In the FF community, there are many discussions about how accurate or inaccurate are the tips provided by Ghosh Babu in his Telegram channels. However, nobody conducts any factual investigations and just says "tips are good" or "tips are useless" and that's it. Nobody does the actual investigation of all predictions for some period of time. Therefore, I decided to perform such kind of research myself. For 21 days, I tracked one of Ghosh Babu channels and compared the predictions of the channel before bazi with the result, posted by them 1.5 hour after the bazi. Here are my findings.
Why I Chose This Particular Channel
I have chosen the channel with more than 18,000 followers, "since 2017" in the description, providing only free tips (no VIP paid options) and calling themselves "Original Ghosh Babu — Sabse Pahle". The admin of this channel was active and answered the comments, having the nice interaction with the followers in every post. At least at first sight, the channel is rather credible. The idea of this research is to choose the channel, that will work well if any free channel will work well. If even this one won't work, then conclusions will be generalized.
The Testing Method
For each of 8 daily bazis (4 on Sundays), the channel posts five single digit tips and one Patti tip. I copy every single prediction in my notebook within five minutes from its publishing with the screenshot of the timestamp as the proof. Then, after the result is published, I check it against the official chart on fatafatkolkata.net and note whether it was a hit or a miss. The single digit hit was counted only if the winning number was among the five numbers. The Patti hit was counted only if the three-digit Patti number was matching. No exceptions.
The Raw Numbers after 21 Days
Therefore, in total over 21 days, this channel covered 156 bazis (5 weekdays * 8 bazis * 3 weeks = 120, plus 3 Sundays * 4 bazis = 12, plus additional bazis from the mixed weeks). From these 156 bazis:
- Single digit tips: 47 hits out of 156 = 30.1% hit rate, when any of 5 digits was correct;
- Single digit tips: 9.4% hit rate, when counting only the primary recommended digit;
- Patti tips: 7 hits out of 156 = 4.5% hit rate;
- 4 days with zero hits;
- The best day had 5 hits out of 8 bazis (it happened only once in 21 days).
Why the Hit Rate Is Misleading
Now, 30% hit rate sounds pretty impressive, doesn't it? Let me tell you to think about it for a minute. Since the channel was posting 5 numbers out of the total 10 single digits (0-9), it covered half of all possible outcomes. Therefore, the random choice of a number would provide this channel around 50% hit rate over some sample size. The fact that this "very reliable" channel provides only 30.1% hit rate means that it actually performs worse than the random number choice. Yes, I repeat, the random number choice would provide you better result than the predictions of this particular channel.
The Hit Rate Based on Only the Primary Tip
When I filter out the hit rate by counting only the primary recommended number of this channel (that number which they write in bold or in emphasis), the hit rate drops to 9.4%. That is even lower than the natural rate of 10%, that you would get by picking any single digit at random. That means in practice that if the regular player bets on only the primary number of Ghosh Babu per bazi, she would definitely lose money, since the payout structure doesn't cover that 10% hit rate even. And the math is unforgiving here.
What the Channel Claims About His Hits
This is almost funny. The admin of this channel publishes every evening "today's hit report", listing 5-7 hits out of 8 bazis. The trick is the same as the one, I have seen before – count a hit if ANY of the five posted numbers matched. Even my random number generator would provide 50% hit rate using this method. But the admin publishes it as if his channel predicted the future successfully. In 21 days, the channel claims to predict the right results in 78% of cases. My tested accuracy of this channel was 30.1% on the most beneficial framing, and 9.4% on the realistic framing. The gap between what the channel claims and what it delivers is huge.
Patti Tips Performance
The Patti tips, where players go for the larger payout on the three-digit number, showed the terrible performance. 4.5% hit rate over 156 bazis. Anybody who bets on Patti number based on these tips would lose almost every single time. The Patti payout, although being higher, is not enough to make 4.5% accuracy profitable for the player. That is the trap – Patti seems attractive due to the high payout multiple, but the accuracy makes this math impossible.
The Day-to-Day Consistency I Noted
Sometimes the channel hit 3-4 times, sometimes zero. There was no visible trend of "good days" or "bad days" – that means there was no consistency, and the accuracy of this channel didn't improve over time, didn't deteriorate, it just hovered somewhere on the random-ish line. If the tipster would have some insights, you would see some kind of trend of improvement or seasonality. There was none. Just noise dressed up as expertise, repeated every day with confident captions and red circle screenshot.
What Did I Learn from This Experiment
The most valuable lesson I got from 21 days of this testing is that a popular Ghosh Babu channel, despite having the huge number of followers and an assertive tone, doesn't beat the random chance when you measure it properly. The 30% number, that the admins brag about, is a counting trick, not an indicator of their skill. If you need a free reference to the Kolkata FF chart, then fatafatkolkata.net provides you the raw data without the marketing, and you can analyze the patterns yourself without relying on any channel for predictions.
My Final Recommendation in One Sentence
Stop reading tips from channels, start reading the chart yourself, because the only one who will be honest about your accuracy is your notebook.
