In last 21 days, I carried out the experiment which could be considered the most boring one for FF community members – I took one particular Ghosh Babu Telegram channel, ignored all other sources of information and compared each number predicted in that channel with the results on fatafatkolkata.net. No VIP plans, no analysis, no calculations. One channel, one notebook, 21 days, 168 bazis on working days plus 12 bazis on Sundays. That particular channel had 8,400 subscribers and its admin assured that it had "80 to 85 percent accuracy daily". I wanted to check if these claims really work and the answer was predictable – "No".
The Ghosh Babu Channel I Chose and My Reasons
I chose a moderate channel in terms of popularity – neither the biggest nor the smallest. The biggest ones are always manipulating the statistics with the edited screenshots, while the smallest ones make no statistical claims at all due to lack of data. That particular channel was active, the admin gave tips on numbers before bazis (not after as usually), it operated for at least 11 months. I used neither screenshots nor anything else, just subscribed to the channel and quietly waited for notifications without making any activity in the chat during the whole period of the experiment.
The Structure of the Tips I Was Using
For every bazi, the admin gave five single digit numbers and one Patti number. This is a common structure for almost any Ghosh Babu channels whether they are paid or free. The tips for morning 1st Bazi were posted at about 10:05 AM, 15 minutes before the announcement of the result – 10:20 AM. Tips for other bazis were posted approximately 30 to 40 minutes before the draw. In total, during 21 days, I ended up with 900 tips on single digits and 180 Patti numbers. This is quite a considerable volume of data.
Raw Statistics of the Experiment Without Any Sugar-Coating
- Total number of bazis observed: 180.
- Total number of tips on single digits posted: 900.
- Total number of tips on single digits hit: 168.
- Accuracy of tips on singles: 18.6%.
- Total number of tips on Patti posted: 180.
- Total number of Patti tips hit: 9.
- Accuracy of Patti tips: 5%.
When I analyzed the "scorecard" posted by the admin during the same period of time, he stated that his accuracy was 79 to 84 percent daily. Now I know how he came to that figure – he regarded a "hit" whenever one of the 5 numbers matched the result, not when the player was winning. During 180 bazis, at least one of the 5 numbers matched in 137 bazis – it is 76 percent of bazis, and it cannot be regarded as a lie. However, for the player betting on one number, the hit rate would be about 19%, which was recorded in my notebook.
The Patterns I Noticed During 21 Days
Of course, some days were better than the others, and I started noticing certain patterns. The 1st Bazi (10:20 AM) and the 2nd Bazi (about 11:50 AM) were the most accurate bazis with the probability of about 21 to 23 percent. The afternoon bazis were less accurate – from 15 to 17 percent. The 7th and 8th Bazi at night were the least accurate ones – from 13 percent. It corresponds with the common knowledge of FF players gained throughout many years of experience – the morning rounds have cleaner patterns. The accuracy of Patti tips was low and quite constant all day long – from 4 to 6 percent, it is just guessing.
Sunday Effect" Exists
There are only 4 bazis on Sundays instead of 8, so I thought that small sample would produce unreliable data. However, I noticed that Sunday accuracy was significantly higher – 24.7% on singles during 12 Sunday bazis. It is still a small sample, so I am not able to make reliable statements, but maybe it is true that Sundays have less rounds, so the admin has more time to analyze them and give tips. Or this could be just an illusion caused by variance. Nevertheless, this is the only interesting discovery of the experiment which requires much more time to be confirmed.
How the Channel Manipulated Reality With Screenshots
Every evening around 9:30 PM, the admin used to post the screenshot called "Today's wins summary". I saved all 21 screenshots. Looking at them now, I noticed that every screenshot had 6 or 7 hits out of 8 bazis. There was not a single screenshot with a bad day. It is statistically impossible – even according to his definition of "hit", my notebook recorded several days with only 3 or 4 hits. He either selected only the good days or edited screenshots in advance. I am not able to prove which way he uses, but the pattern was obvious for 3 weeks of my observation.
What Does It Mean For a Regular Player?
If you are a regular Kolkata FF player who uses tips from any Ghosh Babu channel, here is the bitter truth. Having the hit rate on single digits equal to 18 to 19 percent and taking into account the payout structure of the game, you are losing money over any reasonable sample size. Even the advantage over pure guessing of 8 to 9 percent is not enough to compensate the structural house advantage of the game. Tips on Patti numbers at 5 percent accuracy are pure luck and you should totally disregard them. If you feel like playing, the data suggest that morning bazis are somewhat easier, but it means nothing.
My Conclusion After 21 Days
The channel was not a fraud in the literal sense of the word – they gave tips in time, tips hit a bit better than the level of pure luck, and the admin communicated with the players. However, the gap between the claimed accuracy and the win rate of a regular player is so huge that the channel manipulated the perception of the audience. This is true for almost any Ghosh Babu channel I ever saw. The honest interpretation of my data is that the tips are just guidance at best, and using them as predictions to bet real money is a sure-fire way to lose. Bookmarks fatafatkolkata.net for the public chart, but never replace your own tracking notebook with any channel's claimed accuracy.
