Polymarket prices “Will Russia capture all of Kostyantynivka by December 31, 2026?” at 40.5¢ on the Yes — a 40.5% market-implied probability — against our independently built estimate of 43%, an edge of just +2.5 percentage points (pp) that shrinks to +2pp once you pay the 41¢ ask. On a contract this thin, that is not an edge; it is noise.
The Institute for the Study of War (ISW) assessed on September 1, 2026 that Russian forces hold a presence in 72.23% of Kostyantynivka and have still not completed its seizure. Polymarket’s Yes side trades at 40.5¢ with $1,885 of volume in 24 hours. This Deep Dive walks the resolution rules, the base rates from four comparable Donbas urban battles, our probability working, and the observable signals that would move the number.
The Bet at a Glance:
• Market: “Will Russia capture all of Kostyantynivka by December 31, 2026?” — Polymarket, condition ID 0x627ed310…f863
• Price: Yes 40.5¢ / No 59.5¢; best bid 40¢, best ask 41¢, last trade 40¢ = 40.5% market-implied probability — Polymarket API snapshot, 19:13 UTC, September 2, 2026
• Our estimate: 43% — base-rate model built from four ISW-dated Donbas urban battles plus the current ISW control figure, September 2, 2026
• Edge: +2.5pp on the Yes at the midpoint, +2pp at the 41¢ ask — our model, below our 8pp threshold for a market this illiquid
• Liquidity: $57,425 on the book, $1,885 traded in 24 hours against $840,910 lifetime volume — Polymarket API, September 2, 2026
• Current state: Russian presence in 72.23% of the city; seizure not complete — ISW, September 1, 2026
• Disconfirmation: ISW shading the last 27.77% as assessed Russian control, or the Druzhkivka road being cut — ISW daily assessments
Methodology: how we built the 43%
Our estimate is a base-rate model, not a battlefield forecast. We took four Donbas urban battles that ISW dated in its own daily assessments — Bakhmut, Avdiivka, Vuhledar and Toretsk — and measured the elapsed time between the first ISW-recorded Russian break-in to the built-up area and the day full Russian control was acknowledged or credibly assessed. We then placed Kostyantynivka on that distribution using ISW’s September 1, 2026 control figure and the 120 days remaining to the December 31 deadline, and added a small allowance for the market’s negotiated-settlement clause.
The caveats are large and we will not hide them. The sample is four battles, which is far too small for a confident distribution; each was fought under different drone, artillery and manpower conditions; and “presence” in ISW’s language is not the same thing as the “assessed control” the contract requires. Where sources disagree — and they do, notably on Kostyantynivka’s pre-war population — we say so rather than pick the convenient number.
The market, the price and why the thinness matters
The contract asks whether Russia will capture all of Kostyantynivka, Donetsk Oblast (48.528896° N, 37.702514° E) by 11:59pm ET on December 31, 2026. Resolution is unusually mechanical: per Polymarket’s own rules, Yes requires the entirety of the municipality to be shaded red on the ISW map, and that shading must persist through the next full ISW daily update cycle. Shading marked “Assessed Russian Infiltration Areas in Ukraine” explicitly does not qualify — only assessed control, assessed advance, or gains in the past 24 hours. A negotiated settlement resolves Yes only if actual control is established on the ground.
That specification matters, because the headline number everyone quotes is infiltration-inclusive. Our snapshot, taken from the Polymarket API at 19:13 UTC on September 2, 2026, shows Yes at 40.5¢ against No at 59.5¢, with a best bid of 40¢ and a best ask of 41¢. Lifetime volume is $840,910. Volume over the previous 24 hours was $1,885. Order-book liquidity stands at $57,425.
Is Yes at 40.5¢ value? On our numbers, only marginally, and not enough to act on. A 40.5¢ midpoint implies a 40.5% probability, and our base-rate work lands at 43% — a 2.5pp gap. But the market is quoted 40 bid at 41 offered, so a buyer of Yes actually pays 41¢ and needs the true probability to clear 41% simply to break even, leaving roughly 2pp. Against that, the 24-hour turnover of $1,885 on a $57,425 book means the quoted price is not a confident consensus of many informed participants; it is a handful of orders resting on a market that has traded $840,910 since February 19, 2026. Wide uncertainty on the underlying and a thin book both argue for demanding a much larger cushion than 2pp before treating a price as wrong.
