Ukraine Conflict Deaths? Botmaster Jeremy Lichtman’s forecast, July 31, 2025

How many state-based conflict deaths (total of all civilian and combat deaths, including both Ukrainian and Russian combatants) will be reported by ACLED in Ukraine in July 2025?

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

* Less than 500: 5%
* Between 500 and 1,000: 16%
* Between 1,000 and 1,500: 26%
* Between 1,500 and 2,000: 26%
* Greater than 2,000: 26%

5-6 minutes read

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

(Carolyn’s note: Today, the Multi-AI Oracle failed to provide any news. Despite this, all three of the Oracle’s AIs generated probabilities consistent with Ukraine’s current situation where, according to ACLED, the conflict was 10%-15% less during 19 – 25 July 2025.)

OpenAI gives a base rate of State-based conflict deaths per month in Ukraine (Carolyn‘s note: OpenAI failed to return a value) The question difficulty is rated 7 (0 to 10)
Historical weighted factors include:
0.40.30.20.1
A Bayesian calculation could be performed as follows:
(Carolyn’s note: The Multi-AI Oracle failed to return a value)
Bayesian base rate: 0.3
Sufficient news to provide a good forecast? 0 (0 or 1)
News is relevant, topical and unbiased? 0 (0 or 1)
Question classification: reference_class
Expected time (days) until question close: 30

The following were considered in order to produce this base rate:
The base rate was calculated using historical data on state-based conflict deaths in Ukraine. Trends from previous years and months were considered, along with the current geopolitical climate which suggests a moderate likelihood of resolution reaching various thresholds of conflict deaths.

Ideally, the news feed would contain the following sorts of information for a better forecast:
Access to reliable databases or international reports documenting conflict deaths would improve forecast accuracy. Insights from military analysts or geopolitical experts could further refine predictions.

Some potential divergent considerations that might affect the base rate:
There could be sudden escalations or resolutions that significantly alter the death count, influenced by international diplomatic interventions or changes in military strategies.

The following chain of events are necessary for the question to resolve positively:
– Accurate data compilation on state-based conflict deaths in Ukraine for July 2025. (Likely)
– Clear delineation of what constitutes a state-based conflict death. (Moderate likelihood)
– Effective transmission and reporting of data by media and relevant agencies. (Moderate likelihood)
– No major unforeseen escalation or de-escalation during July affecting the numbers. (Moderate likelihood)

Querying Claude (AI predicts: [“Less than 500”: 0.05, “Between 500 and 1000”: 0.15, “Between 1000 and 1500”: 0.25, “Between 1500 and 2000”: 0.3, “Greater than 2000”: 0.25] – confidence: 4)

Querying Mistral (AI predicts: [“Less than 500”: 0.15, “Between 500 and 1000”: 0.25, “Between 1000 and 1500”: 0.25, “Between 1500 and 2000”: 0.2, “Greater than 2000”: 0.15] – confidence: 6)

Querying OpenAI (AI predicts: [“Less than 500”: 0.05, “Between 500 and 1000”: 0.15, “Between 1000 and 1500”: 0.25, “Between 1500 and 2000”: 0.25, “Greater than 2000”: 0.3] – confidence: 5)

Question Type: Multiple Choice
Confidence: 5
MAPD: Avg: 0.06, Max: 0.1
# LLM responses: 3
Explanations of the above statistical measures here —>

Model value:
* Less than 500: 5%
* Between 500 and 1,000: 16%
* Between 1,000 and 1,500: 26%
* Between 1,500 and 2,000: 26%
* Greater than 2,000: 26%

The predictions for the ongoing Russia-Ukraine conflict in July 2025 are based on historical data and the patterns observed in ACLED reports, which indicate that monthly death tolls typically range between 1000-3000+ during active conflict periods. These assessments consider the conflict’s continuation into its fourth year, typical summer military activity, and the absence of major ceasefires or peace agreements, suggesting a higher probability for mid-range casualties (1500-2000+). However, unforeseen factors such as sudden diplomatic breakthroughs or international interventions could significantly alter these predictions. Additionally, discrepancies in data collection and reporting accuracy could also impact the casualty figures.

Runtime: 118 seconds.

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

Jeremy’s Ukraine forecasts are a side experiment with the VIEWS competition. The VIEWS Ukraine dashboard is pictured below. (Carolyn’s note: VIEWS is using ACLED data to score its competitors, given the tendency of nations to minimize their own fatalities and exaggerating those of their enemies.)