Botmaster Jeremy Lichtman’s Sudan conflict deaths forecast, May 23, 2025

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

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

* Less than 1000: 1%
* Between 1000 and 3000: 4%
* Between 3000 and 5000: 9%
* Between 5000 and 8000: 20
* Between 8000 and 12000: 27%
* More than 12000: 39%

Obtaining outside data.
Checking internet sources, using keywords: ACLED Sudan 2025 conflict deaths
Results from around the web:

Okay, I need to answer the user’s query about finding current web results related to ACLED Sudan conflict deaths in 2025 and summarize them, including tables and a time series. Let me start by looking at the provided search results.

First, the main source is ACLED itself. The search result [1] from December 2024 mentions over 28,700 reported fatalities by November 2024, with 7,500 civilians killed. That’s a key figure. Result [2] seems to repeat the same data, so maybe it’s a different page but same source. Then result [3] from April 2025 talks about SAF gaining upper hand after 23 months, but doesn’t mention deaths directly. Result [4] from February 2025 discusses a study estimating higher deaths in Khartoum state alone, around 26,000 from injuries and over 61,000 total deaths. Result [5] from November 2024 notes 4,204 deaths from 1,792 incidents, mostly by RSF.

The user wants tables. Let me extract the data points. From ACLED [1][2][5], the total reported fatalities are 28,700, civilians 7,500. The Yale study in [4] has higher numbers for Khartoum. The time series part is tricky because the data spans from 2023 to 2024, but the user asked for 2025. Wait, the search results go up to April 2025 in [3], but that’s about SAF control, not deaths. The latest death figures are up to November 2024 in [1] and [5]. So maybe the time series isn’t possible for 2025 data yet. I should note that.

For tables, I can compare ACLED’s national figures with the Khartoum study. Also, the monthly incidents from [5] could be a time series, but they don’t have specific dates. The user wants period, value, change, mean, etc. Since the data is cumulative, maybe annual totals. But without monthly breakdowns, calculating mean and SD might not be feasible. I’ll present what’s available and mention limitations.

URLs to include are the ones provided: [1], [3], [4], [5]. Need to make sure they’re listed correctly. Also, check if all sources are cited in the answer. The answer should start with a summary, then sections with data, tables, and URLs. Avoid headers at the start. Use citations after each relevant sentence. Make sure tables are properly formatted as Markdown. Since time series data isn’t detailed, maybe omit that part but explain why. Conclude with the list of URLs.
The Sudan conflict remains one of the deadliest in the world, with ACLED and other researchers reporting escalating fatalities and humanitarian crises. Below is a synthesis of key data from 2024–2025:
—
### Civilian Casualties and Total Deaths
– **ACLED’s national estimates (December 2024):**
– 28,700+ reported fatalities by November 2024, including **7,500+ civilians killed in direct attacks** [1][2].
– Over 30% of Sudan’s population (14 million people) displaced [1].
– **Violence against civilians:** 1,792 incidents recorded in 2024, causing

OpenAI gives a base rate of probability (Carolyn‘s note: OpenAI failed to return a value.)
The question difficulty is rated 7 (0 to 10)
Historical weighted factors include:
Historical violence levels, 0.3
Consistency of reporting from ACLED, 0.25
Political control shifts, 0.2
International response, 0.15
Civilian displacement figures, 0.1
A Bayesian calculation could be performed as follows:
Based on historical factors, the probability of similar or increased fatalities continuing is calculated by multiplying prior probability (historical data indicates about 28,700 deaths annually) with the likelihood of current conflict dynamics (assumed high due to ongoing conflict). This yields a Bayesian estimate of around 0.85.
Bayesian base rate: 0.85 (Carolyn’s note: the Multi-AI oracle misunderstood the question, instead returning a probability of ???)
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: 223

The following were considered in order to produce this base rate:
The base rate was set at 0.8, considering the high likelihood of ongoing violence based on past data showing a high number of fatalities in previous years. Additionally, factors like control struggles and displacement levels feed into this high probability.

Ideally, the news feed would contain the following sorts of information for a better forecast:
Access to more detailed monthly data in 2025 and updated reports from key conflict zones would significantly enhance the accuracy of forecasts. Additionally, insights from humanitarian organizations on the ground could provide context beyond raw numbers.

Some potential divergent considerations that might affect the base rate:
Potential impacts of unforeseen ceasefires or significant international peacekeeping interventions could lower the fatality rate, although these are currently improbable given the lack of any recent major initiatives.

The following chain of events are necessary for the question to resolve positively:
– The conflict in Sudan must continue to result in significant fatalities throughout 2025, which is a likely scenario given the ongoing dynamics of the conflict.
– The violence should continue at a level similar to or higher than previous years, which seems highly likely based on recent trends and historical data from 2024.
– Continued involvement and activity of armed groups such as SAF and RSF, which is highly likely given the current power dynamics and control over territories.

Querying Claude (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.03, “Between 3000 and 5000”: 0.08, “Between 5000 and 8000”: 0.18, “Between 8000 and 12000”: 0.35, “More than 12000”: 0.35] – confidence: 6)

Querying Mistral (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.15, “Between 3000 and 5000”: 0.2, “Between 5000 and 8000”: 0.3, “Between 8000 and 12000”: 0.24, “More than 12000”: 0.1] – confidence: 7)

Querying OpenAI (AI predicts: [“Less than 1000”: 0.01, “Between 1000 and 3000”: 0.03, “Between 3000 and 5000”: 0.06, “Between 5000 and 8000”: 0.1, “Between 8000 and 12000”: 0.2, “More than 12000”: 0.6] – confidence: 7)

Question Type: Multiple Choice
Confidence: 7
MAPD: Avg: 0.12333333333333, Max: 0.33333333333333
# LLM responses: 3
Explanations of the above statistical measures here —>

Model value:
* Less than 1000: 1%
* Between 1000 and 3000: 4%
* Between 3000 and 5000: 9%
* Between 5000 and 8000: 20
* Between 8000 and 12000: 27%
* More than 12000: 39%

The reasoning across various predictions emphasizes the severe ongoing conflict in Sudan, primarily driven by the power struggle between the SAF and RSF, with the historical context suggesting high fatalities, such as the more than 28,700 reported by ACLED in 2024. Significant factors contributing to continued violence include the lack of meaningful ceasefire or peacekeeping efforts, continued territorial battles, and widespread displacement affecting millions. Despite these grim expectations, some scenarios could alter this trajectory, such as unexpected ceasefires, effective international interventions, or changes in reporting methodologies which could show fewer fatalities. Nonetheless, unless such significant changes or interventions occur, the consensus is that the violence will persist through 2025, with high or escalating fatality numbers likely.

Runtime: 66 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 competitors. Our botmasters, Phil Godzin and Jeremy Lichtman, are conducting a side experiment in collaboration with VIEWS.