| Side | Market price | Implied prob | Our estimate | Edge (pp) |
|---|---|---|---|---|
| Yes (Russia captures all) | 40.5¢ (bid 40, ask 41) | 40.5% | 43% | +2.5 |
| No (does not) | 59.5¢ | 59.5% | 57% | -2.5 |
| Yes, paying the 41¢ ask | 41.0¢ | 41.0% | 43% | +2.0 |
Sources: Polymarket gamma API snapshot, 19:13 UTC, September 2, 2026 (outcome prices 0.405 / 0.595; best bid 0.40; best ask 0.41; last trade 0.40); our base-rate model, September 2, 2026. Prediction-market prices in cents are the implied probability directly, so no de-vig is required; the 1¢ bid-ask spread is the only friction priced in above.
“Russian forces maintain a presence (either through advances or infiltrations) in 72.23 percent of Kostyantynivka and have failed to completely seize the settlement, even after months of increased infiltrations and urban fighting.”
— Institute for the Study of War, Russian Offensive Campaign Assessment, September 1, 2026
(Institute for the Study of War)
The case for Yes: 72% of a city and 120 days
The bull case for the Yes is arithmetic. Russian forces broke into Kostyantynivka in late June 2026 — the Kyiv Independent’s Francis Farrell reported on June 23, 2026 that “the battle for the legendary ‘fortress belt’ of Ukraine’s Donbas region has begun in earnest, with Russian forces breaking into the city of Kostiantynivka in Donetsk Oblast and steadily overrunning the urban area”. From that break-in to ISW’s September 1 assessment is 70 days, over which Russian presence reached 72.23% of the municipality. Extend that rate naively and the remaining 27.77% falls inside a month.
There are 120 days between our snapshot and the December 31 deadline. That is a longer runway than Avdiivka required end to end: ISW recorded the launch of Russian offensive operations in the Avdiivka area on October 9, 2023, and on February 17, 2024 assessed that Ukrainian forces had likely withdrawn after Russian Defence Minister Sergei Shoigu claimed “full control” — 131 days. Kostyantynivka has already used 70 of a possible 190.
The logistics picture also cuts toward Yes. ISW reported on September 1, 2026 that a Ukrainian brigade in the Kostyantynivka direction had observed Russian forces beginning to shell the approaches to Oleksiyevo-Druzhkivka with tube artillery, having previously used only multiple-launch rocket systems, in an effort to interdict Ukrainian logistics to Druzhkivka. Sustained interdiction of a defender’s single supply road has preceded the collapse of every position in the table below. Vadym Lyakh, head of the Slovyansk military administration, said Russian forces are now seven to 10 kilometres from Slovyansk and within artillery range — evidence that the wider fortress belt is under pressure.
| Settlement | ISW-dated break-in / offensive start | Date full control acknowledged or assessed | Elapsed days |
|---|---|---|---|
| Avdiivka | October 9, 2023 (ISW, October 10, 2023) | February 17, 2024 (ISW: Ukrainian withdrawal likely) | 131 |
| Vuhledar | September 23–24, 2024 (ISW, September 24, 2024) | October 2, 2024 (Ukrainian sources acknowledged seizure) | 9 |
| Bakhmut | Claimed complete on May 20, 2023 by Yevgeny Prigozhin | Contested the same day: Ukrainian officials reported fighting in southwest Bakhmut on May 20, 2023 | 0 (claim and reality diverged) |
| Toretsk | Russian MoD claimed advances February 7, 2025 | Still contested: ISW recorded Russian attacks in Toretsk itself on February 6 and 7, 2025 | 180+ from 2024 entry |
| Kostyantynivka | June 23, 2026 (Kyiv Independent) | Not taken: 72.23% presence as of September 1, 2026 | 70 and counting |
Sources: ISW Russian Offensive Campaign Assessments of October 10, 2023; February 17, 2024; September 24, 2024; October 2, 2024; May 20, 2023; February 7, 2025; September 1, 2026. Kyiv Independent, June 23, 2026. Elapsed days are calendar days between the two dated events in each row.
The case against: the last quarter is the slowest quarter
The naive extrapolation above is the single weakest input in any model of this bet, and the base rates are what break it. Urban capture is not linear. It is front-loaded in low-density districts and then stalls where the defender concentrates, which is what the Bakhmut and Toretsk rows record. On May 20, 2023, Yevgeny Prigozhin declared Bakhmut completely captured, and ISW noted the same day that Ukrainian officials reported Ukrainian forces still fighting in a small section of southwest Bakhmut. Toretsk is starker: ISW’s assessment of February 7, 2025 recorded Russian forces still attacking in Toretsk itself, roughly six months after Russian troops first entered a town a fraction of Kostyantynivka’s size.
