When News About a Key NBA Player Breaks, Do the Odds Move Right Away?
A study of sudden player-status news: when a key player is suddenly ruled out or cleared to play, how fast do the odds on the prediction platform Polymarket change?
- Result status
- Preliminary results
- Last updated
- Start date
- Duration
- Day 23 of 14
- Research team
- Mysports.AI Research Team
- Plan registered
NBA Injury News and Odds: Key Findings
Based on the average of 1,460 key-player status reports in the 2025-26 season. Not every report works this way; after some, the odds barely moved.
- Much of the change starts before the official report: about 40% of the whole change happens within 1 hour before it (about half when re-measured trade by trade at exact seconds). News may have come out earlier, or there may be other reasons; we can’t confirm which this time.
- 15 minutes after the report, the odds still move a little further the same way, but so little that trading costs would wipe out any gain.
- How big is the change? From 1 hour before to 3 hours after the report, the win chance implied by the odds moves only about 1 percentage point on average, for example from 50% to 49%.
- This season the NBA switched to updating its injury list every 15 minutes. “Moving very early” seems to happen less now, but the evidence is not strong enough to be sure.
How to use this yourself
Open the NBA’s official injury report page (official.nba.com) and click the latest report of the day (a PDF). Times on the report are US Eastern; since the mid-2025-26 rule change, a new report comes out roughly every 15 minutes. Open source ↗
Compare the latest report with the previous one and look for key players whose “Current Status” changed, for example from Questionable to Out or to Available (cleared to play). This study only counts players averaging 20+ minutes over their last 10 games.
The odds react fast: about 80% (82%) of the change after the report happens in the first 15 minutes, and measured trade by trade, half of it is done about 90 seconds after the PDF goes up. So if you see the report more than 15 minutes after it was posted, the odds have mostly already taken the news in.
Much of the change happens before the report comes out: about 40% of the whole move falls in the hour before the official report (about half when recalculated trade by trade), and more so for the most important players. Waiting for the official report does not necessarily mean you know before anyone else.
Don’t expect big moves: from 1 hour before to 3 hours after the report, the implied win chance changes by only about 1 percentage point on average (for example, from 50% to 49%), and for about 58% of the news the odds did not move at all in the 15 minutes after the report. After those 15 minutes the odds still drift a little further in the same direction, but it is so small that trading costs wipe it out.
For news that a player is now Out, only about 75% of the move is done 15 minutes after the report; after that the implied win chance moves another 0.27 percentage points on average in the same direction (odds of about 2.50 becoming 2.52). This was added after seeing the results and will be rechecked in the 2026-27 season, so don’t treat it as a rule yet.
These results come from 1,460 news items about key players in the 2025-26 regular season, using Polymarket odds (a price of $0.60 ≈ a 60% chance ≈ odds of about 1.67); we did not measure how fast odds react elsewhere, such as Taiwan Sports Lottery. The speed figures depend a lot on how they are measured: using odds taken every 5 minutes gives different numbers from using every single trade (for example, about 40% versus about half for the hour before the report), so treat “15 minutes” and “90 seconds” as rough guides only. They are averages across many news items, not what happens every time, and guarantee nothing for any single game. Don’t treat “the odds are still moving after the report” as a signal to follow.
In plain words
Why study this?
Injury news is the biggest information shock before a game. If the odds lag behind, there is a gap between the news and the odds change that is worth studying. If they adjust all at once, it means the odds on prediction platforms reflect news faster than you might think.
What do we want to know?
If a key player suddenly can’t play, people may think the team is less likely to win. We want to know: after the NBA officially announces whether a player can play, do the odds on Polymarket change right away, or slowly over time?
How do we study it?
Polymarket is a prediction platform where people buy and sell shares in game results. Take “the Lakers win tonight”: the price is between $0 and $1. If the Lakers win, each share pays back $1; if they lose, it pays $0. So a price of $0.60 roughly means the people trading think, as a group, that the Lakers have about a 60% chance to win. That is the crowd’s view, not necessarily the true chance; when news changes that view, the odds change too. A Polymarket price converts directly to odds: a price of $0.60 means about a 60% win chance and odds of about 1.67 (odds = 1 ÷ price). We collected status changes for key players in the 2025-26 season (for example, from “Questionable” to “Out”), analyzed 1,460 of them, recorded the odds every 5 minutes, and looked at how the odds moved before and after each announcement.
