Are Popular Teams Overpriced?
Do popular teams like the Lakers, Warriors, Yankees and Dodgers win less often than other teams at the same odds? We also compare Pinnacle and Taiwan Sports Lottery.
- Result status
- Preliminary results
- Last updated
- Start date
- Duration
- Day 22 of 7
- Research team
- Mysports.AI Research Team
- Plan registered
Popular Team Odds: Key Findings
The range runs from 4.7 lower to 2.3 higher; neither method can rule out luck, so the conclusion is “no sign.” Based on 3,354 games in the 2025-26 NBA and 2026 MLB regular seasons, 913 with exactly one popular team, after removing favorite–underdog bias.
- Extension (pre-registered, prices fetched only after timestamping): in the NHL and NFL, popular teams won 6.9 percentage points less than teams at the same odds (95% interval 3.0–10.7), which meets our pre-set rule. See “Extended analysis” below.
- No sign in the NBA or MLB separately; other popularity definitions (top 3, top 8, Wikipedia lists) point in inconsistent directions.
- Taiwan Sports Lottery: backing popular teams returns NT$85.2 per NT$100 on average versus NT$89.1 for their opponents; net of the lottery’s usual favorite–underdog margin difference, popular teams return about NT$3 less, but this is not conclusive (only 66 popular-team MLB games in Sept–Oct 2026).
- Polymarket and Pinnacle price popular teams almost the same (about 0.1 percentage point apart); no clear difference.
- No sign does not mean the market is right: this data can only rule out being overpriced by more than about 4.7 points or underpriced by more than about 2.3.
How to use this yourself
Open the official Taiwan Sports Lottery website (sportslottery.com.tw) and find the game in its match list. In the list, the team in the first position is the away team and the team in the second position is the home team. Open source ↗
Look at each team’s odds under 不讓分 (picking who wins, i.e. the moneyline). The odds show how much you get back for every NT$1 if you are right (including your stake); the team with the lower odds is the more favored one.
Check whether either team is one of this study’s “popular teams” (the 5 teams whose games had the largest trading volume on Polymarket the season before): for NBA 2025-26 they are the Lakers, Warriors, Celtics, Bucks and Clippers; for MLB 2026, the Yankees, Dodgers, Phillies, Orioles and Mets. Trading volume is only a stand-in for popularity and doesn’t necessarily mean these teams have the most fans.
NBA and MLB: don’t assume a popular team’s odds have been pushed down, or that betting on its opponent is better value, just because it is popular. At the same odds, popular teams’ actual win rate was on average only 1.2 percentage points lower, with a range from 4.7 lower to 2.3 higher, so it can’t be told apart from luck (1 percentage point is, for example, the gap between 50% and 51%).
On Taiwan Sports Lottery, betting on the popular team returned an expected NT$85.2 per NT$100 on average, versus NT$89.1 for its opponent. But after removing the part that comes from Taiwan Sports Lottery already taking a different margin on favorites and underdogs, the popular team was only about NT$3 worse, and with just 66 games this is not certain. These are averages based on win chances estimated from Pinnacle (an international bookmaker), not records of real bets.
Ice hockey (NHL) and American football (NFL) came out differently: in a check whose rules were written in advance, popular teams (again the 5 with the most trading volume the season before, such as the Kansas City Chiefs) won 6.9 percentage points less often than other teams at the same odds, which counted as confirmed under the rules. But the data covers only 10 popular teams with one season each, and with a different definition of popularity the gap disappears, so treat it only as a hypothesis worth tracking (to be rechecked after the 2026-27 season), not as a reason to bet against popular teams.
The main result uses Polymarket odds from 20 to 5 minutes before the start (a price of $0.60 ≈ a 60% chance ≈ odds of about 1.67), across 3,354 NBA 2025-26 and MLB 2026 regular-season games, 913 of which had exactly one popular team; Taiwan Sports Lottery was only compared for MLB in 2026/9–10. With other definitions of popularity (top 3, top 8, a Wikipedia-based list) the direction was not consistent, so don’t treat “popular teams are always overpriced” as a signal; and “overpriced” doesn’t mean betting against them makes money. These are past averages and guarantee nothing for any single game.
