We Analysed 21,777 Bets From Dropping Odds Alerts. Here's What Happened
Daniel Mercer5 min read
We get asked a lot whether following dropping odds actually makes money, or whether it just feels like it does. We had the data to check, so we did.
Every bet in this post was tracked by an OddsHub user straight from the Terminal's drop feed or top drops board between 3 August and 21 September 2026. We then settled them against the final score and compared each price to where the market closed. After cleaning, 21,777 settled bets were left, spread across 9,920 events.
Flat-stake ROI: +3.15% (+674 units at 1 unit per bet)
95% confidence interval: +0.9% to +5.5%
Bets that beat the closing line: 55.6%
Average closing line value: +1.5%
How we cleaned the data
We only counted bets tracked from a live drop. Manual entries and CSV imports were left out because there is no way to verify when they were placed. On top of that we removed:
- bets where the user set or changed the result by hand, so every result you see here was graded automatically
- duplicates, where someone tracked the same selection on the same event twice
- bets where the odds entered were more than double, or less than half, the price the drop showed. These are nearly always typos.
Everything is counted at 1 unit per bet. People stake in different currencies and amounts, and a handful of big stakes can drag an average in either direction. We also grouped the confidence interval by event, because lots of users often track the same drop and those bets are not independent.
Overall: +3.15% on flat stakes
Over the full sample, the bets returned +3.15%, or +674 units. For context, a typical recreational bettor gives up somewhere around 5% to 8% of their turnover to the bookmaker's margin. Turning that into a clear profit over more than 20,000 bets is a strong result, and even the bottom of the confidence interval sits above zero.
The curve climbs steadily across the whole period. Like any real betting record it has quieter spells along the way, but the direction never changes.
Results by closing line value
This is the chart we are most excited about. We grouped every bet by how much its price beat the closing no-vig price, then looked at the results in each group:
CLV 0 to 5%: +236 units, +4.5% ROI (5,332 bets)
CLV 5 to 10%: +245 units, +8.2% ROI (3,060 bets)
CLV 10 to 20%: +307 units, +14.4% ROI (2,168 bets)
CLV 20%+: +311 units, +30.0% ROI (1,047 bets)
Every bracket that beat the close made money, and the bigger the CLV, the bigger the return. Even better, the ROI in each bracket lands almost exactly on its average CLV. Bets in the 10% to 20% bracket averaged 13.8% CLV and returned 14.4%. Bets in the 5% to 10% bracket averaged 7.2% CLV and returned 8.2%. That is closing line value doing exactly what it is supposed to do, and it shows the edge in these drops is real.
The flip side proves the same point. Bets that ended up behind the closing price lost money, with the worst bracket down 16.4%. Get the price and you get paid. Miss it and you do not. If you want the background, our guide on reading line movement explains why the close is the benchmark that matters.
Across the whole sample, 55.6% of bets beat the close, with an average CLV of +1.5%. That share rose from 52% in the first full week to 60% in the week of 7 September as users got sharper about which drops they took.
EV at track time
Partway through this period, the Terminal started saving each bet's expected value against the sharp no-vig price at the moment it was tracked, so this chart covers 14,565 bets rather than the full sample.
Bets tracked at negative EV roughly broke even (+0.2%). Bets tracked at 10% EV or higher returned +13.6% and beat the close 83.6% of the time.
We also looked at bets where users got a price 15% to 50% higher than the reference bookmaker's, usually at a slower soft book that had not moved yet. About 75% of those beat the close and they returned close to +10%. Spot the drop, then grab the best number, and you have the whole recipe.
A note on variance
Results come in waves. The week of 7 September had the best CLV of the whole sample and finished roughly flat, while the week of 24 August returned +10.7%. That is exactly why we judge everything on CLV. Over enough bets, the price you get is what shows up in your results, and this data shows it clearly. We will keep running this analysis as the sample grows and publish the updates.
How does +3% compare?
It puts OddsHub users right in the range serious value bettors aim for, and ahead of what arbitrage typically makes per bet. The ROI is also running a little ahead of the +1.5% average CLV. The standout result is what happens when users are selective: bets that beat the close by 5% or more returned between +8% and +30%. Follow sharp drops, check them against the no-vig price, and the numbers speak for themselves.
Putting it into practice
- Filter on EV rather than drop size. Drop size on its own had a much weaker link to results.
- Compare prices before you bet. Most of the edge sits in the gap between the sharp price and a soft book that has not caught up yet.
- Track your bets and look at your CLV, not just your profit.
The OddsHub Dropping Odds Terminal handles all of that for you, with real-time sharp drops, EV against the no-vig price when you track, and automatic result and CLV grading. It is the same tool the bets in this post were tracked with.
Frequently asked questions
- Are these results from real bets?
- Yes. Every bet was tracked by a real OddsHub Terminal user from a live dropping odds alert and settled automatically by our grading system. Manual entries, imports and bets with hand-edited results were left out.
- Why use flat stakes instead of the stakes users actually placed?
- Users bet in different currencies and sizes, so a few large stakes can distort the average. Counting every bet at 1 unit shows how the selections themselves performed.
- Does beating the closing line guarantee profit?
- No single bet is guaranteed, but CLV is the best known predictor of long-term profit. In this sample, bets that beat the close by 10% to 20% returned +14.4% and bets that beat it by 20% or more returned +30.0%, while bets more than 10% behind the close lost 16.4%.