Report: Covid-19 predictions for TTUHSC El Paso

How were these predictions generated?

Metaculus enjoys a highly engaged forecaster community is made up of thousands of subject matter experts. We are currently running a modelling competition that invites our community to submit open-source data-driven and mechanistic models.

Metaculus has an aggregation engine built by our physics professor founders, which cleverly aggregates individual predictions to generate forecasts that are better than the sum of their parts.

These predictions generally will update in real time, as new information emerges. However, as these updates are made manually by our community of forecasters, delays may sometime occur.

How do I interpret these predictions?

Metaculus produces probabilistic predictions. That is, instead of producing a single number, our predictions involve probabilities attached to a range of different outcomes. This enables us to communicate the degree of uncertainty involved in the predictions.

Predictions are displayed on prediction charts. these charts consist of two parts: the probability graph (also known as the probability density), and a "summary part" that presents some key parameters that describe the prediction, including the median and the interquartile range (25th and 75th percentiles).

More details on reading Metaculus prediction charts

Here we provide a brief summary on how to interpret Metaculus' prediction charts.

Charts consist of two parts: the probability graph (also known as the probability density), and a "summary part" that presents some key parameters that describe the prediction.

Our predictions are displayed on prediction charts, which presents a interactive probability distribution, and presents additional information about the prediction.

The median prediction is an important summary statistic. This is the most likely value to be realized, and so can, in some cases be seen as our "best guess".

Our forecasters expect that, with a 25% chance, the value will be realized as anything below this number. Conversely, with a 75% chance, the value will turn out to be at least as great at this.

Our forecasters expect that, with a 75% chance, the value will be realized as anything below this number. Conversely, with a 25% chance, the value will turn out to be at least as great at this.

The probability graph is interactive. By hovering your mouse over this graph, you can focus attention on any specific number (or date) and can examine the probability of the value being realized being at least as large as this number (or date).

Important note: these charts might not properly render on the Firefox web-browser.

Predicted peak-date and number of new infections on the peak day

Predicted date when the number of infections will peak in El Paso County

Probability distribution for the Metaculus prediction of the date when the number of newly reported Covid-19 infections will first peak in El Paso County.

Note that our forecasters might expect the peak to have already passed. Hence, some probability might be assigned to the peak being before today's date.

Predicted number of new reported infections on the day on which the peak occurs

Probability distribution for the Metaculus prediction of the number of newly reported Covid-19 cases in El Paso County, on the day of the peak.

How quickly will the number of newly reported cases decline, after the peak day?

Predicted number of new infections for two periods after the peak

Probability distribution for the Metaculus prediction of the number of the average newly reported Covid-19 cases in El Paso County on the 3rd, 4th and 5th days after the peak (left panel) and the 6th, 7th and 8th days after the peak (right panel).

Required hospital resources

Predicted peak number of in-hospital lab-confirmed COVID-19 patients in El Paso County

Probability distribution for the Metaculus prediction of the number of hospitalized Covid-19 cases in El Paso County, according to reports by the Texas Dept Of State Health Dashboard

Predicted peak number of ICU-admitted confirmed COVID-19 patients in El Paso County

Probability distribution for the Metaculus prediction of the number of Covid-19 cases in El Paso County requiring ICU admission on the day when the number of hospitalizations peak.

Predicted peak number of COVID-19 patients requiring invasive ventilation in El Paso County

Probability distribution for our prediction number of Covid-19 requiring ventilation in El Paso County on the day when the number of hospitalizations peak.

About us

The Metaculus Pandemic Project delivers pro-bono forecasting and modeling resources to public health and policy professionals as they navigate critical decisions that will affect the shape of the epidemic’s curve around the world.

Our highly engaged forecaster community is made up of thousands of subject matter experts, and our mission is to scale our community's ability to rapidly and accurately provide predictions and decision support for frontline organizations’ most important questions.

Metaculus predictions have been able to perform on par with or better than prominent epidemiologist surveys in accurately predicting a variety of critical COVID-19-related outcomes, and is also able to provide fine-grained predictions on specific issues of interest to healthcare and policy professionals.

We are currently participating in a competition with University of Massachusetts Amherst to evaluate our predictions against top infectious disease experts.

Our forecasts are also helping to inform modelling groups at the CDC.