BY 
The problem with a record you can't check
I taught math before I moved into projections work, and grading was the least glamorous part of the job. You can't give a student credit for homework they say they finished at home. Sports handicapping runs on exactly that kind of credit. A capper posts a hot streak, and you're expected to take his word for the denominator.
The wins get posted loudly. The losses get deleted quietly, or never posted at all. Once a record is curated after the fact, it stops being a record and becomes an ad.
CAPTRACKER exists because this problem has a clean technical fix: take the grading away from the person being graded. Every pick on the platform is settled automatically against ESPN data, and no human gets a vote.
How settlement against ESPN data actually works
The mechanics are simpler than people expect. A pick enters the system with a timestamp and gets locked. When the game goes final, the system reads the result from ESPN's data and grades the pick, with pushes handled the way a sportsbook would handle them. Nobody types anything in.
That last sentence is the whole ballgame. Manual settlement means a person decides what counts, and people who sell picks have an incentive problem you don't need a statistics degree to spot. Automated settlement means the grade comes off the scoreboard. You can read the full pipeline on the methodology page, and I'd encourage you to, because a verification system you can't audit is just another claim.
Locked picks, the part everyone skips
Auto-grading only matters if the pick itself can't move. On CAPTRACKER, a pick is timestamped when it's submitted and can't be edited or deleted afterward. Not by the capper, and not by anyone running the site.
Think about what edit access does to a dataset. If losses can vanish, the win rate you're shown is a survivorship number, and survivorship numbers always flatter. My old analyst job beat this into me: the moment anyone can touch historical data, you have to treat the whole series as suspect. A locked, append-only record is the only kind worth analyzing.
How to read a verified record like an analyst
Start with sample size before you even glance at the percentage. A capper who goes 12-8 has told you almost nothing, because twenty picks is a coin rattling around in a cup. The same capper holding an edge across several hundred locked picks is a genuinely interesting object. Small samples run hot all the time, and variance doesn't care about anyone's feelings.
Then anchor everything to break-even math. At standard -110 juice you need to win 52.38% of your bets just to tread water, because you're risking 110 to win 100 every time. A capper bragging about winning 52% of his plays is, after juice, losing money slowly. That single number filters out most of the noise you'll ever see on a sales page.
Units won and ROI tell you more than raw win percentage, which is why the CAPTRACKER leaderboard ranks cappers on those instead. Someone who lives on underdogs can sit below 50% and still be up units, while a favorite-heavy capper can win a lot of bets and bleed anyway. Percentage without price context is a vanity stat. I'll take a modest ROI over a shiny win rate every single time.
Convergence, or when independent cappers agree
With 900+ tracked handicapper profiles on the platform, disagreement is the normal state of things. So when independent cappers land on the same side of the same game, CAPTRACKER flags it as a convergence signal on the daily feed. Agreement between people who don't know each other and don't share a spreadsheet carries different information than one loud opinion.
I treat convergence the way I treated agreement between independent models in my analyst days: a reason to look closer, never a conclusion on its own. No signal replaces your own judgment, and nothing on any platform, this one included, guarantees a profit. What a good signal does is spend your attention wisely, and attention is the scarcest resource a bettor has.
Why free matters here
Everything I've described costs nothing. The leaderboard, the capper profiles, the daily feed, and the methodology writeup are all open, with no card on file. That's worth pausing on, because the standard business model in this industry is charging you for numbers nobody checked.
If you follow anyone's picks, free or paid, ask one question first: who graded this record, and could they have changed it after the fact? If the honest answer involves trusting the seller, walk away. If the answer is that the picks were locked before the games started and ESPN's data graded them automatically, you finally have something a careful person can work with.
Start with the verified leaderboard, sort by units won, check the sample sizes, and run the break-even math yourself. It's the good kind of homework.