Botmaster Jeremy Lichtman’s Sudan conflict deaths forecast, July 21, 2025

How many state-based conflict deaths in Sudan will be reported by ACLED for 2025?

Here’s what Jeremy Lichtman’s Multi-AI Oracle predicts today:

Model value:
* Less than 1,000: 1%
* Between 1,000 and 3,000: 5%
* Between 3,000 and 5,000: 12%
* Between 5,000 and 8,000: 25%
* Between 8,000 and 12,000: 35%
* More than 12,000: 22%

11–13 minutes read

Obtaining outside data.
Checking internet sources, using keywords: Sudan conflict deaths ACLED
News from various sources:

The recent reports and studies regarding the ongoing civil war in Sudan reveal a grim picture of the conflict’s human toll. According to research conducted by the London School of Hygiene & Tropical Medicine, deaths have been “substantially underreported,” especially in Khartoum State, where over 61,000 people died of all causes between April 2023 and June 2024. This includes 26,000 deaths directly due to violence, exceeding the 20,178 intentional-injury deaths reported for the entire country by the Armed Conflict Location and Event Data (ACLED) project.

The conflict began in April 2023 between the Sudanese Armed Forces (SAF) and the Rapid Support Forces (RSF), with devastating impacts throughout the country. The violence has led to the deaths of thousands and displaced millions, making it one of the world’s worst humanitarian crises. It is reported that more than 14 million people have been displaced, with profound impacts on children, who face severe malnutrition and increased vulnerability to disease.

ACLED reports an average of 1,200 direct conflict deaths per month, while estimates suggest as many as 150,000 people may have died due to the conflict, considering both direct and indirect causes such as disease and starvation. Various regions, including Darfur and Kordofan, have experienced high levels of violence, with allegations of atrocities and ethnic cleansing.

Obstacles in humanitarian efforts, accusations of using “starvation tactics” by both sides, and international involvement complicate the situation. Recent reports highlight the use of foreign arms, with Amnesty International accusing the UAE of supplying the RSF with weapons incorporating French technology.

Overall, the data suggests that the conflict has not only led to massive direct loss of life but also resulting in a grave humanitarian crisis, exacerbating already existing issues of food insecurity, lack of medical care, and mass displacement. Humanitarian organizations and international actors continue to call for urgent support and intervention to address the escalating crisis.

Fact-checking news summary:
Here is a list of specific facts from the summary, along with assessments of their correctness, importance to the resolution of the question, and relevance to the outcome:
1. Fact: Over 61,000 people died of all causes in Khartoum State between April 2023 and June 2024.
– Check: This fact is consistent with the context given.
– Importance: Important (provides context for scale of mortality).
– Relevance: True.
2. Fact: 26,000 deaths in Khartoum State were directly due to violence.
– Check: Consistent with the statement about high conflict-related mortality.
– Importance: Important (pertains to conflict-specific deaths).
– Relevance: True.
3. Fact: The Armed Conflict Location and Event Data (ACLED) project reported 20,178 intentional-injury deaths for the entire country.
– Check: Needs external verification, but internally consistent.
– Importance: Important (ACLED is directly relevant to the question).
– Relevance: True.
4. Fact: The conflict in Sudan began in April 2023.
– Check: Consistent with external timelines of the conflict’s onset.
– Importance: Lesser (background information on the conflict’s timing).
– Relevance: True.
5. Fact: The conflict involves the Sudanese Armed Forces (SAF) and the Rapid Support Forces (RSF).
– Check: Consistent with external reports about parties in the conflict.
– Importance: Important (identifies the parties involved).
– Relevance: True.
6. Fact: More than 14 million people have been displaced in Sudan due to the conflict.
– Check: Consistent with external humanitarian reports.
– Importance: Important (indicates the scale of displacement, though not directly about deaths).
– Relevance: False.
7. Fact: ACLED reports an average of 1,200 direct conflict deaths per month.
– Check: Needs external verification, but has internal consistency.
– Importance: Critical (direct inference to the question on deaths).
– Relevance: True.
8. Fact: Estimates suggest as many as 150,000 people may have died due to the conflict, considering both direct and indirect causes.
– Check: Needs external verification for the overall estimate.
– Importance: Critical (provides broader estimate of conflict toll).
– Relevance: True.
9. Fact: Regions including Darfur and Kordofan have experienced high levels of violence.
– Check: Consistent with historical and current contexts of the conflict.
– Importance: Important (background on conflict zones).
– Relevance: True.
10. Fact: Amnesty International accuses the UAE of supplying RSF with weapons incorporating French technology.
– Check: Consistent with documented arms trade issues and accusations.
– Importance: Lesser (more about conflict dynamics rather than death counts).
– Relevance: False.
11. Fact: The conflict has exacerbated existing issues of food insecurity, lack of medical care, and mass displacement.
– Check: Consistent with external humanitarian assessments.
– Importance: Important (contextualizes crisis’ humanitarian impact, not direct death count).
– Relevance: False.