There is also a definitional trap. The 72.23% figure is explicitly “presence (either through advances or infiltrations)” — and Polymarket’s rules exclude infiltration shading. The number that matters here is smaller than the headline, and neither ISW nor the market publishes the split.
Then there is the theatre-wide trend, which is going the wrong way for the Yes. ISW assessed that Russian forces seized or consolidated about 90.35 square kilometres across Ukraine in August 2026, an average of 2.91 square kilometres per day, or 123.53 square kilometres including infiltrations. In August 2025 the figure was 505.55 square kilometres. Russia’s monthly rate of advance is running at under a fifth of the previous year’s. In the same month, ISW assessed Ukrainian forces liberated at least 129.81 square kilometres — more ground than Russia took. And on the defensive side of the ledger, ISW reported on September 1, 2026, citing Hromadske, that Brigadier General Andriy Biletsky will command a new formation with an area of responsibility running from the Lyman direction to the Kramatorsk-Kostyantynivka area, with Brigadier General Denys Prokopenko taking the Dobropillya sector to the south — a fresh command structure aimed squarely at this sector.
“During July 3, the enemy carried out 11 assaults in the specified direction, but no success was achieved. Instead, the enemy, not for the first time, resorts to the dissemination of outright disinformation and fakes by the highest officials.”
— Major Andriy Kovalev, Spokesperson, General Staff of the Armed Forces of Ukraine, speaking to Ukrainska Pravda
(The Kyiv Independent)
Kovalev’s statement matters for a tradeable reason. On July 4, 2026, Russian President Vladimir Putin claimed Kostyantynivka was under Russian control. President Volodymyr Zelensky rejected it — “Of course, this is not true – it is simply yet another Russian lie designed to generate some sort of news” — and two months later the city was still not seized on ISW’s map. Expect further capture claims that do not move the resolution source. The contract resolves on ISW shading, not announcements.
Sources also disagree on the basic scale of the place. The Kyiv Independent’s June 2026 analysis describes Kostyantynivka as “once home to a population of over 100,000”; its July 2026 news report, quoting Russian chief of the general staff Valery Gerasimov, calls it a “once-67,000-person city”. The first is plausibly a historic peak and the second a more recent census-era figure. We flag the discrepancy rather than resolve it, because the size of the built-up area is a live input into how long the final districts take.
Where this bet breaks
The lean on Yes at 40.5¢ is thin by construction, and it rests on assumptions that could fail. It breaks if any of these fire:
- The Druzhkivka road is cut and stays cut. ISW recorded Russian tube artillery striking the approaches to Oleksiyevo-Druzhkivka on September 1, 2026 to interdict Ukrainian logistics. If resupply and casualty evacuation to the remaining Ukrainian districts stop entirely, the timeline compresses hard and 43% is far too low.
- ISW starts shading the western districts as assessed control rather than infiltration. This is the single cleanest signal, because it is the resolution source moving. Presence converting to assessed control in the last 27.77% is what a Yes actually needs.
- The price moves past 55¢ or below 30¢ on real volume. Above roughly 55¢ our 43% estimate is beaten by the market and the small edge inverts to the No side; below 30¢ the gap becomes wide enough to matter even after the spread. A move on $1,885-a-day turnover, however, is not information.
- The new Ukrainian corps commands hold, or a ceasefire freezes the line. The Biletsky and Prokopenko formations reported by ISW on September 1, 2026 are an untested variable. Equally, a negotiated settlement that hands Russia the city on paper without establishing actual control resolves No under Polymarket’s rules, not Yes.
What to watch between now and December 31
The catalyst here is not an event but a cadence. ISW publishes a Russian Offensive Campaign Assessment daily, with a stated data cutoff of 1:00pm ET and an assessment time of 9:00pm ET, and the contract requires any qualifying shading to survive the next full daily update cycle. That means a Yes resolution needs two consecutive ISW publications to agree — a meaningful brake on resolution from a single ambiguous map day.
