How we got our poll results and interpreted them

BernieWorks.com launched with data from 100 matchups between Democratic and GOP presidential candidates. Our source: RealClearPolitics.com, which aggregates poll results.

We extracted matchups from voter surveys published there between the last day of February and the last day of March, 2016. On our Data page you can view the pollsters and their findings.

We included every poll that pitted both Sanders and Clinton against at least one GOP candidate. We call these "paired matchups."

The paired Sanders and Clinton results came from same survey, polled on the same day(s), conducted by the same pollster, in the same locale(s), using the same methodology and demographic sampling.

Thus our 100 results consist of 50 matched pairs. (The round numbers were coincidental.)

Every result you see on this site meets our "paired matchup" standard. We believe it's the most even-handed way to compare Sanders' and Clinton's likely electoral performance against GOP candidates.

We keep in mind that polls can be wildly inaccurate. What, then, was the point of this study?

     1. Reliable or not, many people pay attention to poll results. During our March 2016 study period a new poll was published an average of once every couple of days. There must be an audience for these voter surveys, or pollsters would not invest substantial time and money in them.

     2. We are not looking at individual poll results but seeking patterns in 50 matched  pairs. If there is any reality to be discovered, it's in a consistent pattern among paired matchups.

     3. Yes, we sought (and found) results that aligned with our hopes. We do want to see Bernie Sanders inaugurated as the next President of the United States on January 20, 2017. We celebrate his progressive policies and admire his integrity and 30-year record. We  believe he will brook no nonsense with plutocrats and will do what he can for ordinary folks.

That is why we were especially thrilled at the pattern we found in the data. Beyond Sanders being the nation's best leader, this pattern says he is also the most electable.

Before launching this study, we knew Sanders was outperforming Clinton in poll matchups against the GOP. Robert Reich pointed it out in a January 2016 video. But we were amazed by the strength of Sanders' consistent performance in these matchups. We were also surprised at Clinton's relative weakness.

If Trump gets the GOP nomination — far from a given at this early-April point — the evidence indicates Clinton would defeat him but by a less commanding margin than would Sanders.

But if Cruz or Kasich get the GOP nod, Clinton could be in trouble. That's what the March polling pattern shows. Clinton failed to win a third of her 50 matchups.

This means Clinton is a liability. With the stakes so high, It's a risk we cannot afford.

Sanders, by contrast, won 90 percent of his matchups and often by wider margins than Clinton. The few matchups he did lose were by smaller margins than those of Clinton.

So there you have it: we're partisans who sought to paint a fair picture of reality (such as it may be), and were fortunate enough to discover a remarkably favorable outcome.

If you live in an upcoming primary state, please get out and vote for Sanders. He's not only the best candidate, but also the most electable.

The evidence speaks loud and clear: Sanders is our safest bet against the GOP.