Sudan updates from AI Botmaster Jeremy Lichtman of April 17, 2025

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: 4%
* Between 3000 and 5000: 15%
* Between 5000 and 8000: 30%
* Between 8000 and 12000: 30%
* More than 12000: 20%

Obtaining outside data.
Checking internet sources, using keywords: Sudan conflict 2025
Results from around the web:
{think}

Okay, I need to tackle the user’s query about Sudan conflict deaths in 2025 using the provided search results. Let’s start by understanding what’s being asked. The user wants current web results on the topic, including news sites, Wikipedia pages, official party websites, and prediction markets. They also want a summary and any relevant data tables, including time series with statistical measures.

First, I’ll look through the search results to find the most recent and relevant information. The search results go up to April 2025, which is the current date mentioned (April 17, 2025). Results 1 through 6 and 9 seem relevant to the Sudan conflict, while 7, 8, and 10 discuss other topics like LLM benchmarks, South Sudan’s civil war, and the Darfur conflict, so I’ll focus on the first six and ninth.

Result 1 from ACLED (December 2024) mentions over 28,700 reported fatalities by November 2024, including 7,500 civilians. It also notes some estimates as high as 150,000. Result 2 (February 2025) from Science.org corroborates ACLED’s numbers and adds that many deaths are from starvation and disease. Result 3 (CFR, February 2025) cites a former U.S. envoy’s estimate of 150,000 killed since the war began. Result 4 (NRC, April 2025) states over 150,000 deaths. Result 5 (Genocide Watch, January 2025) cites LeMonde’s estimate of at least 150,000 civilian deaths and specific figures for Khartoum. Result 6 (Wikipedia, as of April 2025) lists 12,000+ killed but seems outdated compared to others.

I need to compile these figures, noting discrepancies. The ACLED and Science.org reports are more conservative (28,700 intentional deaths) versus higher estimates from NGOs and the U.S. envoy (up to 150,000). The time series part is tricky because the data isn’t presented in periodic intervals. However, some results mention periods like April 2023 to June 2024 (Result 2’s model) and up to November 2024 (Result 1).

For tables, I can create one summarizing death estimates from different sources and another showing time-specific data if possible. Since the user mentioned prediction markets, but none of the search results include data from Kalshi, Metaculus, or Polymarket, I’ll note that no relevant market data was found.

I should structure the answer with a summary of the key figures, a table of estimates, and another table for time series if applicable. The time series might be challenging because the sources don’t provide monthly or quarterly breakdowns, but I can use the periods mentioned in the studies. Also, calculate statistical measures like mean, median, etc., for the available data points.

Check for any mentions of 2025 specifically. Result 3 mentions events in February 2025, but the death tolls are cumulative up to that point. The conflict is ongoing, so 2025 data is part of the continued reporting.
Ensure all citations are correctly referenced with their respective result numbers. Note that some results might be overlapping or from the same source but different articles. For example, Result 1 and 2 both cite ACLED’s 28,700 figure but add different contexts.

Finally, format the answer according to the guidelines: no headers at the start, use markdown for tables, cite sources after each relevant sentence, and avoid any unsupported claims. Since the user requested market prediction sites but none were found in the results, mention that explicitly. The Sudan conflict has resulted in catastrophic human losses, with death toll estimates varying widely due to challenges in data collection and the complex nature of casualties from both direct violence and war-related deprivation. Here’s a synthesis of current reports and data:

### Key Estimates of Conflict Deaths (2023–2025)

### Critical Observations:
– **Discrepancies in counts**: ACLED’s conservative tally (28,700 direct deaths) contrasts sharply with NGO estimates (up to 150,000 total deaths)[1][3][5], reflecting differences in methodology and access to conflict zones.
– **Indirect deaths dominate**: Starvation, disease outbreaks (e.g., cholera, malaria), and lack of healthcare account for 70–80% of fatalities[2][5].
– **Ethnic targeting**: The RSF is implicated in genocidal violence against non-Arab groups in Darfur, with 10,000–15,000 Masalit killed in El Geneina alone[3][5].

### Geographic Breakdown (2024–2025)
– **Khartoum**: 61,000 total deaths[5]
– **Darfur**: 15,000+ direct deaths[3][5]
– **El Fasher**: 782+ civilians killed in siege warfare since September 2024[5]

### Prediction Market Data
No relevant forecasts from platforms like Kalshi or Polymarket were identified in the search results. Most mortality projections rely on epidemiological models and satellite-derived grave counts[2][6].

The lack of humanitarian access and systematic data collection has led to significant underreporting, particularly in RSF-controlled areas[2][5]. With the SAF and RSF continuing to block aid, the UN warns of impending famine affecting 30 million people[3][4].

OpenAI gives a base rate of an ongoing death rate of 1,500-2,000 per month
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: 258

The following were considered in order to produce this base rate: The base rate is derived from past conflicts of similar nature and intensity levels, considering reports from credible sources such as ACLED, NRC, and other aligned forecasts, which provide an estimated 150,000 deaths from both direct and indirect causes. Consistent reports support this number, given continuing conditions minimally altered from their previous state.

Ideally, the news feed would contain the following sorts of information for a better forecast: Additional data on the current activity of international humanitarian aid operations, potential ceasefire negotiations, and more detailed regional casualty reports would help refine the forecast.

Some potential divergent considerations that might affect the base rate: Possible de-escalation due to international diplomatic interventions or successful humanitarian responses could lower the death toll from predictions, though currently, there is limited evidence to strongly support this scenario.

The following chain of events are necessary for the question to resolve positively:
– Continuation of the conflict with high-intensity clashes Highly likely
– Worsening humanitarian conditions and access limitations Very likely
– Increased displacement and population vulnerabilities Highly likely

Querying Claude (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.05, “Between 3000 and 5000”: 0.15, “Between 5000 and 8000”: 0.35, “Between 8000 and 12000”: 0.3, “More than 12000”: 0.14] – confidence: 6)

Querying Mistral (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.04, “Between 3000 and 5000”: 0.15, “Between 5000 and 8000”: 0.3, “Between 8000 and 12000”: 0.3, “More than 12000”: 0.2] – confidence: 6)

Querying OpenAI (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.01, “Between 3000 and 5000”: 0.01, “Between 5000 and 8000”: 0.02, “Between 8000 and 12000”: 0.05, “More than 12000”: 0.9] – confidence: 8)

Question Type: Multiple Choice
Confidence: 6
MAPD: Avg: 0.16888888888889, Max: 0.50666666666667
# LLM responses: 3

Model value:
* Less than 1000: 1%
* Between 1000 and 3000: 4%
* Between 3000 and 5000: 15%
* Between 5000 and 8000: 30%
* Between 8000 and 12000: 30%
* More than 12000: 20%

The analyses from various AIs converge on the prediction that the ongoing conflict in Sudan, particularly intense in regions like Khartoum and Darfur, will lead to a high death toll in 2025. Based on data indicating approximately 28,700 direct conflict deaths by November 2024 and an ongoing death rate of 1,500-2,000 per month, the expected annual rate for 2025 stands between 7,000 and 10,000. However, sources like ACLED and NRC highlight significantly higher estimates, considering direct and indirect causes, suggesting potential totals exceeding 150,000 deaths. The predictions are contingent on several volatile factors such as the potential for peace agreements, international intervention, and seasonal fighting variations, as well as challenges in data accuracy. Any major change in these areas, such as successful diplomatic efforts or improved humanitarian access, could significantly alter the anticipated outcomes.

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.

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/