 Show Posts
|
|
Pages: [1]
|
that is not what sentiment means!
the real sentiment is seen in the actual market and on the actual charts. everything else is just white noise like a fart in the wind. on all these platforms like bitcointalk and reddit, there are always a lot of trolls who come out of their hiding whenever there is a price drop. remember kuwakduck? these people are the ones you think are the "sentiment" and it is not the first time someone makes this mistake. that's why they keep coming back to fool you.
For sure there is a lot of bots and crap. But we ARE getting influenced by that. There are many examples that show that also "fake" information has an effect on public opinion: look at propaganda in the 1930s 1940s, look at the rise of PR (just google the relationship between Freud and Bernays, the founder of PR) or the last US election. The stories we read and hear do form our opinion – even if they are faked or even if we know they are not true (recommended reading: "thinking fast and slow" by Nobel Prize winner Kahnemann). However, I admit that taking action based on sentiment is a whole different topic. of course. i am not denying the influence that FUD has always had on this market specially since it is still a small one filled with weak hands. my example was a well known user around here who have been very active in spreading FUD. but the thing about FUD is that it loses its effectiveness the more it is used. take the China news for example in 2012-2013 when they first said China banned bitcoin and then in following years the panic was big but in 2017 when they said China banned bitcoin nobody gave fuck anymore. in 2018 they tried it again and the news didn't even spread. in 2019 you no longer even heard about it anymore. interesting perspective regarding the deterioration of the effect of FUD! I suppose this holds true for all types of newd |
|
|
|
that is not what sentiment means!
the real sentiment is seen in the actual market and on the actual charts. everything else is just white noise like a fart in the wind. on all these platforms like bitcointalk and reddit, there are always a lot of trolls who come out of their hiding whenever there is a price drop. remember kuwakduck? these people are the ones you think are the "sentiment" and it is not the first time someone makes this mistake. that's why they keep coming back to fool you.
For sure there is a lot of bots and crap. But we ARE getting influenced by that. There are many examples that show that also "fake" information has an effect on public opinion: look at propaganda in the 1930s 1940s, look at the rise of PR (just google the relationship between Freud and Bernays, the founder of PR) or the last US election. The stories we read and hear do form our opinion – even if they are faked or even if we know they are not true (recommended reading: "thinking fast and slow" by Nobel Prize winner Kahnemann). However, I admit that taking action based on sentiment is a whole different topic. |
|
|
|
All this data is just a lagging indicator. If you were around those Reddit posts back in Dec 2018 when we hit $3100 people were saying that BTC will hit 3 digits sometime in the next few months. Same with the June 2019 bull market, people were saying that it will hit a New ATH before it goes to $100K
Right now people are saying we will hit $5000. And there was a slight rally since then. So based on this data you can't really trade off it because people start complaining about BTC after it falls and people praise BTC after it rallies, so unless you got some time machine it's useless reading those posts.
That's why you need to learn to trade on your own. Because that way you can't blame other people for your losses.
You are right, in many cases it is laggin. I think we should take a statistical perspective though. How many times does it need to be right for you to make profits? For example this backtest from yesterday is based on the same data input  I assume you don't trust other people's backtest... If so, download the sentiment scripts I used, check & control the parameters yourself:https://github.com/augmento-ai/quant-reseach (scripts are in th folder called examples. you need to run them in the order of the numbers from 0 to 4) |
|
|
|
Bearish sentiment across Twitter, Reddit & Bitcointalk has not been that high since Dec 2018 + folks start calling the bottom:  |
|
|
|
Bearish sentiment across Twitter, Reddit & Bitcointalk not been that high since Dec 2018! And folks start calling the bottom: |
|
|
|
ah that's nothing. we got down to at least 5 back in august. i agree that sentiment is getting there though---lots of people around here are becoming legit fearful. there's also lots of bears cheering on lower prices. Strangley or not though, I'm not worried. For me this just seems like part of the plan. that's one anecdotal indication that we haven't seen the bottom yet.  which is also in line with the index only hitting 12 so far...... Bearish sentiment across Twitter, Reddit & Bitcointalk not been that high since Dec 2018! And people start calling the bottom again: |
|
|
|
You use tensorflow?
Cool AI 😎
Thanks! Nope – do you? |
|
|
|
Hi all, We just launched a dashboard with a Bull & Bear trading signal. So far only quants use our data. With the dashboard, our goal is to make a quantitative sentiment signal that is accessible for discretionary traders, too.  Some background on our technology: We spent the last months training an AI that identifies 93 emotions (FOMO, Bullish/Bearish, optimism, FUD, ....) and topics (hacks, scam, technical analysis,...) in social media data (Reddit/Twitter/Bitcointalk) in real-time. This is the full picture of what our AI detects in online crypto discussions:  We would be happy to hear what you think and we’re looking for testers who want to test the signal (for free) and provide feedback. You can sign up here: https://www.augmento.ai/sentiment-signal-signup/Looking forward to learning if you find this useful Cheers |
|
|
|
Hi ayou the dashboard is super simple. That's the whole idea behind what we built: making a complex (quant) method accessible for everyone. Feel free to give it a try: https://www.augmento.ai/sentiment-signal-signup/Would be interested to hear what you think. |
|
|
|
Hi all, We just launched a dashboard with a Bull & Bear trading signal. So far only quants use our data. With the dashboard, our goal is to make a quantitative sentiment signal that is accessible for discretionary traders, too. Some background on our technology: We spent the last months training an AI that identifies 93 emotions (FOMO, Bullish/Bearish, optimism, FUD, ....) and topics (hacks, scam, technical analysis,...) in social media data (Reddit/Twitter/Bitcointalk) in real-time. This is the full picture of what our AI detects in online crypto discussions: We would be happy to hear what you think and we’re looking for testers who want to test the signal (for free) and provide feedback. You can sign up here: https://www.augmento.ai/sentiment-signal-signup/Looking forward to learning if you find this useful Cheers |
|
|
|
Hi all, We just launched a dashboard with a Bull & Bear trading signal.  So far only quants use our data. With the dashboard, our goal is to make a quantitative sentiment signal that is accessible for discretionary traders, too. Some background on our technology: We spent the last months training an AI that identifies 93 emotions (FOMO, Bullish/Bearish, optimism, FUD, ....) and topics (hacks, scam, technical analysis,...) in social media data (Reddit/Twitter/Bitcointalk) in real-time. This is the full picture of what our AI detects in online crypto discussions:  We would be happy to hear what you think and we’re looking for testers who want to test the signal (for free) and provide feedback. You can sign up here: https://www.augmento.ai/sentiment-signal-signup/Looking forward to learning if you find this useful Cheers |
|
|
|
|
