After NBA Injury News, Do Big or Small Trades Move First?
In the 5 minutes after an official injury report is posted, we compared trades on Polymarket in the news direction: which shows up first, trades of $1,000 or more or smaller ones, and how often each follows the news.
- Result status
- Preliminary results
- Last updated
- Start date
- Duration
- Day 1 of 1
- Research team
- Mysports.AI Research Team
- Plan registered
Big vs Small Trades After NBA Injury News: Key Findings
555 news events in 403 games, 2025-26 regular season; 95% range −0.9 to +8.5 seconds, which includes 0 = no difference can be seen (not the same as equally fast). The same method also shows a +10.6-second lean before any report.
- In practice the first small news-direction trade comes before the first big one (median 12 seconds; mean 49 seconds); only 27.8% of first news-direction orders are big, versus 26.8% with shuffled amounts — about the same.
- Direction (descriptive, not tested): in the first minute, 67.7% of big trades (63.9–71.6%) and 49.3% of small ones (47.2–51.4%) go with the news; within 5 minutes 60.9% vs 49.8%; before any report 52.1% vs 49.3%.
- Frequency (descriptive, not tested): in the 90 seconds after the report, big news-direction trades happen 8.3 times (7.2–9.7) as often per minute as 60–15 minutes before it, small ones 2.1 times (2.0–2.3). By count, news-direction trades are still mostly small (1,088 big vs 4,996 small).
- The direction is not consistent across settings: for $5,000 and up, big trades arrive about 6 seconds earlier, relatively (−6.0, range −11.2 to −0.7); with a $250 cut-off, small trades arrive about 7 seconds earlier instead (+7.3, +2.4 to +12.3). This data cannot give a stable answer to “who comes first”.
- Trades against the news: big ones arrive clearly later relative to small ones (+16.7 seconds, +10.9 to +22.7), consistent with big trades mostly going with the news.
How to use this yourself
Open the NBA’s official injury report page (official.nba.com), click the latest report of the day (a PDF), compare it with the previous one and look for key players whose “Current Status” changed. Times on the report are US Eastern. Open source ↗
Work out the news direction (a simplified example): Questionable changing to Out is bad for his team, so the news direction is against that team; changing to Available is the opposite. This study uses R1’s definition, which has extra rules when one team has several changes within 15 minutes; this is only an intuitive example.
Use this page’s numbers: on past average, in the first minute after the report about 68% of trades of $1,000 or more went with the news, and about 49% of smaller trades did — roughly half and half. This is an average over many games and cannot be used to judge any single trade.
Do not use “whether big or small trades come first” as a signal: the first news-direction trade is usually small only because small trades are more common; accounting for that, no difference in speed can be seen, and the result changes with the cut-off.
These are past averages from the 2025-26 regular season and Polymarket’s first-generation contracts, with no guarantee for any game or trade; the direction shares are descriptive and were not formally tested. We cannot tell who placed the trades, so a big trade is not the same as a “whale”. This page is not betting advice.
In plain words
Why study this?
Answers whether big or small trades come in first when news breaks. When you look at trades after the news, it helps to know that the first one being small is normal, and that big trades do not all go the same way.
What do we want to know?
Our study R1 measured, trade by trade, that after the NBA’s official injury report is posted, Polymarket trade prices complete half of their move in about 90 seconds. In that time, is it big money or small money that trades in the news direction first?
How do we study it?
We used the same 1,456 injury news events as R1 (2025-26 regular season, 763 games) and public trade-by-trade data. Trades in the same second and of the same kind were combined into one order; $1,000 or more counts as big. Looking only at orders in the news direction within 5 minutes of the report, we compared when the first big and the first small order appeared. Because small orders are far more common (about 1 big for every 16), we shuffled the dollar amounts 2,000 times as a comparison, to see how the real order of arrival differs from one where size and timing are unrelated. The plan and the analysis code were timestamped before we looked at any results.
What have we found?
The first trade in the news direction is usually a small one (the first small one comes 12 seconds before the first big one, median), but that is because small trades are more common. Compared with the shuffled amounts, big trades arrive 3.7 seconds later relative to small ones, with a range of −0.9 to +8.5 seconds that includes 0, so no difference in speed can be seen; with other cut-offs ($5,000 or $250) the direction is not consistent, and the same method also leans +10.6 seconds before any report. The direction is clearer (descriptive, not tested): in the first minute after the report, about 68% of big trades and about 49% of small ones go in the news direction; big trades in the news direction happen 8.3 times as often per minute as 60–15 minutes before the report, small ones 2.1 times.
What does this mean?
After news, the first trade being a small one is normal and does not mean small traders react faster; big trades do not clearly jump ahead either. The difference is in direction: big trades go in the news direction more often than small ones, and their frequency rises more; small trades split roughly half and half. We cannot tell who placed the trades, so a big trade is not the same as a “whale”.
What can’t we say yet?
What we cannot tell yet is who comes first: the direction changes with the cut-off, and the same method also shows a lean before any report, so this data cannot give a stable answer. We also cannot tell who placed the trades, and we cannot see orders being posted or cancelled; the first to react may be someone pulling old quotes. There is no comparable trade-level data for the new season, so we cannot test it again.
Terms used on this page
- News direction
- Trading in the direction of the news. If a team’s key player goes from Questionable to Out, that is bad for the team, so the news direction is against that team (selling it or buying the opponent).
