OK Win NotesOK Win Win Rate Calculator: Luck or Skill?
Type your wins and losses, the payout and any fee. The tool prints the win rate, the rate that breaks even, the 95% range the true rate could sit in and the exact chance of luck alone. It opens on 4,700 wins in 10,000 rounds.
A record is a pile of wins and losses. It means something only if you know where it came from and how many rounds stand behind it. The tool above sets one against the line it has to clear. The tools page lists the other sums on this site.
At the time of writing (8 October 2026) we could not read any OK Win payout or fee, so the boxes hold an example: a win that returns 2× the stake, and a 2% fee. Replace both with what your own rules screen states.
Where did the record come from?
Start with the source, then the number.
| The record came from… | Then… |
|---|---|
| Your own tally, kept from a start you chose in advance | Type the totals. This is the only kind the tool can test fairly. |
| A chat or channel post that carries an address | Score the address in the link checker first. At the time of writing the button’s address forwarded us to a different domain each time, so the address guide shows what a moving address does and does not mean. |
| A post showing “the last 20” or “today” | Ask for every call since the first, wrong ones too. A window picked after the fact breaks the test. |
| The page’s “Winning information” list or “Today’s earnings chart” | Not a record. Both list masked names, and a list of amounts cannot say how many rounds were lost on the way. |
The opening record, line by line
The tool opens on 4,700 wins and 5,300 losses at a 2× payout with a 2% fee.
- Win rate: 4,700 ÷ 10,000 = 47.0%.
- Break-even rate: 1 ÷ (2 × 0.98) = 51.02%, which takes 5,103 wins in 10,000 rounds.
- Range: the true rate could plausibly be 46.0% to 48.0%. It never reaches 51.02%.
- Luck: if the true rate were exactly 51.02%, fewer than 0.0001% of records would sit this far away.
- Money: at 47.0%, ₹100 staked comes back ₹92.12 on average, or ₹90.21 to ₹94.04 across the range.
So on these example terms the true rate is probably below break-even. A record suggests; it does not prove.
The same rate, weighed five ways
Same example terms throughout, so break-even stays at 51.02%.
| Record | Win rate | Range | Luck alone this far off | Rounds to tell |
|---|---|---|---|---|
| 14 of 20 | 70.0% | 48.1% to 85.5% | 11.7% | 52 |
| 17 of 20 | 85.0% | 64.0% to 94.8% | 0.268% | 15 |
| 240 of 500 | 48.0% | 43.7% to 52.4% | 18.0% | 2,150 |
| 2,400 of 5,000 | 48.0% | 46.6% to 49.4% | 0.0021% | 2,150 |
| 4,700 of 10,000 | 47.0% | 46.0% to 48.0% | <0.0001% | 1,213 |
Rows three and four share a win rate and tell different stories. At 500 rounds, luck alone sits 3 points off break-even about 18 times in 100. At 5,000 rounds it almost never does. The rate did not change. The evidence grew.
What a post of 14 of 20 leaves out
Fourteen of twenty reads as 70.0%. Count only the high side: a caller whose true rate is exactly break-even posts 14 or more wins about 6.9% of the time. In a channel of 1,000 such callers, about 69 would show it.
Seventeen of twenty is rarer, 0.172% on the high side. It is also what a post looks like when it picked the best 20 calls of a long run. The tool needs the calls it hides.
How many rounds it takes
For a true rate below the 51.02% line, the tool estimates the rounds needed to tell it from break-even (an 80% chance of catching it, at the 5% level):
- 10 points below: 195 rounds
- 5 points below: 784 rounds
- 3 points below: 2,179 rounds
- 2 points below: 4,904 rounds
- 1 point below: 19,617 rounds
Before you type your own numbers
- Count every round of one game and one kind of bet. WINGO and Aviator are the two captions we could read on the recommendation row; they are different games, so each gets its own tally.
- Type the payout with your stake counted in: even money is 2.
- Type the fee the rules screen states, or 0 if you looked and saw none.
- If you keep no tally, the play ledger is a place to start one. The Wallet tab we read named a Deposit history and a Withdrawal history, which are money lists. We opened neither and cannot say whether the page lists rounds anywhere.
What this calculator cannot do
It assumes the same chance every round and a complete record. It cannot predict the next round, and it knows nothing of OK Win’s games. A seller’s screenshot of “proof” is a picked record, so the hack guide tests such claims. For the price of a round, start with the games guide, and for “how likely is that many wins?” use the probability calculator.
Weighing a record from a chat? Start with where it came from. If you open the platform’s page, read its address bar before you type.
Opens the platform’s own page in a new tab. Referral link. 18+ only.
Common questions
My own count is 47 wins in 100. Does that say anything?
Not yet. On the example terms (a 2× payout and a 2% fee), 47 in 100 is 4 points under break-even, and luck alone lands at least that far away 42.6% of the time. The range runs from 37.5% to 56.7%, which includes the 51.02% line. Telling a gap that size from luck takes about 1,213 rounds.
A screen says a win pays 1.95. What do I type for the payout?
Type 1.95 if that figure already counts your stake, and 0 for the fee if you looked and saw none. Break-even is then 51.28%, which is 1 divided by 1.95. We read no OK Win payout, so 1.95 is only an example here. If the screen means 0.95 on top of your stake, the box still wants 1.95.
Can I add my WINGO rounds and my Aviator rounds into one record?
No. They are different games, and the tool assumes one chance for every round in the record, so mixing them blurs it. Keep one tally per game and per kind of bet. We opened neither tile and read no rules for either, so each chance and payout has to come from its own rules screen.
Why does 17 wins out of 20 not show that a caller has skill?
On the example terms, a caller sitting exactly at break-even posts 17 or more wins of 20 about 1.7 times in 1,000 (0.172%). Among a few thousand callers, a handful would post it by luck alone. A post that picked its best 20 from a longer run changes the question, because the tool counts the whole record.
How many rounds should I count before I judge my own record?
It depends on the gap you want to detect. On the example terms, a true rate 5 points under break-even takes 784 rounds to tell from luck, and a gap of 1 point takes 19,617. A hundred rounds can only show a very large gap. Count from a start fixed in advance, and do not stop on a good stretch.
Does the earnings chart on the page count as a record I can test?
No. At the time of writing (8 October 2026) it listed masked names, and the top entry was the same on every load. It is the page's own display and we cannot check it. A list of earnings by masked name is not a tally of every round played, so there is no win rate to type. A ranking shows the top of a crowd, not the crowd, and we quote no amount from it.