Topic: [ANN][PRESALE]⭐ REPU ⭐ - Smart reputation managment - Pre-ICO Soon🚀🚀 - page 56. (Read 19313 times)
February 18, 2018, 04:42:23 PM
How many people do you have on the whitelist? I think this project deserves a high number which should not be that hard because there is a lot of buzz with ICOs at the moment.
February 18, 2018, 04:19:39 PM
It seems this project REPU will grow in future become very popular for it's nice platform it is nice to follow this project for more updates.
That is a valid point since we are now getting accustomed to rating platforms, I guess adoption for this project will go in a faster rate.
February 18, 2018, 04:14:29 PM
I don't think averaging extreme ratings is or should be the solution.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
I am sure that any ratings can be artificially adjusted. The question is who will pay for it and how much. Conspiracy can never be ruled out.
And it is a trivial task to use bots. Some people might say that it is simple for people/AI/machine-learning to detect bots, but if people are competent enough it is easy to make multiple accounts that can't be differentiated from real accounts.
February 18, 2018, 04:04:43 PM
Yes i also think that use repu for ico rating site could be a good application of the project. There is so much application for smar reputation management. It’s a big need especially in crypto world where reputation is a really important thing.
There are some small project what are tackling reputation scores, but not on a big scale or as an important feature, like lifID, so an exclusive project handling this sector will be better, especially for HR companies and such, if this is down well, and Repu get key partners, this will help a lot to the mainstream of cryptos in general, ICOs will just be a small part of the platform, though very need for the start.
February 18, 2018, 03:16:26 PM
Yes i also think that use repu for ico rating site could be a good application of the project. There is so much application for smar reputation management. It’s a big need especially in crypto world where reputation is a really important thing.
February 18, 2018, 03:16:22 PM
It seems this project REPU will grow in future become very popular for it's nice platform it is nice to follow this project for more updates.
February 18, 2018, 03:01:40 PM
I don't think averaging extreme ratings is or should be the solution.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
I am sure that any ratings can be artificially adjusted. The question is who will pay for it and how much. Conspiracy can never be ruled out.
And it is a trivial task to use bots. Some people might say that it is simple for people/AI/machine-learning to detect bots, but if people are competent enough it is easy to make multiple accounts that can't be differentiated from real accounts.
February 18, 2018, 02:45:43 PM
I don't think averaging extreme ratings is or should be the solution.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
I am sure that any ratings can be artificially adjusted. The question is who will pay for it and how much. Conspiracy can never be ruled out.
And it is a trivial task to use bots. Some people might say that it is simple for people/AI/machine-learning to detect bots, but if people are competent enough it is easy to make multiple accounts that can't be differentiated from real accounts.
February 18, 2018, 02:37:55 PM
I don't think averaging extreme ratings is or should be the solution.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
I am sure that any ratings can be artificially adjusted. The question is who will pay for it and how much. Conspiracy can never be ruled out.
February 18, 2018, 02:32:11 PM
The Repu token was invented to give people a reason to give ratings. People usually vote on the extreme ends. It's either a perfect 5 stars or it totally blows. Is the Repu token reward going to be worth enough that people also bother with meh ratings? And even then how are you going to compensate for the extreme ratings?
Extreme ratings are part of the process. It is taking the average of all votes. So even every vote is extreme it will make an average point around somewhere logical.
I agree, we all have different perception. What is ok for one person, can be extreme for others. Besides, probably some extreme ratings won't go through because of AI/ML.
So you are saying that AI or some sort of algorithm can intervene to rating process? Maybe i understood wrong but if so, i think it is wrong that somebody or machine intervening the reputation process in decision phase( but i'm excluding that intentional fake crowd reputations, 1000 people gives 1 to a movie or else)
February 18, 2018, 02:14:25 PM
The Repu token was invented to give people a reason to give ratings. People usually vote on the extreme ends. It's either a perfect 5 stars or it totally blows. Is the Repu token reward going to be worth enough that people also bother with meh ratings? And even then how are you going to compensate for the extreme ratings?
Extreme ratings are part of the process. It is taking the average of all votes. So even every vote is extreme it will make an average point around somewhere logical.
There should also be some more complex logic in play but, in general the system will need lot of data to calibrate
It's really interesting how complex the topic is. Fake ratings, extreme ratings etc. I'm beginning to understand why there is no demo yet.
