Most groups quietly bury the calls that went wrong and post only the winners. Gem Hunters logs every call, win or lose, and has done it since 2017. Here is the documented record, with the grading rule attached so you can check it yourself.
Almost every crypto call group does the same thing. They screenshot the one trade that went up 40x and post it everywhere. The trades that went to zero quietly disappear. You never see the full picture, so you can never tell if the group is actually good or just loud.
Gem Hunters does it the other way. Every call goes in the log, win or lose. We have kept it that way since 2017. The grading rule is published so you can apply it yourself. The calls carry timestamps so you can check what happened after they went out. This page lays the whole thing out. The win rate, how it is graded, how the calls are logged, and exactly how you verify it.
There is a reason most groups do not show you this. A real win rate is uncomfortable. It includes the calls that died, the ones that chopped you out, and the ones that looked great for an hour and then bled. Show all of that and the headline number drops. Most groups would rather you never do the math. We would rather hand you the math and let you decide. That is the whole bet behind this page.
A documented win rate is a record of the past. It is not a promise about your next trade. Read it that way, and read the risk note before you act on anything.
A win rate only means something if you know how a win is counted. The grading here is built to be hard on itself. Read these four rules first, because they are the reason the number stays honest.
A gem call counts as a win only if it gains more than 10%. Anything under that bar is logged as a loss, even if it ended slightly green. A small gain is not treated as a victory here.
Meme coins move faster and harder, so the bar is higher. A meme call has to gain more than 20% to count as a win. Below that, it goes in the log as a loss.
A leverage call is graded at the first take-profit target, not at some peak it touched for a single minute. We mark the result where a real person could have taken it, not at the best wick on the chart.
Breakeven trades and small gains are counted as losses, not wins. This is the rule most groups skip. It drags the headline number down, and that is the point. We would rather show a lower honest figure than a flattering one.
Why set the bar this high? Because a win rate is easy to inflate. Count every flicker of green as a win and you can push almost any record past 90%. That number would be useless. By grading against fixed targets and treating small moves as losses, the figure you read describes what a disciplined member could have actually captured. It is meant to be honest rather than impressive, and on most months those two things pull in opposite directions.
The leverage rule deserves a closer look, because it is where most records cheat. A leverage trade can spike, print a huge unrealized number for a few seconds, and then snap back. A dishonest group screenshots that spike and calls it the result. Here, a leverage call is marked at the first take-profit target, the level a real person could have closed at without perfect timing. That choice alone strips out a lot of fantasy gains. It also means the leverage wins on this record are gains someone could have actually banked, not wicks that existed for one candle.
The same discipline applies to how a call is even counted as a call. A call is a clear, posted entry with a token and a reason, time-stamped in the group. It is not a passing mention or a maybe. That keeps the denominator honest too. You cannot quietly drop a weak call out of the count after the fact, because it was logged the moment it went out. Both halves of the fraction, the wins and the total, are fixed in place by the timestamp.
These are a record of what has been called and logged. They describe the past, not a promise about any future trade.
Documented win rate across the graded sample.
Independently graded, April 2024 to November 2025.
Winning calls inside that graded sample.
Member wins posted in the wins log to date.
Of those wins came with a screenshot attached.
Calls logged year after year, the good and the bad.
One blended number can hide a lot. So here is the record split by year, graded under the same rule each time. The table below shows the win rate and the number of calls behind it for each period. More calls means a steadier figure, because one lucky run carries less weight when it sits inside hundreds of entries.
| Year | Win rate | Calls |
|---|---|---|
| 2024 | 71.9% | 246 |
| 2025 | 72.7% | 765 |
| 2026 (year to date) | 69.9% | 249 |
| Independently graded sample | 70.18% | 929 |
A few things are worth reading off this table. The win rate sits in a tight band, between roughly 70% and 73%, across three very different years of crypto. 2024 ran 246 calls. 2025 was the busiest year on record at 765 calls, and it held the highest rate at 72.7%. 2026 is still in progress, and so far it sits at 69.9% across 249 calls. The figure moves a little, the way an honest figure should, but it does not swing wildly.
The bottom row is the independently graded sample. It covers April 2024 through November 2025, which is why its call count of 929 does not simply add up the yearly rows above it. Those rows count calls by calendar year. The graded sample counts a fixed window inside that span. Across that window, 652 of 929 calls were winners, which is the 70.18% headline. Different slices of the same record, graded the same way, all land in the same neighborhood. That consistency is the point of breaking it out.
It also tells you something about sample size. A win rate built on 30 calls can be luck. A win rate built on hundreds is much harder to fake, because one hot streak gets diluted by everything around it. The 2025 row alone holds 765 calls. That is a big enough base that the 72.7% figure is doing real work, not coasting on a handful of moonshots. When you read a track record anywhere, the call count next to the percentage matters as much as the percentage itself. A high rate on a tiny sample tells you almost nothing. A steady rate across thousands of logged calls tells you a lot.
Notice too that the rate is not climbing toward some impossible number. It hovers near 70%. If a group ever shows you a 95% or 99% win rate, that is not skill, that is the grading rule counting every green tick as a victory. The fact that this record sits in the low 70s, year after year, under a rule that throws out small wins, is the strongest sign it is honest. An inflated number would look better and mean less.
