Twitterwatcher - introduction

Introduction

I have spent the past few years in crypto, trying to understand what is happening around me. I have a technical background and many tech friends who invest much time trying to understand every project entirely.

It did not work. The projects that, according to myself, were doomed to succeed, were unprofitable.

It got me thinking.

What if the crypto market is not knowledge-driven, but rather hype-based?

If you think about it - it makes sense.

1. The price variations happen for a number of reasons - main being positive press coverage, big news, or precisely the opposite - a scam with a similar name can drop the price more than 10%.
2. People tend to spend less and less time tracking the progress and technicalities of the projects and more time watching the tweets from the big guys in the company.
3. Finally, the crowd psychology is here at its finest - if I see crypto losing, I will sell.

Knowing this - I kept thinking - what do I need to know to be able to use this information before the rest of the pack notices it?

Here is the approach I would like to share with you.

What is twitterwatcher?

My thought process for being ahead of the curve:

1. I want to know what is the overall mood of the crowd
2. I want to know when the mood swings happen and in what direction they are going
3. I want to be able to observe it live and see the correlations between the price and mood

Assumptions I made:

1. I will watch twitter with pre-selected tags - as this is the most common communication medium in crypto
2. I will watch the most significant exchange and use the data from there - as this is the most common platform
3. I will not automate anything until I have traction from the community

So now - with all these in mind, I would like to introduce you to the https://twitterwatcher.com

A simple website that is tracking the mood swings in the twitter community for #TRX tag (more to come in the next few weeks).

What we do:

Tweets

We get all the tweets with tag #TRX and perform both numerical (how many) and sentimental (are they positive, negative or neutral) analysis. We then present this data in a histogram. You will get the default view on the website a six-hour graph with a five minutes step. The chart is auto-updated every 5 minutes, so you can leave it on your screen and observe the changes live! 

Legend - in the red circle - is usable as well - it allows you to hide/show tweets type you want.

The graph above has positive tweets hidden.

What you can see in this graph:

• The mood swing of the crowd - if you will notice a trend (I will give you an example from past in the bottom of this article) you can ride on the wave and get the most out of it.
• The graphs are synchronised - this means that a tooltip and a pointer are visible for all of the charts.

Exchange

We get all the data from the Binance and present them without any changes. At the default view, you will see an average price graph calculated from formula (max_price+min_price)/2.

After clicking the Candlestick graph - you will receive the standard candlestick graph with information about the open/close and min/max values in the timeframe.

The second graph presents the total number of transactions with pair TRX:USDT.

What you can observe in these graphs:

• The changes in the price in correlation with the mood swings of the crowd as read from tweets.
• The graphs are synchronised - this means that a tooltip and a pointer is visible for all of the charts.

Analytics

In this section, we will be presenting various analytical data - as this is still WIP part of our product. We want to show a relation between the tweets/mood and the price/size of transactions.

For now, we decided to present two graphs for more advanced users:

Figure 1 - tweets mood VS delta average price change

Where:

• tweets mood → number of positive tweets/(number of negative tweets + number of neutral tweets)
• delta average price change → average price (from the current step)/average step (from the previous step)

For our Graph 2 - delta tweets mood VS delta average price change

Where:

• delta tweets mood → tweets mood (from the current step)/tweets mood (from the previous step)
• delta average price change → average price (from the current step)/average step (from the previous step)

How to use it

Use Case - TRX and the price drop due to false allegations:

Below you can observe an actual event that happened on the 8th of July 2019 around 5:00 AM (GMT+2). A piece of news - as it turned out later, a fake one - was spreading through the web - and read that the authorities are raiding the Tron offices in Beijing. This message caused panic in the markets - being visible on twitter first. The number of negative tweets significantly increased in a short amount of time. No action made by the Tron officials were able to stop this wave.

When you are using our solution - look for changes in the negative tweets - this will allow you to notice the potential fall of prices in the upcoming moments.

You must know - that we cannot be held responsible for future events as many different elements can change the outcome. What we are offering to you is a tool to keep an eye on one piece of the puzzle.

Limitations and plans

Limitations

The twitterwatcher is and will be a free solution, created after hours by a few believers, for everyone to review and read the crowd mentality for the crypto world. As such we do not have the resources to monitor the most common tags - #BTC, #ETH - due to the number of tweets that are there and the enormous costs with us to perform a sentiment analysis on them.

Plans:

Below you can see some of the high-level goals we have in our plans:

• to monitor more currencies (#BNB, #ADA, #LSK, #LINK)
• to enable custom querying data and step for the graphs
• alerts when a new mood swing is starting
• to optimise our analytics graphs for more visibility
• to present the content of the tweets within the timeframe - for you to get the info about who said what
• to introduce - the messiah* and antichrists* of the twitter

Where: Messiah - a twitter account that causes the price to rise. Antichrists - a twitter account that causes the price to fall

• to start earning on the https://twitterwatcher.com

We are currently toying with the idea of introducing ads, paid users/white-labelling our solutions with more features.