In summary, facts 3, 7, and 8 particularly are directly relevant to the reported number of conflict deaths as sought by the original question.

Fact-checked summary:
In assessing future reports from ACLED on state-based conflict deaths in Sudan in 2025, several verified facts are pertinent. ACLED reported 20,178 intentional-injury deaths for the entire country, and the project also averages 1,200 direct conflict deaths per month, indicating substantial ongoing violence (fact 3 and 7). Additionally, broader estimates suggest that as many as 150,000 people may have died due to the conflict when considering both direct and indirect causes (fact 8). These figures are significant because they help outline the scale of the conflict’s impact in recent history, which is directly relevant to estimating future conflict-related deaths.

OpenAI gives a base rate of probability (Carolyn‘s note: OpenAI failed to retuirn a value.)
The question difficulty is rated 5 (0 to 10)
Historical weighted factors include:
Monthly average death rate, 0.7
Seasonal fluctuations in conflict intensity, 0.1
Attempts at peace negotiations, 0.1
Influence of external actors, 0.1
A Bayesian calculation could be performed as follows:
Using the historical factors, we assume a starting point from the previous year, where the monthly average was approximately 1,200 deaths. This gives a yearly estimate of 14,400 deaths without interventions. Taking the weights assigned to historical factors, the base probability is adjusted to account for fluctuations and potential peace efforts, leading to an 80% likelihood estimate for outcomes above 8,000 deaths.
Bayesian base rate: 0.8
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: 163
The following were considered in order to produce this base rate:
The base rate is calculated using the ongoing average death rate of 1,200 per month, considering the conflict’s continued intensity and historical lack of resolution. This accounts for consistent levels of state-based conflict deaths reported historically.

Ideally, the news feed would contain the following sorts of information for a better forecast:
Additional insights into peace negotiation progress and external geopolitical forces would enhance forecasting accuracy. Understanding potential reporting discrepancies by ACLED would also aid the analysis.

Some potential divergent considerations that might affect the base rate:
Potential unforeseen ceasefires or intervention by international communities could significantly alter the outcome, reducing the number of deaths. Additionally, if the reporting by ACLED becomes unreliable, estimates could also diverge.

The following chain of events are necessary for the question to resolve positively:
Monthly rates of conflict-related deaths continue at an average rate similar to previous years. Very likely – No significant de-escalation through peace agreements or foreign intervention. Likely – Accurate and timely reporting of deaths by ACLED throughout 2025. High confidence

Querying Claude (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.05, “Between 3000 and 5000”: 0.12, “Between 5000 and 8000”: 0.25, “Between 8000 and 12000”: 0.35, “More than 12000”: 0.22] – confidence: 6)

Querying Mistral (AI predicts: [“Less than 1000”: 0.02, “Between 1000 and 3000”: 0.05, “Between 3000 and 5000”: 0.13, “Between 5000 and 8000”: 0.25, “Between 8000 and 12000”: 0.35, “More than 12000”: 0.2] – confidence: 6)

Querying OpenAI (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”: 0.3, “More than 12000”: 0.5] – confidence: 8)

Question Type: Multiple Choice
Confidence: 6
MAPD: Avg: 0.066666666666667, Max: 0.2
# LLM responses: 3
Explanations of the above statistical measures here —>

Model value:
* Less than 1,000: 1%
* Between 1,000 and 3,000: 5%
* Between 3,000 and 5,000: 12%
* Between 5,000 and 8,000: 25%
* Between 8,000 and 12,000: 35%
* More than 12,000: 22%

The AIs base their predictions on data from ACLED, which reports an average of 1,200 conflict-related deaths per month in Sudan, extrapolating this to approximately 14,400 deaths for 2025 if the trend persists. As of July 21, 2025, around 60% of the year has passed, implying cumulative deaths could be in the 8,000+ range thus far. The reasoning highlights consistent high-intensity conflict and the absence of significant peace agreements or interventions that might lower the death toll. However, the estimates could be affected by unforeseen ceasefires, international interventions, changes in reporting accuracy, or alterations in conflict intensity and methodology. While historical trends affirm continued high death rates, factors such as unreported deaths and the nuanced difference between state-based and broader conflict deaths could influence the final count. Potential developments not captured in the existing data might lessen the current projections.

Runtime: 144 seconds.

Past forecasts by Phil’s and Jeremy ’s bots —>

Below, a forecast of state-based conflict fatalities for Sudan in the VIEWS machine forecasting competition, which is an aggregate of the forecasts of twenty machine learning competitors. Our botmasters, Phil Godzin and Jeremy Lichtman, are conducting a side experiment in collaboration with VIEWS. The VIEWS aggregate forecast is much lower than ours.