Ukraine Conflict Deaths? Botmaster Jeremy Lichtman’s forecast, Oct. 16, 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 October 2025?

Jeremy’s Multi-AI Oracle predicts today:

Less than 500: 5%
Between 500 and 1,000: 15%
Between 1,000 and 1,500: 30%
Between 1,500 and 2,000: 35%
Greater than 2,000: 15%

12–15 minutes read

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

In 2025, the conflict in Ukraine has led to significant human loss and continued devastation. According to various reports, Ukraine’s counter-intelligence and other open intelligence sources have estimated Russian military losses to be over 1 million when combining dead, wounded, and missing. Specifically, a leaked document reported 281,550 Russian soldiers lost just in the first eight months of the year, with Ukraine claiming that their military intelligence largely corroborated these figures.

In contrast, Ukraine has sustained lower losses, as they have been more focused on defense. Ukrainian President Volodymyr Zelenskyy reported over 46,000 Ukrainian soldiers killed and 390,000 wounded by February 2025, indicating a strong effort to rescue and treat the wounded compared to Russian practices.

The war’s toll on civilians has also been severe, with the United Nations reporting that 14,116 Ukrainian civilians have been killed and 36,481 injured since the conflict began. Despite intense international monitoring, civilian casualties in Russia, due to cross-border shelling and drone attacks, also continue but remain largely undocumented.

Mass recruitment efforts are evident in both nations, from Ukrainian streets flooded with recruitment ads to Russia seeking soldiers as far afield as Africa and the Middle East. Russian military recruitment faces challenges due to high casualty rates and troubling conditions for recruits.

Russia has dramatically increased the production of drones, using enhanced combat drones as a national priority to exert pressure on Ukrainian defense systems. Despite substantial drone attacks, Ukraine manages to intercept about 85% of them, although the threat remains high with vigorous assaults on energy infrastructures causing intentional power outages.

On the geopolitical stage, diplomatic efforts continue. Donald Trump’s administration has put renewed pressure on Russia to negotiate peace, focusing on energy sanctions targeting Russian oil sales to India and potentially China. There’s also discussion on providing Ukraine with advanced weaponry like Tomahawk missiles, though significant debates on its implications persist.

Infrastructure damage in Ukraine continues to mount, requiring enormous recovery efforts estimated at $524 billion by a joint World Bank, UN, and EU assessment. In Gaza, similar devastation is reported, exacerbating longstanding humanitarian issues but representing a different conflict dynamic compared to Ukraine.

Both conflicts have drawn substantial international legal and diplomatic attention, with ongoing debates at the UN General Assembly and other bodies reflecting geopolitical divides and humanitarian advocacy.

Fact-checking news summary:
Here is a list of specific facts from the summary, along with an analysis of their correctness, importance to the resolution of the original question, and relevance:
1. **Fact:** Ukraine’s counter-intelligence and other sources estimate Russian military losses to be over 1 million (including dead, wounded, and missing).
– **Correctness:** This figure might be internally consistent within the context provided, but its correctness could be disputed externally as it’s a very high estimate.
– **Importance:** Important (provides context on the scale of conflict).
– **Relevance:** True (provides context but not directly related to state-based conflict death counts).
2. **Fact:** A leaked document reported 281,550 Russian soldiers lost in the first eight months of 2025.
– **Correctness:** Internally consistent but likely not verifiable externally; very high estimate.
– **Importance:** Important (significant for understanding the conflict scale).
– **Relevance:** True (relevant for understanding military conflict scale, not direct death counts).
3. **Fact:** Ukraine sustained over 46,000 soldiers killed and 390,000 wounded by February 2025.
– **Correctness:** Internally consistent but high; could be disputed externally.
– **Importance:** Important.
– **Relevance:** True (relevant for understanding the conflict scale).
4. **Fact:** The United Nations reported 14,116 Ukrainian civilian deaths since the conflict began.
– **Correctness:** Internally consistent; possibly accurate with external verification.
– **Importance:** Critical (directly pertains to calculating total conflict deaths).
– **Relevance:** True (relevant to the outcome of the question).
5. **Fact:** 36,481 Ukrainian civilians injured since the conflict began.
– **Correctness:** Internally consistent; possibly accurate with external verification.
– **Importance:** Important.
– **Relevance:** True (provides context on civilian impact).
6. **Fact:** Civilian casualties in Russia remain largely undocumented.
– **Correctness:** Likely accurate; true in practice due to limited data.
– **Importance:** Important (affects calculation of total deaths if considered).
– **Relevance:** True (relevant for comprehensive casualty context).
7. **Fact:** Ukraine intercepts about 85% of Russian drones.
– **Correctness:** Internally plausible; could be reasonable with external verification.
– **Importance:** Lesser (tactical detail).
– **Relevance:** True (provides context but not directly related to death counts).
8. **Fact:** Donald Trump’s administration pressures Russia for peace talks.
– **Correctness:** Internally consistent with narrative possibilities.
– **Importance:** Lesser.
– **Relevance:** False (not directly relevant to conflict death count).
9. **Fact:** Infrastructure damage in Ukraine requires $524 billion in recovery efforts.
– **Correctness:** Internally consistent; possibly excessive or not externally corroborated.
– **Importance:** Lesser.
– **Relevance:** False (not directly relevant to conflict death count).
10. **Fact:** Substantial international legal and diplomatic attention on the conflicts at the UN.
– **Correctness:** Likely accurate; generally true.
– **Importance:** Lesser.
– **Relevance:** False (not directly relevant to conflict death count).

