Quote from: Cauchy on September 14, 2017, 01:20:30 PM

Quote from: KryptoKings on September 14, 2017, 01:00:45 PM

Quote from: Cauchy on September 13, 2017, 06:19:05 PM

Quote from: KryptoKings on September 13, 2017, 04:39:31 PM

Quote from: heyjoe on September 13, 2017, 01:02:45 PM

Love the idea of being able to be a part of a trading bot without shelling out the high upfront costs or needing to spend hours figuring out how to tweak and set it up.

The token holder payout of 85% of gas seems very fair as well - as someone else mentioned, a calculator would be a great tool to allow potential investors to play around with how the numbers may work out.

Currently going through the whitepaper in more detail to gain a stronger understanding but looking forward to following along.

Hello heyjoe,

I put the answer done to levyashin:

Quote from: KryptoKings on September 13, 2017, 03:50:42 PM

Quote from: levyashin on September 13, 2017, 10:00:43 AM

Some projects make a little calculator tool, will you do one?

It's like:
User will write amount of npx you hold or thinking to hold or just random number just see the result
He will choose a revenue on menu,
and calculator gives the montly profit will token holder will get.
It gets very helpful when investing. You know what to expect.

Hello levyashin,

We will think about that. However, we have to check if it is possible to do so from a legal point of view. But the math is quite easy to do. You can get an idea of the performances of our trading bots from the blackpaper : http://blackpaper.napoleonx.ai
From the performance, you derived the gas which equals 25% of the performance and then NPX tokens holders gets 85% of that. You will quickly see that it is a business volume. The more is raised in DAFs, the more to the NPX token holders at the end!

The performance metrics is from backtesting the strategies on historical data?

The advantage of developing algorithm is that you can backtest them. So what we show in our Blackpaper is indeed a backtest. However, we should stress that one of the developed model has been used for several underlyings to demonstrate its robustness. We have also a very low trading frequency to avoid having large tracking error between our backtest and the live version. At last, Stephane Ifrah, one of the founder, has been using algos for more than 9 years for his own use. We have also put some of our trading strategy on an index publisher that we have developed using a test ethereum blockchain. We will release it very soon.

Thanks for the feedback. Indeed backtesting is beneficial but to render the metrics you have the algo must have been tweeked to that specific dataset imo. Looking forward to have a look on the index publ

You are raising an important point here on overfitting. We are very acute of any such risk and the strategies presented in our Blackpaper are for a large part the same strategy applied to different and unrelated markets. This is one of our selection criteria when we select an algo. We want it to be applicable to different markets so that we believe that we are closer to the modelisation of human behavior than overfitting.