BY 
This week's top verified leaderboard report starts with one wild number
Scroll to the top of CAPTRACKER's leaderboard this week and one line jumps out. TheLineTamer is running a 94.8% ROI in MLB. Every pick on the site gets timestamped and locked before the game, then auto-settled against ESPN results the second the final score posts, so the number itself is real. Slide your eyes one column over, though, and you'll see the sample behind it: 10 picks, seven wins and three losses.
That gap, a huge percentage sitting on a tiny pick count, is basically the whole story of this week's report. Compare it to X_Underscore_X, who leads the football board at 27-17-7 over 51 picks, +389.8 units, and a 51.6% ROI. Same idea, five times the sample, and the number still holds up under the weight. That's the difference between a hot streak and something closer to a track record.
A few other names on the board show the same shape this week. Jared2338 sits at 8-5 over 13 picks for a 71.2% ROI. Bugleboy98 is 9-2 over 11 picks at 76.7%. Initial_Increase_242 is 7-4 over 11 picks at 55.0%.
Numbers like that grab attention fast, and they should. Attention just isn't the same thing as evidence, not on its own.
A hot week on the leaderboard isn't a full season
CAPTRACKER also tracks a season-long pool of results, and it only ranks cappers on the season list once they've reached 15 settled picks. Ten picks can turn on one bad Tuesday. Fifteen starts to mean something more, and thirty or fifty means quite a bit more than that.
Run this week's hot names against that season list and the picture splits into two groups. Timely-Conclusion532, wezvidz, and Certain-Challenge202 all show up in both places alongside X_Underscore_X, with identical records in each column. Their big week and their season are the same sample, just viewed at different zoom levels.
TheLineTamer, Jared2338, Bugleboy98, and Initial_Increase_242 don't appear on the season list yet, because none of them have crossed the 15-pick threshold. That doesn't make their numbers fake. It means the sample hasn't had enough time to prove itself in either direction.
There's a quieter case worth naming too. nbagamblingpodcast doesn't show up on this week's hot list at all, but sits at 38-31 over 69 picks for the season, a modest +13.4 units and a +19.5% ROI. None of that is flashy. It's also a lot harder to argue with 69 picks than with 10, no matter how the percentage compares.
The math mistake hiding inside a small sample ROI
The mechanic underneath all of this is simple math. At standard -110 odds, break-even sits at 52.38%. Win half your picks and the vig eats you alive. Win a bit more than half, over a real sample, and you've actually got something worth following.
Small samples don't average out that fast, though. Flip one result in a 10-pick week and the win rate jumps or drops by ten full percentage points in an instant. Flip one result in a 121-pick season, like Timely-Conclusion532's current 83-38 run, and the number barely twitches. Same coin, wildly different amount of noise riding on it.
Nobody running hot right now did anything wrong. That's just how variance behaves at low volume, and it's exactly why CAPTRACKER separates the weekly window from the season pool instead of blending them into one score.
I'll admit the rookie in me used to circle the highest ROI on the page and stop reading right there. It took a few too many small heaters cooling off fast for me to learn that the pick count matters more than the percent sitting next to it.
What CAPTRACKER's full handicapper pool says about base rates
Zoom all the way out and CAPTRACKER's tracked pool includes 1,185 cappers and 7,135 settled picks. Of those, 615, or 52%, are currently profitable. The median ROI across that entire pool sits at +8.1%, which tells you the middle of the pack is barely above water.
Filter down to the 82 cappers with 15 or more settled picks and the number ticks up a little: 44 of 82, or 54%, are profitable. That's close to a coin flip, tilted slightly toward the cappers who've actually built a real sample. It's not a wide gap, but it points the right direction.
Volume alone doesn't fix that either. Three of the highest-volume cappers in the pool make the point on their own.
- SportsBettingPicks: 348-310-5 over 663 picks, +0.2 units, 0.0% ROI
- GuyBostonSports: 267-300-3 over 570 picks, -57.6 units, -10.1% ROI
- CallingOurShot: 70-74-1 over 145 picks, -13.8 units, -9.5% ROI
Hundreds of settled picks and none of that adds up to an edge. A big sample proves a record is real. It doesn't automatically make that record good.
How to read next week's leaderboard report without getting fooled
None of this means skip the weekly report or write off anyone with a short track record. It means reading the pick count before the percent next to it. The leaderboard shows both columns side by side for exactly that reason, and it costs nothing to check both before you trust either.
Watch for convergence too, the cases where independent cappers land on the same side of a game without coordinating with each other. One capper on a small sample is a single data point. Several agreeing at once, across different sports and different books, is a signal worth more of your attention.
If you want to watch picks lock and settle in real time instead of trusting a screenshot, the daily feed is where that happens live. If a term on the leaderboard isn't clear, the glossary spells out how units and ROI actually get calculated, and the methodology page walks through how auto-settlement against ESPN works from pick to final score.
CAPTRACKER's leaderboard is free to browse, covers more than 900 tracked handicapper profiles, and settles every pick automatically against ESPN data, no cherry-picked screenshots, no quietly deleted losers. Check the pick count before you check the percent, and this week's report will tell you a lot more than the top line alone ever could.