In plain words
Why study this?
A common claim is that everyone loves betting on popular teams, so their odds get pushed down and backing them is worse value. If true, a model using market odds should correct for it; if not, that is one myth fewer.
What do we want to know?
Teams like the Lakers, Warriors, Yankees and Dodgers draw huge attention. A common claim is that everyone bets on them, so their odds get pushed down and backing them is worse value. We check with real data: at the same odds, do popular teams win less often than other teams?
How do we study it?
“Popular teams” are fixed in advance, not picked afterwards: the five teams whose games had the most Polymarket trading volume last season (NBA: Lakers, Warriors, Celtics, Bucks, Clippers; MLB: Yankees, Dodgers, Phillies, Orioles, Mets). Favorites and underdogs may be mispriced in their own right, so we always compare at the same price: popular teams versus ordinary games at the same odds. We use Polymarket odds 20–5 minutes before tip-off for the 2025-26 NBA and 2026 MLB regular seasons, and also compare how Polymarket, Taiwan Sports Lottery and Pinnacle price popular teams.
What have we found?
No sign that popular teams are overpriced or underpriced. At the same odds they won 1.2 percentage points less on average, but the range runs from 4.7 less to 2.3 more, so luck cannot be ruled out; the NBA and MLB look the same separately. Compared with Pinnacle, neither Taiwan Sports Lottery nor Polymarket shows a clear difference after accounting for price level.
The popularity rule changed once—all disclosed
We first chose popular teams by English Wikipedia page views, but found it measured news events (such as a team relocating or renaming). Before looking at any odds or results, we switched to the volume rule above and timestamped it first. The MLB list was also recomputed twice because of code errors; the old lists are kept public.
What can’t we say yet?
High volume may reflect strong teams, news or good time slots rather than fans, so any difference will be described as “last season’s high-volume teams,” without claiming a cause. We also do not answer whether betting against them would make money.
Terms used on this page
- Polymarket
- A prediction platform for trading game outcomes; prices run from $0 to $1. A price of $0.60 roughly means traders give that side about a 60% chance—not necessarily the true probability. Price converts directly to odds: odds = 1 ÷ price.
- Popular team
- The five teams whose games had the most Polymarket trading volume last season (adjusting for opponent and month). It can be decided before the season and uses no results from the season being tested.
- Same-price comparison
- Comparing popular teams with ordinary games at the same odds, so that favorite–underdog bias is not mistaken for a popularity effect.
- Percentage point
- The difference between two probabilities. For example, 50% to 51% is an increase of 1 percentage point.
Detailed results
Expected return per NT$100 (Taiwan Sports Lottery, MLB Sept–Oct 2026)
Using Pinnacle’s margin-free win estimates; games with exactly one popular team only.
Secondary test; no clear difference after netting out the usual favorite–underdog margin (adjusted p = 0.37) | Not actual betting records.
HypothesesFor deeper reading
- H1At the same odds, popular teams’ actual win rate differs from other teams → no sign (1.2 points lower, range 4.7 lower to 2.3 higher)Insufficient evidence
- H2Polymarket prices popular teams differently from Pinnacle → no sign (about 0.1 point apart)Insufficient evidence
- H3On Taiwan Sports Lottery, backing popular teams is worse or better value → no sign (about NT$3 less per NT$100, not conclusive)Insufficient evidence
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.
- We re-ran the test on two new leagues, NHL 2025-26 and NFL 2025 (the plan was time-stamped before any prices were downloaded): at the same odds, popular teams (the 5 teams with the most Polymarket volume the previous season) won an estimated 6.9 percentage points less often than other teams at the same odds (95% interval 3.0 to 10.7), and both pre-specified tests passed — in these two leagues, popular teams' odds were too short — they were overpriced. Without adjusting for price level, popular teams' average implied win probability was 53% (a Polymarket price of $0.53, odds about 1.89), but they won only 46% of the time.