What have we found?
Adding up all the news: the odds change after the report is mostly in the first 15 minutes (looking only at the change after the report, the first 15 minutes make up about 80% of it), with only small changes after that. Also, about 40% of the whole odds change we saw happened within 1 hour before the official report (about half when re-measured trade by trade at exact seconds). Reporters or other sources may have spread the news earlier, or there may be other reasons.
What does this mean?
If you wait for the official report before acting, you won’t necessarily know the news before others do. For our models, information on whether a player can play should go in as early and as accurately as possible.
What can’t we say yet?
“15 minutes” comes from adding up many news items; it does not mean every game is the same. For about 58% of the news items, the odds did not move at all in the 15 minutes after the report. We only have one odds reading every 5 minutes, so we can’t see what happens within those 5 minutes. The small change after 15 minutes is not clear enough once we leave out news items where the odds never moved. This study also does not prove that trading right after the news makes money.
Terms used on this page
- Polymarket
- A prediction platform where people buy and sell shares in game results. Prices run from $0 to $1 and roughly show the chance the traders think an outcome has — not necessarily the true chance. A price converts directly to odds: odds = 1 ÷ price.
- Percentage point
- The gap between two percentages. For example, going from 50% to 51% is a rise of 1 percentage point.
- Toward the news
- If news makes a team more likely to win, its odds go down (the implied win chance goes up). That counts as “toward the news” and is recorded as a positive number; the opposite direction is negative.
- Lagging
- The odds keep moving a little in the same direction more than 15 minutes after the announcement.
- Sample
- The set of news items or games this study actually analyzed. “Supported in this sample” means it held in this data and still needs checking on new data.
Detailed results
- 011 hour before
Check whether the odds already move before the official announcement.
- 02Announcement
Separate the periods before and after the news.
- 0315 minutes after
Measure the initial change in odds after the announcement.
- 04Then up to 3 hours
Check for continued changes, stopping no later than 5 minutes before tip-off.
The curve below averages many news items; it is not an individual game. A change in odds does not establish that the news caused it.
How the team’s win chance moved before and after the news, on average (implied by the odds)
Starting 1 hour before the report. The green band is the first 15 minutes after it; after that, the line goes almost flat.
This chart was drawn after seeing the results, to show how the change unfolds｜Average of 1,460 news items, 2025-26 regular season｜Ends no later than 5 minutes before tip-off｜Looking only at the 504 items at least 3 hours 5 minutes before tip-off, the shape is the same: almost no movement after 15 minutes.
View detailed chart values
How much the team’s win chance changed before and after the news, on average (implied by the odds)
Units are percentage points: going from 50% to 51% is 1 percentage point. A positive number means the change went the way the news points, for example, a player is confirmed “Out” and the team’s win chance drops. Most of the change happens within 15 minutes of the report. “After 15 minutes” counts up to at most 3 hours after the report; if the game starts earlier, it counts up to 5 minutes before tip-off.
Calculated following the plan written in advance. On 2026/10/20, while using another AI tool to check the code line by line, we found a timing error, fixed it and recomputed (see “Update log”)｜1,460 news items on key players and 763 games, 2025-26 regular season｜Polymarket odds recorded every 5 minutes, converted to implied win chance.
