Botmaster Jeremy Lichtman’s Sudan conflict deaths forecast, June 23, 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 1000: 2%
* Between 1000 and 3000: 8%
* Between 3000 and 5000: 15%
* Between 5000 and 8000: 25%
* Between 8000 and 12000: 28%
* More than 12000: 22%

11–13 minutes read

Obtaining outside data.
Checking internet sources, using keywords: Sudan conflict deaths 2025
News from various sources:

A recent report by the London School of Hygiene & Tropical Medicine highlights significant underreporting of deaths due to the ongoing conflict in Sudan. This study, which focused primarily on Khartoum State from April 2023 to June 2024, found over 61,000 deaths attributed to various causes, a 50% increase from pre-war levels. Specifically, 26,000 of these were due to violence. This is notably higher than the entire country’s reported 20,178 intentional-injury deaths by the Armed Conflict Location and Event Data (ACLED).

The conflict, beginning in April 2023, is contested between the Sudanese army and the Rapid Support Forces (RSF). It has exacerbated a humanitarian crisis, with preventable diseases and starvation as leading causes of death, particularly harmful in regions like Kordofan and Darfur.

Data collection obstacles stem from limited access to accurate and comprehensive mortality figures. Researchers used public surveys, social media obituaries, and private network surveys to compile their findings, employing a “capture-recapture” analysis method. However, underreporting and difficulty in obtaining raw data present challenges in assessing the true impact.

In light of the severe humanitarian crisis, with 11 million displaced and food scarcity exacerbating already dire conditions, the study urges international intervention and increased aid. Both military factions have been accused of severe human rights abuses, with the RSF primarily blamed for ethnically driven violence in Darfur. Despite these challenges, the conflict’s resolution remains elusive with continued international diplomatic efforts proving insufficient.

Fact-checking news summary:
### Extracted Facts and Analysis
1. **Fact:** A recent report by the London School of Hygiene & Tropical Medicine highlights significant underreporting of deaths due to the ongoing conflict in Sudan.
– **Check**: True. The report is real and matches the description.
– **Importance**: Important. Provides context to the issue of underreporting and frames the overall data challenge.
– **Relevance**: True. Directly relates to understanding the discrepancy in reported deaths, helping approximate future counts.
2. **Fact:** The study focused on Khartoum State from April 2023 to June 2024 and found over 61,000 deaths attributed to various causes.
– **Check**: True, as per the summary’s data.
– **Importance**: Important. It provides specific data measures critical to understanding the extent of the conflict in key areas.
– **Relevance**: True. Helps gauge the report’s findings against other data, influencing the 2025 death estimate.
3. **Fact:** 26,000 of those deaths were due to violence.
– **Check**: True, consistent with the summary.
– **Importance**: Critical. This directly relates to conflict-related deaths, the core focus of the original question.
– **Relevance**: True. Helps define the measurement scope relevant to ACLED figures.
4. **Fact:** This is notably higher than the entire country’s reported 20,178 intentional-injury deaths by ACLED.
– **Check**: True, assuming the report’s figures are accurate.
– **Importance**: Important. It highlights discrepancies between different reports.
– **Relevance**: True. It illustrates the potential underreporting that needs adjustment in 2025 projections.
5. **Fact:** The conflict began in April 2023 between the Sudanese army and the Rapid Support Forces (RSF).
– **Check**: True. Consistent with external sources.
– **Importance**: Important. Provides the timeline and parties involved in the conflict.
– **Relevance**: True. Key to understanding current and future conflict dynamics related to casualty estimates.
6. **Fact:** Preventable diseases and starvation are leading causes of death in regions like Kordofan and Darfur.
– **Check**: True. Matches summary and wider analysis of humanitarian issues.
– **Importance**: Less Important. While it provides context, it’s secondary to direct conflict deaths.
– **Relevance**: True. Affects overall mortality but not direct state-based conflict deaths.
7. **Fact:** Data collection obstacles come from limited access to accurate and comprehensive mortality figures.
– **Check**: True. Universally acknowledged in similar conflict zones.
– **Importance**: Important. Explains part of the reporting discrepancies.
– **Relevance**: True. Directly influences potential 2025 reporting by ACLED.
8. **Fact:** Researchers used a “capture-recapture” analysis method.
– **Check**: True, as outlined in the summary.
– **Importance**: Less Important. Technical detail regarding methodology.
– **Relevance**: True. Influences understanding of data reliability which affects estimations.
9. **Fact:** The human rights abuses in Darfur are primarily blamed on the RSF.
– **Check**: True, consistent with numerous reports.
– **Importance**: Less Important. Background information on conflict but not direct mortality figures.
– **Relevance**: False. Impacts understanding of conflict nature, but not directly relevant to the death count estimation for 2025.
10. **Fact:** There are 11 million persons displaced due to the conflict.
– **Check**: Assume true based on current crisis assessments.
– **Importance**: Important. Illustrates the conflict’s scale which indirectly relates to potential death reports.
– **Relevance**: True. Highlights the severity of the humanitarian crisis, potentially affecting future death tolls.
Overall, while many extracted facts are of varying importance and relevance, they help form a clearer basis for estimating potential state-based conflict deaths in 2025 by ACLED in Sudan.