That's how AI and machine learning work, the more data they gather the more accurate is the output. You can think of a human being, he/she'll predict some situations only with his/her past experience
February 18, 2018, 02:10:57 PM
I don't think averaging extreme ratings is or should be the solution.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
Your two described problems are correct. But at the end the people know this. If I buy something on Amazon I never read the 5 start or one star feedback.
I read the longer ones and these which were liked much.
I agree that the are problems. There was ip web cam on Amazon and you get a management software for free if you leave a feedback.
February 18, 2018, 12:54:48 PM
I mean...chinese new year is over, so everything should go back to normal prett soon
Chinese people need some time do get new money, we cant expect, that on 15.2. prises raise like hell.
The rising we see just now is the expectation of the non chinese people, not chinese people are buying again.
They come later in march.
At least price is now over 10k and market looks better than it was.
If it keeps like that, it will also make a positive effect for icos.
February 18, 2018, 12:53:47 PM
I don't think averaging extreme ratings is or should be the solution.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
There are 2 conflicting problems at the same time.
The first is that ratings are exaggerations. Ratings attract people who have very strong feelings about the thing they are rating. Especially if they are angry. Let's say 20 hotel customers had a bad experience. 15 of them rate the hotel 0 stars, while from the other 980 customers who thought it was OK only 5 bother rating 3 to 5 stars.
The other problem is the opposite. If you have thousands of ratings the most extreme ratings disappear in the statistics. But those extreme ratings contain the most important information. Let's say most bank customers don't really read the small print of a loan they got, they're happy enough they got it at all and rate the bank 4/5. Only 3 customers notice misleading terms in the documents and rate the bank 0 stars, but are completely averaged away by the happy customers ratings.
February 18, 2018, 12:39:30 PM
We should use REPU to give point to ICO rating sites, this way people know to trust the rating or not.
would certainly be a sector where repu should enter, in addition to the ratings of industry experts on the ico comparison sites, it would be interesting to read the opinions, ratings of common investors/users, i think this will be the year of the utilities token, a video on the operation of the platform would be welcome... February 18, 2018, 12:12:30 PM
I mean...chinese new year is over, so everything should go back to normal prett soon
Chinese people need some time do get new money, we cant expect, that on 15.2. prises raise like hell.
The rising we see just now is the expectation of the non chinese people, not chinese people are buying again.
They come later in march.
At least price is now over 10k and market looks better than it was.
If it keeps like that, it will also make a positive effect for icos.
February 18, 2018, 11:35:29 AM
The Repu token was invented to give people a reason to give ratings. People usually vote on the extreme ends. It's either a perfect 5 stars or it totally blows. Is the Repu token reward going to be worth enough that people also bother with meh ratings? And even then how are you going to compensate for the extreme ratings?
Extreme ratings are part of the process. It is taking the average of all votes. So even every vote is extreme it will make an average point around somewhere logical.
There should also be some more complex logic in play but, in general the system will need lot of data to calibrate
It's really interesting how complex the topic is. Fake ratings, extreme ratings etc. I'm beginning to understand why there is no demo yet.
That's how AI and machine learning work, the more data they gather the more accurate is the output. You can think of a human being, he/she'll predict some situations only with his/her past experience
February 18, 2018, 10:57:53 AM
I mean...chinese new year is over, so everything should go back to normal prett soon
Chinese people need some time do get new money, we cant expect, that on 15.2. prises raise like hell.
The rising we see just now is the expectation of the non chinese people, not chinese people are buying again.
They come later in march.
At least price is now over 10k and market looks better than it was.
If it keeps like that, it will also make a positive effect for icos.
February 18, 2018, 10:56:53 AM
We should use REPU to give point to ICO rating sites, this way people know to trust the rating or not.
February 18, 2018, 10:19:11 AM
The current crypto market ‘recovery’ is disappointing me at the moment (my alts are in the red), but I think Repu should still get the token sales over as soon as possible without any further delays because investors are keen to see progress in the project.
The main growth nowadays may be a good sign of more values applied in projects like Repu, soon, unfortunately alts bleed for the growth of Btc, Eth, but I do not think it is an improper moment for Icos, I see more and more users if joining and spreading the wordI think that in the future more and more projects will be more independent from ETH's or BTC's price movements. These days it's still this feeling that either all succeed or fall together, but once these concepts demonstrate their real world use cases they can exist on a standalone basis.
Actually, utility tokens are now beginning to take center stage in crypto verse and are expected to boom in the near future.
it's gonna take a while man. utility tokens are still tied to ETH blockchain, so it they take off, then ETH will take off.
Jump to: