How many conflict deaths will there be in Sudan in 2025?
Jeremy’s most recent version of his Multi-AI Oracle is a side competitor with the machine forecasting bots of the VIEWS competition.
Here’s what Jeremy‘s latest version of his Multi-AI Oracle predicts: (Carolyn’s note: Both Jeremy and I agree, these probabilities are, so far, just an experiment. It is too early to know how good these will turn out to be)
* Less than 1000: 1%
* Between 1000 and 3000: 5%
* Between 3000 and 5000: 7%
* Between 5000 and 8000: 15%
* Between 8000 and 12000: 24%
* More than 12000: 48%
Obtaining outside data.
Checking internet sources, using keywords: Sudan conflict 2025
Results from around the web:
Okay, let me tackle this query about the death toll in Sudan’s conflict. The user wants current web results summarized, including tables if there’s relevant data. First, I need to go through each search result provided to extract accurate information.
Starting with [1], Science.org mentions estimates ranging from 20,000 to 150,000 deaths, with many dying from starvation and disease. The Sudanese government isn’t cooperative with data collection, so numbers are uncertain.
ACLED in [2] reports 28,700 fatalities by November 2024, including 7,500 civilians, but they note this is an underestimate. The BBC study cited there suggests up to 150,000. The CFR tracker in [4] echoes the 150,000 figure from the former U.S. envoy and mentions over 14 million displaced. Human Rights Watch [3] and NRC [5] both discuss the humanitarian crisis but don’t add new death toll numbers.
Now, the user asked for tables. The ACLED data from [2] has specific numbers: 28,700 total deaths, 7,500 civilians. The CFR [4] mentions 150,000. The Science.org article [1] gives a range. I can create a table comparing these sources.
Since there’s no time series data provided in the search results—only different estimates from various times—I can’t make a period-based table with changes, means, etc. The user also wanted data from prediction sites, but none of the sources mention Kalshi, Metaculus, or Polymarket, so I can’t include that.
I should structure the answer with a summary of the death toll estimates, causes, and the humanitarian impact. Then present the table comparing sources. Note the lack of time series data and prediction site info. Ensure all citations are correctly placed, using the source numbers provided. Make sure to mention the displacement figures as part of the crisis context. Also, highlight that the numbers are uncertain due to restricted access and underreporting.
The ongoing civil war in Sudan, which began in April 2023, has resulted in catastrophic human losses and widespread suffering. Estimates of total deaths vary significantly due to restricted data access and underreporting, with sources citing figures ranging from **28,700 documented fatalities** to **150,000 estimated deaths** when accounting for indirect causes like starvation and disease[1][2][4]. Below is a synthesis of key findings and data:
—
### **Documented and Estimated Death Toll**
—
### **Key Drivers of Mortality**
1. **Direct Violence**
Both the Sudanese Armed Forces (SAF) and Rapid Support Forces (RSF) have committed atrocities, including mass killings, torture, and targeted attacks on civilians[1][3]. The RSF faces genocide allegations for targeting non-Arab ethnic groups[1].
2. **Indirect Causes**
– **Starvation**: Blockades of humanitarian aid have left 30 million dependent on assistance[1][4].
– **Disease**: Destroyed healthcare infrastructure and lack of clean water exacerbate outbreaks[1][5].
3. **Displacement Crisis**
Over 14 million people displaced (30% of Sudan’s population), creating overcrowded refugee camps in Chad, Ethiopia, and South Sudan[2][4].
—
### **Humanitarian and Regional Impact**
– **30 million people** require urgent aid, with the World Food Programme warning of a looming famine[4][5].
– The conflict has destabilized neighboring countries, straining resources and risking regional spillover[4].
—
### **Data Limitations**
No time-series data or market predictions (e.g., Kalshi, Polymarket) were identified in the sources. Death toll figures remain speculative due to:
– Hostility toward independent data collection[1].
– Displacement of aid workers and local researchers[2].
– Underreporting of indirect deaths[4].
This crisis underscores the urgent need for international intervention to address both immediate humanitarian needs and long-term political resolution.
OpenAI gives a base rate of probability of exceeding 8000 deaths (Carolyn’s note: Open AI failed to return a value.)
The question difficulty is rated 8 (0 to 10)
Sufficient news to provide a good forecast? 1 (0 or 1)
News is relevant, topical and unbiased? 1 (0 or 1)
Question classification: scenario_based_forecast
Expected time (days) until question close: 90
(Carolyn’s note: the correct answer is 254.)
The following were considered in order to (fail to) produce this base rate:
The base rate considers multiple sources reporting varying estimates on death tolls in Sudan from direct violence and indirect causes like disease and starvation. Historical data suggests a significant likelihood due to the scale and duration of the conflict.
Ideally, the news feed would contain the following sorts of information for a better forecast:
Additional information on the Sudanese government’s stance, real-time humanitarian interventions, and a more granular breakdown of casualty reports would significantly enhance forecasting accuracy. A consistent stream of data from non-partisan international observers would be invaluable.
Some potential divergent considerations that might affect the base rate:
The lack of accessibility to accurate data and underreporting could result in actual figures being either under- or over-represented in statistical models. Additionally, future geo-political events could either exacerbate or alleviate the current conditions, affecting the final outcome.