Three specific things are worth tracking. First, ISW’s monthly totals: August 2026 came in at 90.35 square kilometres seized against 505.55 in August 2025, and a September figure that reverses that decline would be a genuine signal. Second, the settlements north of the city: a Russian milblogger claimed on September 1, 2026 that Russian forces had seized Fedorivka, north of Kostyantynivka — a claim ISW records but does not confirm, and exactly the kind of unverified assertion that moves thin markets without moving maps. Third, DeepStateMap, which Polymarket names as the backup resolution source if the ISW map becomes unavailable.
TL;DR
Polymarket’s Yes on “Will Russia capture all of Kostyantynivka by December 31, 2026?” trades at 40.5¢, a 40.5% market-implied probability. Our base-rate model — four ISW-dated Donbas urban battles, plus ISW’s September 1, 2026 finding of Russian presence in 72.23% of the city and 120 days to the deadline — puts the true number at 43%. That is a +2.5pp edge, +2pp after paying the 41¢ ask, well inside the noise on a market that turned over $1,885 in 24 hours. The price is roughly fair. It breaks if ISW converts the western districts from infiltration shading to assessed control.
FAQ
What are the odds on Russia capturing all of Kostyantynivka by December 31, 2026?
Polymarket quoted the Yes at 40.5¢ and the No at 59.5¢ at 19:13 UTC on September 2, 2026, with a best bid of 40¢, a best ask of 41¢ and a last trade at 40¢. In a prediction market the price in cents is the implied probability, so 40.5¢ equals a 40.5% market-implied chance. Lifetime volume on the contract is $840,910 since it opened on February 19, 2026, but only $1,885 changed hands in the previous 24 hours against $57,425 of resting liquidity.
Is Yes at 40.5¢ value?
Only marginally, and not by enough to matter. Our base-rate estimate is 43% against a 40.5% midpoint, a 2.5pp gap that falls to 2pp once you cross the 1¢ spread and pay 41¢. On a market this illiquid we want at least 8pp before calling a price wrong, because a handful of orders on a $57,425 book is not a considered consensus and the underlying uncertainty is genuinely wide. The honest verdict is that the market is roughly fair here.
What exactly does the market require to resolve Yes?
The entirety of Kostyantynivka municipality must be shaded red on the ISW map by 11:59pm ET on December 31, 2026, and that shading must persist through the next full ISW daily update cycle. Assessed Russian control, assessed advance, and gains in the past 24 hours all qualify; areas marked as Russian infiltration do not. A negotiated settlement counts only if it establishes actual control on the ground, not de jure control.
How much of the city does Russia hold now?
ISW assessed on September 1, 2026 that Russian forces maintain a presence, through advances or infiltrations, in 72.23% of Kostyantynivka, and had failed to completely seize it after months of urban fighting. Because the infiltration component does not count toward Yes resolution, the share that is actually relevant to this contract is smaller than 72.23%, and neither ISW nor Polymarket publishes that split.
What would change our estimate?
Two things above all. If ISW begins shading the remaining western districts as assessed Russian control rather than infiltration, our number rises sharply. If Russia’s monthly rate of advance keeps falling — August 2026 was 90.35 square kilometres against 505.55 in August 2025, while Ukraine liberated 129.81 square kilometres in the same month — and the new Ukrainian corps commands stabilise the sector, our number falls toward the low 30s.
Why does the market’s low volume matter so much?
Because price only carries information when informed money is willing to move it. This contract traded $1,885 in 24 hours against a $57,425 order book, so the 40.5¢ midpoint reflects a few resting orders, not a deep consensus. Thin markets can be persistently mispriced, but they also make small edges unreliable and expensive to exit — which is why our threshold here is high.
For related single-contract analysis on this site, see our Deep Dives on the Russia Duma 2026 odds and the Bab el-Mandeb closure market, both of which face the same December 31, 2026 deadline structure, and our Israel election 2026 market write-up. Our Polymarket aliens market analysis covers the same thin-book failure mode.
This article is informational analysis only and is not betting or financial advice. Odds and prediction-market prices move constantly; every price quoted is a timestamped snapshot, not a live line. There is no such thing as a guaranteed bet — past results and model estimates do not guarantee outcomes. Do your own research.
Gambling carries financial risk and can be addictive. 18+ / 21+ depending on jurisdiction. Never bet more than you can afford to lose. If you or someone you know needs help, visit GamCare (UK), call 1-800-GAMBLER (US), or see our Responsible Gambling page.
