throughout the primary race, although in reality Bernie Sanders has been trending upward consistently in recent national polls. The data and the scripts used to transform the data into the tabular format required by the Google Charts API are all freely available on GitHub.

In the first visualizisation, we have simply the raw data given that Bernie Sanders is tied nationally. The three charts below show the data projected onto a map of the United States, the data as a timeline of the primaries and caucuses leading up to the nomination, and the same timeline but with the difference in delegates plotted, where a positive difference indicates a Bernie Sanders lead.

The above charts show a scenario in which Bernie Sanders would come out ahead in pledged national delegates by a margin of 20 delegates. Note that this does not include Democrats voting from abroad or from American territories. It is also interesting to note the stark separation between North and South when it comes to support of Bernie Sanders and Hillary Clinton. Even in the case of a national tie, Super Tuesday, and many of the states that vote in early-mid March are expected to go heavily in favor of Hillary Clinton. States that vote from mid-March to mid May are more likely to go Sanders's way.

In the next set of charts, we have given Hillary Clinton a 5% boost in every state to roughly account for what would be a 5% Clinton lead in national polls.

Something as small as a 5% national boost is enough to give Hillary Clinton the nomination by more than 200 delegates. It is important to note, however, that a 5% national boost would not necessarily be distributed equally between each state as our model would.

In our final set of visualizations, we have the Democratic presidential primary race as it stands today, with Hillary Clinton winning Nevada and Iowa and Bernie Sanders winning New Hampshire, leaving Hillary Clinton with a single delegate lead.

As the race stands, Bernie Sanders is closer to the simulation in which Hillary Clinton leads by 5% nationally. What will end up shaping the race is if Bernie Sanders loses Super Tuesday by roughly 50 delegates (putting him in the national tie scenario) or by closer to 100 delegates (putting him in the Clinton 5% lead scenario).