The following chain of events are necessary for the question to resolve positively:
– Continued access to independent data within Sudan despite ongoing conflict. Unlikely
– Sustained international focus and humanitarian aid reaching affected regions. Moderately likely
– Complete reporting from on-ground agencies regarding direct and indirect deaths. Unlikely
Querying Claude (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.02, “Between 3000 and 5000”: 0.05, “Between 5000 and 8000”: 0.12, “Between 8000 and 12000”: 0.25, “More than 12000”: 0.55] – confidence: 6)
Querying Mistral (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.05, “Between 3000 and 5000”: 0.1, “Between 5000 and 8000”: 0.15, “Between 8000 and 12000”: 0.25, “More than 12000”: 0.44] – confidence: 6)
Querying OpenAI (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.05, “Between 3000 and 5000”: 0.07, “Between 5000 and 8000”: 0.15, “Between 8000 and 12000”: 0.22, “More than 12000”: 0.5] – confidence: 6)
Question Type: Multiple Choice
Confidence: 6
MAPD: Avg: 0.027777
777777778, Max: 0.073333333333333
# LLM responses: 3
Model value:
* Less than 1000: 1%
* Between 1000 and 3000: 5%
* Between 3000 and 5000: 7%
* Between 5000 and 8000: 15%
* Between 8000 and 12000: 24%
* More than 12000: 48%
The conflict in Sudan, which began in April 2023, has resulted in significant fatalities, with estimates ranging from 28,700 to 150,000 deaths, driven by direct violence and indirect causes like starvation and disease. The displacement crisis and restricted access hinder accurate data collection, contributing to varied estimates. Despite challenges, historical trends and current intensity suggest death counts could exceed 12,000 in 2025. However, predictions may change dramatically depending on potential developments such as peace agreements, improved humanitarian efforts, or increased international intervention, which might reduce mortality rates. The reliability of current estimates is also questioned due to potential double-counting and underreporting, making the actual situation potentially better or worse than depicted.
Past forecasts on Sudan’s conflict fatalities —>
Guide to Multi-AI Oracle’s Reports (some of these only apply to older versions)
MAPD: Mean Absolute Percentage Deviation calculates the average percentage difference between actual values and predicted values, providing a relative measure of error.
Median: We collect the output values from the LLMs in an array. We then take the median value. This is between 0 and 1 (i.e. 0.15 = 15%).
Base rate: Currently we take OpenAI’s value as the base rate. We are working on a better value using a different prompt that takes into account process. The base rate helps to determine if the median makes sense.
SD: The standard deviation between the median and the base rate.
Confidence: We query each of the LLMs on how confident they are of their predictions (between 0 and 10) and take a median. Because the LLMs tend to be overconfident, we take anything lower than 6 as being low confidence. This factors into the overall model.
Conf Mode: Based on the confidence value. >=9 is high confidence. Below 6 is low confidence (this is also triggered by an exceptionally high SD).
Mellers: This refers to Barbara Mellers, specifically a paper she wrote that includes a formula for moving values towards an extreme (i.e. 0 or 1).
Reverse Mellers: This uses the formula from above, but with a sub-1 coefficient to move the values closer to 50%.
Theory of Mind: We ask the LLMs what they think other LLMs would predict. We hope that this makes them consider the questions more deeply.
Beta Distribution: Currently unused, but possibly of interest. This is based on the median, the base rate and the SD.
Close Type: We noticed that the appropriate base case for some questions is closer to the extremes, while others are closer to 50%. When we have a low confidence value, this helps us to determine whether to extremize or de-extremize the value. ‘A’ implies closer to zero. ‘B’ implies closer to 50%. ‘C’ implies closer to 100%.
# LLM Responses: The count of LLMs that responded. We query 5 currently, but they have a relatively high API failure rate.
Model value: For normal confidence mode, this is the median. For high confidence, this is the Mellers value. For low confidence, this is either the Mellers or Revers
Background on VIEWS: Near the end of 2024, the latest forecasting competition of the Violence and Impacts Early-Warning System (VIEWS) was launched. On Dec. 4, 2024, during a telecon, a group of machine forecasting experts involved in VIEWS observed Jeremy’s demonstration of his Multi-AI Oracle. Although it was too late for Jeremy to sign up as a VIEWS competitor, they urged him to run a side competition for everyone’s benefit. Jeremy chose their Sudan conflict deaths question.
Background on the Sudan conflict, as evaluated by the team of Jeremy Lichtman and Carolyn Meinel:
(1) The importance of the Sahel, a region in Africa that includes Sudan:
Why the Sahel Actually Matters (Possibly To You, Too)
(2) April 29, 2024: Today, a Russian diplomat is negotiating gold mining with the Sudanese Minister of Minerals.
(3) June 3, 2024: Today the African Union Announced Plans to Mediate in Sudan’s Civil War
(4) Sept. 11, 2024 Over 100K starvation deaths 96% likely in Sudan by 1 Jan 2025
Below, map of control in Sudan’s civil war as of April 15, 2025.
Below, before and after pictures of the devastation of the civil war in Sudan.
Carolyn has checked Sudan’s news sites below to ensure that they are still in operation. However, given the widespread violence in Sudan, any of these could go offline without warning.
https://www.alnilin.com/
https://www.alrakoba.net/
https://www.altaghyeer.info/ar/
https://www.altaghyeer.info/en/
https://www.alttahrer.com/
https://www.alzaeemsd.com/
https://www.assayha.net
https://www.cover-sd.com/
https://kushnews.net/
https://rakobanews.com
https://www.sudanakhbar.com/
https://sudanjem.com/
https://suna-sd.net/en
https://sudantimes.net/
https://www.sudantribune.net/
https://www.sudantribune.com/
https://sudaneseonline.com/
https://www.sudanile.com/
https://sudanow-magazine.net/
https://www.sudaress.com/
https://www.tagpress.net/