- All 10 popular teams won less often than their odds implied — for example, the Kansas City Chiefs averaged a 64% implied win probability but won only 29% (14 games). Neither league passes on its own, however; the result passes only because both point the same way.
- The earlier hint (seen only after the fact in NBA and MLB) that popular teams are especially overpriced at home was tested again under rules written in advance, and it showed up in the new data. But in the new data popular teams were also overpriced on the road, so the pattern looks like popular teams being overpriced in general, not only at home.
- Pooling all four leagues (NBA, MLB, NHL, NFL), popular teams won 3.1 percentage points less often than other teams at the same odds (95% interval 0.3 to 5.8). NBA and MLB alone show no clear effect, and that half of the data had been seen before, so this pooled figure is a description, not a new test.
Limits of this analysis
- Only 10 popular team-seasons; the NFL has just 68 popular-team games, so its effect size is very unstable (95% interval 5.6 to 26.6 percentage points).
- The result depends on how "popular" is defined: using the top 5 by average betting volume, or treating each team's volume score as a continuous measure, shows no clear difference.
- We cannot tell whether this is a feature of hockey and football or of the 2025-26 season; lists for 2026-27 are already frozen and will be re-tested after the season.
- Prices exclude bid-ask spreads and fees; "overpriced" does not mean betting the other side is profitable. We also do not infer who pushed popular teams' prices up, or why.
- We also tried social-media follower counts as a popularity measure (post-hoc exploration): free, cleanly licensed public data is very incomplete (X counts stop in 2023 and miss the Warriors, Yankees and Dodgers); only MLB YouTube subscriber counts were usable, and they correlate moderately (0.56) with the volume ranking.
Popular Teams: Research Methods
Method summary
- The popularity list was fixed before looking at any odds or results; the rule changed once and was recomputed twice for code errors, all disclosed.
- Always compare at the same price: actual win rates of popular teams versus ordinary games at the same odds.
- One primary test (H1); the other two are corrected together for multiple testing.
- Robustness: top 3, top 8, the Wikipedia list, and dropping one team at a time.
Data and limitations
Data we use (all public and free)
- Polymarket single-game odds and results, 2025-26 NBA and 2026 MLB regular seasons
- Extension 1: Polymarket single-game odds for NHL 2025-26 and NFL 2025, with official scores
- Last season’s Polymarket trading volume (to set the popularity list)
- Pinnacle and Taiwan Sports Lottery closing odds (Mysports.AI odds system, MLB 2026/9/10–10/19)
- Post-hoc exploration: Wikidata social-media follower counts (CC0)
Limitations
- High volume does not necessarily mean more fans; it may reflect strong teams or news. We only describe “last season’s high-volume teams.”
- The list was computed after the tested season, so it is not a true out-of-sample test; the 2026-27 season will re-test.
- The Taiwan Sports Lottery and Pinnacle comparison covers only MLB in Sept–Oct 2026.
- No betting advice; spreads and fees are not computed.
Latest progress
Latest：Extension 1 (pre-registered): re-tested on the NHL and NFL, popular teams won 6.9 percentage points less often than other teams at the same odds, which meets our pre-set rule
Next step：Re-test with the frozen lists in NBA and NHL 2026-27, NFL 2026 and MLB 2027 (both the original per-group checks and the combined four-league check are pre-registered)
Study timeline
- Novelty and data checked; plan finalized and timestamped after two AI-tool code reviews
- Analysis done; corrected after AI-tool code review
- Re-test with new lists in the 2026-27 NBA and 2027 MLB seasonsIn progress
- Pre-registered a combined new-season check across four leagues (NBA and NHL 2026-27, NFL 2026, MLB 2027; using the frozen lists)
Update log
Version history
- v1.0-results
Main results; the analysis code was timestamped before running and 6 AI-tool review points were fixed
- v1.0
Plan finalized and timestamped (T0 = 2026-10-21 11:06 UTC); the popularity rule changed once before seeing any odds, disclosed
Study record
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 page is a research log for reference only and is not betting advice or a guarantee of profit.