This analysis assumes the statements’ correctness based on consistency within the provided summary and typical reports in conflict settings. External verification with reliable sources would be necessary for confirmation.

Fact-checked summary:
In determining the number of state-based conflict deaths reported by ACLED in Ukraine in October 2025, relevant and verified facts include the United Nations’ report of 14,116 Ukrainian civilian deaths since the conflict began, which is critical information for calculating total conflict deaths. Additionally, Ukraine sustained over 46,000 soldiers killed and 390,000 wounded by February 2025, which is important for understanding the scale of military casualties. Civilian casualties in Russia remain largely undocumented, an important consideration affecting the overall calculation of total deaths. These facts are relevant to the original question as they pertain to the ongoing assessment of both civilian and combatant losses in the conflict.

OpenAI gives a base rate of probability range of total conflict deaths falling within a given category (Carolyn‘s note: OpenAI failed to return a value)
The question difficulty is rated 7 (0 to 10)
Historical weighted factors include:
Historical monthly casualties in the Ukraine conflict, 0.5
Intensity of conflict in 2025 compared to prior years, 0.3
Availability and reliability of casualty reporting, 0.2
A Bayesian calculation could be performed as follows:
Using historical data, if the average monthly total has been around 1000 due to varying intensity, and recent intensifications in 2025 suggest an increased casualty count. Assuming P(H) = 0.3 for high casualties based on historical frequency, P(E|H) = 0.7 for increased intensity, and P(E) = 0.5, then: P(H|E) = (P(E|H) * P(H)) / P(E) = (0.7 * 0.3) / 0.5 = 0.42.
Bayesian base rate: 0.42 (Carolyn‘s note: In cases where the Multi-AI Oracle is only forecasting a single percent probability, this Bayesian rate is easy to understand, an adjustment to a probability. But in this case, with five probability buckets, this apparently is an overall adjustment factor across all buckets.)
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: 45
The following were considered in order to produce this base rate:
The base rate was derived by considering historical monthly conflict casualties, the expected increase due to ongoing intense military operations, and verified reports indicating heavy losses in 2025. Past patterns and reports suggest a substantial likelihood of high casualty numbers, aligning with the ‘Between 1500 and 2000’ category.

Ideally, the news feed would contain the following sorts of information for a better forecast:
Additional insights into undocumented or less verified casualties, both military and civilian, especially in Russian territories or areas with limited reporting, would increase forecast accuracy. Detailed intelligence on shifts in conflict strategy or peace negotiations would provide valuable context.

Some potential divergent considerations that might affect the base rate:
Potential underreporting of civilian casualties, especially in Russia, and potential ceasefire developments or changes in conflict dynamics that could suddenly drop the casualty rate could diverge the actual number from the base projection.

The following chain of events are necessary for the question to resolve positively:
Continued active conflict in Ukraine throughout the entirety of October 2025. Very likely – Sufficient reporting and verification of deaths by organizations such as ACLED. Likely – High-intensity conflict resulting in significant civilian and military casualties. Likely

Querying Claude (AI predicts: [“Less than 500”: 0.05, “Between 500 and 1000”: 0.15, “Between 1000 and 1500”: 0.3, “Between 1500 and 2000”: 0.35, “Greater than 2000”: 0.15] – confidence: 6)

Querying Mistral (AI predicts: [“Less than 500”: 0.05, “Between 500 and 1000”: 0.15, “Between 1000 and 1500”: 0.3, “Between 1500 and 2000”: 0.4, “Greater than 2000”: 0.1] – confidence: 5)

Querying OpenAI (AI predicts: [“Less than 500”: 0.05, “Between 500 and 1000”: 0.1, “Between 1000 and 1500”: 0.2, “Between 1500 and 2000”: 0.35, “Greater than 2000”: 0.3] – confidence: 7)

Question Type: Multiple Choice
Confidence: 6
MAPD: Avg: 0.053333333333333, Max: 0.13333333333333
# LLM responses: 3
Explanations of the above statistical measures here —>

Model value:
Less than 500: 5%
Between 500 and 1,000: 15%
Between 1,000 and 1,500: 30%
Between 1,500 and 2,000: 35%
Greater than 2,000: 15%

The various AI analyses converge on a prediction of high casualty numbers in the range of 1500-2000 deaths, based on established base rates and historical patterns of conflict intensity in Ukraine. These forecasts consider the ongoing intense nature of military operations, significant documented losses in 2025, and the reliability of sources like the United Nations and ACLED for casualty reporting. However, several uncertainties could influence these projections, including potential underreporting, especially of civilian casualties and Russian losses, as well as unexpected changes in conflict dynamics, such as ceasefires or de-escalations, that might alter casualty figures. The challenges of accurate data collection in conflict zones further complicate the certainty of these estimates.

Runtime: 216 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.)