- One order
- Each row in the data is one resting order being filled. We add up rows in the same market, same second and same kind of trade as one order; we cannot tell who placed it, and several people in the same second may be combined.
- Big and small trades
- An order of $1,000 or more is big; less is small. Around the events, the median order is only $12 and about 6% are $1,000 or more.
- Shuffled-amount comparison
- For each news event, we randomly shuffle the dollar amounts of the news-direction trades in the 5 minutes (keeping their times) 2,000 times, to see how much later the first big order would come than the first small one if size and timing were unrelated, and compare with what actually happened.
- Range
- A 95% interval from resampling whole games 2,000 times. We only say there is a difference when the range does not include 0.
Detailed results
Seconds the first big news-direction trade arrives later (+) or earlier (−) than the first small one, relative to shuffled amounts
2025-26 regular season. The first row is the pre-registered main result; the others are pre-listed alternative calculations and comparisons (descriptive only). Negative = big trades move first, relatively.
The main result’s range is −0.9 to +8.5 seconds, which includes 0 = no difference can be seen. The other rows are pre-listed alternative calculations and comparisons, for reference only. This is research, not betting advice.
HypothesesFor deeper reading
- H1Big trades move first, relatively → no difference can be seen: +3.7 seconds (range −0.9 to +8.5), and the same method also leans +10.6 seconds before any report. This does not mean they are equally fastInsufficient evidence
- D1In the first minute after the report, 68% of big trades and 49% of small trades go in the news direction (descriptive, not tested)Exploratory
- D2In the 90 seconds after the report, big trades in the news direction happen 8.3 times as often per minute as 60–15 minutes before it; small ones 2.1 times (descriptive, not tested)Exploratory
Big vs Small Trades After Injury News: Research Methods
Research questions
- After accounting for small trades simply being more common, does the first big trade in the news direction come earlier or later than the first small one?
- After the report, what share of big and of small trades go in the news direction? (descriptive, not tested)
- How many times more often than usual do big and small trades in the news direction happen? (descriptive, not tested)
Method summary
- There is one main result: for each news event, the actual gap in seconds (first big minus first small order in the news direction) minus the same gap under shuffled amounts, averaged. Negative = big trades move first, relatively. Only the 555 events with both a big and a small news-direction order within 5 minutes are used.
- This is the timing of big relative to small trades: it could come from big ones being earlier or small ones being later, so we also split it (big +3.2 seconds, small −0.5 seconds; no difference can be seen in either).
- “Report time” is when the official PDF was posted; trades in that very second count as before the report (as in R1).
Data and limitations
Data we use (all public and free)
- Polymarket-v1 trade-level data (Boka Qin, Rui Yang, arXiv 2606.04217), CC BY 4.0; we used only trade amount, side and time, and this page shows only aggregated numbers, with no accounts
- NBA official injury reports (each with an upload time accurate to the second; the 2025-26 regular-season events compiled in R1)
Limitations
- Retrospective, not blind: R1 had already measured reaction speed and the first-minute direction tilt on the same events and trades (stronger by dollars than by count); big trades leaning toward the news fits that result and is not an independent new finding.
- We cannot identify people or complete orders: one person splitting an order over several seconds looks like several small orders, and several people in the same second are combined into one.
- Before any report (30 to 25 minutes before), the same method also leans +10.6 seconds, so the comparison is not perfectly neutral; the main number cannot be read directly as an effect of the news.
- Only trades that take liquidity are seen; posting and cancelling orders are not, and the first to react may be someone pulling old quotes.
- The dataset’s trade volume is about half the market’s published volume (reason not verified); conclusions apply only to trades in the dataset. Only the 2025-26 regular season; no comparable trade-level data exists for the new season.
Latest progress
Latest:Main results are in: after accounting for small trades being more common, no difference can be seen in whether big or small trades move first; in the first minute about 70% of big trades go with the news, small ones about half and half. The plan and analysis code were timestamped first, and results were reviewed by an AI code-review tool
Next step:No comparable trade-level data exists for the new season, so there is no forward test; we will reconsider if new public trade-level data appears
Study timeline
- Data gate passed (counts and amounts only, no direction or timing looked at)
- Plan finalized and timestamped after two rounds of AI-tool code review; analysis code tested on synthetic data and timestamped before running
- Main results in; one round of AI-tool code review (numbers unchanged; some wording fixes)
Update log
Version history
- v1.0-results
Main result: no difference can be seen in whether big or small trades move first (+3.7 seconds, range includes 0); descriptive: in the first minute about 68% of big and 49% of small trades go with the news
- v1.0
Plan finalized and timestamped (T0 = 2026-11-16 18:19 UTC); retrospective, not blind; no forward test because no comparable trade-level data exists for the new season
Study record
b795c7c62eee96e35efda3703be407e08f8a2ba98feacf48a43f503809d032a045766c27eb99a9376f273d0ba12bde2bd8585bfdb6522124dec65fd474b3ca0b47fd2099d012fdde3d0becfa7ffd77e73be352e5116b28dd0c66c22767ed0c578a5f7360641b6679bb9ba8eb70eeab59afcd09be8a7ec045bd8d447138c3c3dfThis page is a research log for reference only and is not betting advice or a guarantee of profit.