Fact-checked summary:
A recent report by the London School of Hygiene & Tropical Medicine reveals significant underreporting of deaths due to the ongoing conflict in Sudan, which is critical as it provides context to data discrepancies affecting future death estimates. The report, focusing on Khartoum State from April 2023 to June 2024, found over 61,000 deaths, with 26,000 attributed to violence, illustrating the conflict’s severe impact. Notably, this figure surpasses ACLED’s reported 20,178 intentional-injury deaths for the entire country, highlighting potential underreporting that might inform projections for 2025. The conflict that began in April 2023 between the Sudanese army and the Rapid Support Forces underscores the timeline and involved parties, aiding in understanding ongoing and future conflict dynamics. Data collection challenges, due to limited access to accurate mortality figures, remain a significant barrier affecting potential reporting by ACLED in 2025. Additionally, the displacementof 11 million people due to the conflict illustrates the crisis’s scale, which could indirectly influence future death tolls.

OpenAI gives a base rate of probability of exceeding 12,000 deaths in 2025
The question difficulty is rated 8 (0 to 10)
Historical weighted factors include:
Past underreporting trends based on LSHTM report, 0.3
ACLED’s historical reporting accuracy, 0.2
Sudan’s historical conflict severity, 0.3
Displacement impact on mortality rates, 0.2
A Bayesian calculation could be performed as follows:
Using historical death rates adjusted for underreporting, and accounting for conflict severity, I calculated the probability of exceeding 12,000 deaths by adjusting the ACLED historical baseline upwards by a factor commensurate with the discrepancy noted by the LSHTM report, combined with ongoing conflict severity weighting.
Bayesian base rate: 0.2
Sufficient news to provide a good forecast? 0 (0 or 1)
News is relevant, topical and unbiased? 1 (0 or 1)
Question classification: scenario_based_forecast
Expected time (days) until question close: 192
The following were considered in order to produce this base rate:
Based on the historical factors, there is a 0.15 base probability that deaths in 2025 will exceed 12,000, mostly driven by the established pattern of underreporting and conflict severity from past data.

Ideally, the news feed would contain the following sorts of information for a better forecast:
Additional detailed data on recent conflict developments and humanitarian access improvements in Sudan would allow for more refined death toll estimates.

Some potential divergent considerations that might affect the base rate:
Potential peace negotiations or ceasefire agreements, which were not previously common, might reduce the death toll significantly. Conversely, if ongoing displacements exacerbate health crises, it could lead to higher mortality directly related to the conflict.

The following chain of events are necessary for the question to resolve positively:
Continuation of the conflict between the Sudanese army and the Rapid Support Forces throughout 2025 Likely – Significant underreporting of deaths continues or worsens Possible – Increased access to accurate mortality figures remains limited Likely

Querying Claude (AI predicts: [“Less than 1000”: 0.02, “Between 1000 and 3000”: 0.08, “Between 3000 and 5000”: 0.15, “Between 5000 and 8000”: 0.25, “Between 8000 and 12000”: 0.28, “More than 12000”: 0.22] – confidence: 6)

Querying Mistral (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.1, “Between 3000 and 5000”: 0.25, “Between 5000 and 8000”: 0.35, “Between 8000 and 12000”: 0.25, “More than 12000”: 0.04] – confidence: 5)

Querying OpenAI (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”: 0.3, “More than 12000”: 0.39] – confidence: 6)

Question Type: Multiple Choice
Confidence: 6
MAPD: Avg: 0.09, Max: 0.23333333333333
# LLM responses: 3
Explanations of the above statistical measures here —>

Model value:
* Less than 1000: 2%
* Between 1000 and 3000: 8%
* Between 3000 and 5000: 15%
* Between 5000 and 8000: 25%
* Between 8000 and 12000: 28%
* More than 12000: 22%

22%The various AI analyses converge on the notion that the ongoing conflict between the Sudanese army and Rapid Support Forces, coupled with substantial underreporting of death figures, has created a severe humanitarian crisis. Historical patterns of conflict and data underreporting suggest a high likelihood for continued violence and inaccurately reported death tolls through the end of 2025. The displacement of 11 million people exacerbates the situation by increasing indirect mortality rates. However, potential developments, such as successful peace negotiations, ceasefire agreements, or improvements in data collection, could lead to a more favorable outcome with lower casualties. International interventions and improved humanitarian access could further alleviate the mortality rate and alter the current outlook.

Runtime: 207 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.