A record is only as good as the way it is kept. Here is how a call moves from a post in the group to a line in the log you can check later.
The call goes out in the group in real time, so it carries the date and time it was made. There is no quiet edit after the fact.
Once the trade plays out, it is graded by the published rule. Gem over 10%, meme over 20%, leverage at the first take-profit. Small moves count as losses.
Members add their own wins to the wins log, almost always with a screenshot. 2,500+ wins have been posted, and nearly all of them have proof attached.
The calls that did not clear the bar are not deleted. They sit in the log next to the winners. That is what keeps the win rate honest.
Each graded call rolls into the year it belongs to. The yearly rates you see on this page come straight out of that running tally, not from a hand-picked highlight reel.
The reasoning behind the calls is on tape too. 370+ livestreams are logged, so you can watch how a setup was discussed before it became a call.
The grading sample tracks the calls. The wins log tracks what members actually did with them. These are two different things, and both matter.
Members have posted 2,500+ wins in the log to date. Of those, nearly all came with a screenshot attached. That is real people showing real positions, not a marketing graphic built after the fact. On top of that, the brand counts 550+ wins as team-verified, the subset that has been checked closely rather than simply posted.
Add it up over the years and the community has reported more than $20M in profits since 2017. None of that is a promise that you will see the same. It is a record of what members have logged. Some trades go to zero. Crypto does that. The honest version of a track record shows both sides, and that is the version on this page.
Why does the wins log matter on top of the graded sample? Because a call being right and a member making money are not the same event. A call can clear its bar while a member sits out, sizes too small, or sells too early. The wins log is where you see the second half of the story. Real entries, real exits, real screenshots. It is also why the screenshot rate matters so much. A win with proof attached is a claim you can inspect. A win with no proof is just a sentence. Almost every entry in the log is the kind you can check.
A number is only worth as much as your ability to check it. You do not have to take this record on faith. Here is how to test it for yourself.
Every call carries the time it was posted. Pull up the chart for that token at that moment and see what happened next. The call cannot move after the fact, so the timestamp is your anchor.
The grading rule is published on this page. Take any call, apply the 10% gem bar or the 20% meme bar, mark leverage at the first target, and grade it the way we do. You will get the same answer we did.
The wins log holds 2,500+ posted wins, almost all with screenshots. Scroll it. Read the dates. Cross-check a few screenshots against the chart history for that day.
Join the free group and watch new calls post in real time alongside 40,000+ members. The fastest way to trust a record is to watch the next entries go into it yourself.
The graded sample covers recent years in fine detail, but the group itself has been calling and logging since 2017. That is several full market cycles. Bull runs, deep winters, and the choppy stretches in between. A record that only exists during a green market is easy to build. One that survives a bear market and keeps getting kept is a different thing.
Across that span the community has logged more than $20M in member profits and grown past 40,000 members. The calls span gems, meme coins, and leverage setups, each graded under its own bar. The point of keeping it this long is simple. A track record is only convincing when it includes the years you would rather forget, and this one does. No junk, just gems, logged in good markets and bad.
A documented win rate shows how the calls have been graded over time. It is not a forecast. Past performance does not guarantee future returns, and crypto is high risk. Many low-cap tokens go to zero, and any single trade can lose. A 70% historical win rate still means roughly three in ten logged calls did not clear the bar. Never put in money you cannot afford to lose. Nothing here is financial advice.
The documented win rate is 70.18%, measured across 929 independently graded calls from April 2024 to November 2025. Of those, 652 were winning calls. The grading is deliberately conservative, so breakeven trades and small gains count as losses, not wins.
Every call is graded against a published rule. A gem call counts as a win only if it gains more than 10%, a meme call only above 20%, and a leverage call is marked at the first take-profit target. Breakeven trades and small gains are counted as losses. The record is logged win or lose, with screenshots on almost every entry.
In 2024 the win rate was 71.9% across 246 calls. In 2025 it was 72.7% across 765 calls. For 2026 year to date it sits at 69.9% across 249 calls. The independently graded April 2024 to November 2025 sample lands at 70.18% across 929 calls, with 652 winners.
Every call is posted live inside the group, so it carries the date and time it went out. The entry stays in the log whether it won or lost. Members then post their own results in the wins log, and almost every one of those comes with a screenshot. 2,500+ wins have been posted to date, and nearly all of them have a screenshot attached.
No. Small gains and breakeven trades are counted as losses, not wins. A gem call has to clear more than 10% to count, and a meme call has to clear more than 20%. The bar is set high on purpose, so the win rate you read is harder on itself than most published numbers.
Yes. The main group is free to join at $0, and that is where the calls post live in real time alongside 40,000+ members. There is a paid tier on top of it, but the free group is where you can watch the record build for yourself.
No. Past performance does not guarantee future returns, and crypto is high risk. A documented win rate shows how the calls have been graded over time. It is not a promise of any result on any future trade. Never invest money you cannot afford to lose.
Join 40,000+ Gem Hunters. The free group posts daily market updates and gems that clear the checklist, with the reasoning attached, and every call lands in the log. No junk, just gems.
Join 40,000+ Hunters Free