HypothesesFor deeper reading
- H1The odds don’t finish adjusting all at once → holds under the method set in our plan: 15 minutes after the report, they still move a tiny bit further the same way on average (about 0.1 percentage points of implied win chance); but with a different data filter, it is no longer clearSupported in this sample
- H2Games with less money bet on them are more likely to lag → not enough evidence to tellInsufficient evidence
- H3The more surprising the news, the more likely the odds lag → not enough evidence to tellInsufficient evidence
- H4After the switch to updating the list every 15 minutes, “moving before the report” happens less → right direction, but not enough evidence to tellInsufficient evidence
- H5After the switch to updating the list every 15 minutes, the change after the report is different → not enough evidence to tellInsufficient evidence
- E1(Exploratory, added after seeing the results) How much the odds move before the official report → about 40% of the whole change happens within 1 hour before the report (about half when re-measured trade by trade at exact seconds)Exploratory
Further analysis
Further analysis done after the main results. Items marked “post-hoc” were added after seeing the data and are for reference only; items marked “pre-registered” or “Extension 1” had their rules written and timestamped before anything was calculated.
- "Ruled out" news keeps moving a little after the announcement (post-hoc exploration): for the 360 updates where a player became "Out", prices had completed only about 75% of the move 15 minutes after the official report, then drifted another 0.27 percentage points of implied win probability in the same direction (e.g. from 40.00% to 39.73%, odds about 2.50 to 2.52). This split was chosen after seeing the data; it will be re-tested on the 2026-27 season under rules written in advance.
- Net movement toward the news before the report is concentrated in the last 15 minutes (post-hoc exploration): after the NBA moved to a report every 15 minutes, about 81% of the net movement toward the news in the 2 hours before the report fell in the last 15 minutes; net movement 1–2 hours earlier was about zero. One possible explanation is that information spreads shortly before the PDF is posted, but the data cannot tell when the information appeared or who knew first.
- The more important the player, the more prices move before the report (post-hoc exploration): for the top third of players by value, implied win probability had already moved 0.79 percentage points on average in the hour before the report, about 42% of the whole move; for the bottom third, prices barely moved.
- (Trade-by-trade re-measurement: plan and code timestamped first, but the first results were already known, so this is not a blind test) Measured trade by trade after the official injury-report PDF goes up: within the first five minutes, Polymarket trade prices complete half of the post-release adjustment (already by the 90-second checkpoint; 95% interval 45 seconds to 4 minutes), and about 70% by 5 minutes. In the first minute, trades come more than 5 times as fast as in the 15 to 60 minutes before release.
- (Added after seeing the results.) Re-measuring 1,449 updates at exact seconds, the hour before release accounts for about half of the price move from 1 hour before to 3 hours after (or 5 minutes before tip-off, whichever is earlier); the 5-minute price data had put it at 40%. In our breakdown, about 8 of the 11 percentage points of difference correspond to sampling time: the 5-minute price used 'at release' was usually about 5 minutes old.
- (Trade-by-trade re-measurement) Since the league moved to reports every 15 minutes (December 2025), about 70% of the net move in the news direction during the 2 hours before release happens in the last 15 minutes, and this still holds trade by trade. This describes timing only; it cannot tell who knew first and does not identify any account.
Limits of this analysis
- These splits were chosen after seeing the results (more than 20 of them), so some may look special by chance; two of them are now pre-registered and will be re-tested on the 2026-27 season.
- The 81% figure is a share of net movement toward the news; earlier ups and downs cancel out, so it does not mean 81% of all price changes happened in the last 15 minutes.
- We wanted to re-check with minute-level bid/ask data from another platform (Kalshi), but its data terms restrict such use without written permission, so this was not done.
- This re-measures the same 2025-26 injury updates with finer data; it is not a blind test, since the earlier 5-minute results were already known. The starting sample is 1,460 updates; the main trade analysis includes 1,456; the 'last 15 minutes' share uses the 1,029 updates after the December 2025 change.
- Trade prices only update when someone trades and we have no order-book data, so quotes may move earlier or later; 'release' means the time the official PDF was posted, and reporters may break news earlier.
- Odds conversion: a Polymarket price of $0.60 means about a 60% chance, or decimal odds of about 1.67; percentage points refer to that implied chance. Trade data: Polymarket-v1 (Qin and Yang, CC BY 4.0). This study gives no betting advice.
NBA Injury News: Research Methods
Research questions
- Do the odds adjust all at once, or slide slowly and take a long time to finish?
- Or do they go too far at first and then pull back?
