The prediction model takes dozens of polls to estimate each candidate's probability to become Prime Minister.

### Who will be PM?

The model generates thousands of possible Congresses resulting from the election and analyzes the options for each candidate under each composition of Congress. The specific chances for each candidate are debatable, but they rely on a transparent methodology that is consistent across time, making the trend particularly valuable.

### How will Congress look?

*Average composition across 5,000 simulations*

This graph shows the average result from the 5,000 election simulations. These simulations rely on all the published polls and historic data about polling errors and geographic distribution of the vote.

### Polling Average

Here you can check the weighted average of all the polls included in the model and its change over time.

### Composition of Congress

*80% probability bracket for each party*

There is an 80% chance that the final result for each party will fall within its bracket. Each party falls under the bracket in less than 10% of simulations and over the bracket in less than 10%.

### Methodology

- Here I explain where these numbers are coming from.
- Here I give more detail about the first model, which predicts the composition of the Spanish Congress. This model generates the second, third and fourth graphs in this page.
- Here I explain the second model, which predicts the election of the PM. This model generates the first graph in this page.

The links above are all in Spanish.