- Do the odds start moving before the official report comes out?
Method summary
- One piece of “injury news” = a change in a player’s status between two back-to-back official reports (for example, from “Questionable” to “Out”).
- Only important players count: those averaging 20+ minutes over their last 10 games.
- Main test: from 15 minutes after the official report until at most 3 hours after it (only up to 5 minutes before tip-off if the game starts earlier), do the odds keep moving in the same direction? We also compare before and after the NBA’s rule change, using 6 weeks on each side.
- We check five guesses at the same time. The more guesses you test, the easier it is to see a result that “seems to work” just by luck, so we set a stricter standard.
- Two ways to measure “how much is done within 15 minutes,” worked out after seeing the main results (using implied win chance): looking only at the change after the official report, the first 15 minutes make up about 82% of it (95% range 70%–99%); counting the 1 hour before the report as well, about 89% is done by 15 minutes after it (range 81%–99%). Both are shares of the total across all news items.
Data and limitations
Data we use (all public and free)
Limitations
- We only have one odds reading every 5 minutes, and no record of buy and sell prices. Very small changes may not show up, and trading costs can only be assumed.
- News times are the upload times of the official reports. Reporters may reveal the news earlier, so “moving before the report” does not mean anyone jumped the gun.
- When the NBA changed its rules, it changed both how often the report is updated and when it has to be filed, so we can only see the combined effect. There were too few playoff games, so they are not part of the conclusions.
Latest progress
Latest：Trade-by-trade re-measurement (plan timestamped first): about 90 seconds after the official PDF goes up, trade prices have completed half of the post-release adjustment; re-measured at exact seconds, the hour before release accounts for about half of the move (40% with 5-minute prices)
Next step：Check again on new games from the 2026-27 season; we compare the answers in mid-February 2027
Study timeline
- Research direction set
- Plan finalized and timestamped
- Downloading official injury reports
- Organizing injury news and rating how important each player is
- Analyzing odds changes before and after the news
- Comparing before and after the rule change
- Write up the results, compare them with R4, and let our prediction system update its data as soon as a player is confirmed outIn progress
- Checking again on new 2026-27 games: plan finalized, collecting data every dayIn progress
Data progress
Every possible publishing time slot is checked once. In the days around the switch from the old system to the new one, both file-name formats are tried, so some slots simply never had a report.
Update log
Version history
- ext1
Further analysis (post-hoc) and a 2026-27 prospective check plan, timestamped
- v1.1-forward
The plan to check again on new 2026-27 games is finalized and timestamped, with the rules for what counts as “still holds” written in advance
- v1.0-fix
When another AI tool checked our code, it found that the timing was calculated wrong for 18 news items posted “less than 20 minutes before tip-off.” We fixed it and recomputed everything. We also added the results under a different data filter, with their ranges.
- v1.0
Finished the main analysis of the 2025-26 regular season (fixes made along the way are all recorded under plan changes)
Study record
43f952dfab971c945e07378e677b5873c7b3dfe2d725aa6b334cb65fe2bfdcad6593506ba3db9d67cddc2c542d6e312ce145a5fe8a524a2b8219604781a5865c095c89b71d3bfafb39cbc01df914ae0b867100d4d85ca81871203319fd5471b42fe346a93e1b3754c6e7b538cc9db2dd68f1a179dc0b69cf82345544bf7e569e165ea8f2ca03ab4939d73cba68477ca17df4488f95a7264afc72028cee791ad21929802dec2425fd3c05066d8c1a1b977d438c3523b6e02ad414ae3a09e3cdd8d79a9431905a7a018cc50a685cbdb6547c628d37133bfbfdf90fd4af9d91e86e4c65029af39c4729821ed19b0af6a8857f61c0785af7ff5b1ae54b49dfe3fa4d8fa7a71ec4eaf48cee9e947505ec1b34d061f1442936f25ed6e488a60c2bbeb69cd01e8550b69eaf21e18e3cdc44e937d1f50173bffd93fb25960d0d973b6676This page is a research log for reference only and is not betting advice or a guarantee of profit.