[Grimeta]

New Predictions

2021-07-13 and 2021-07-14 DOWN

2021-07-14 and 2021-07-15 UP

So 14 is up or down? 😆

[SIM-3867235]

So 14 is up or down? 😆

Its between 13 and 14 as well as 14 to 15. Predictions are for 24 hour periods.

[m8than]

What's backtesting results like with this?
Also what inputs are you using for your model?

[Don Vito Corleone]

I like that you're sharing your way to training your model with us so we can see how your model improves. You state that it's not financial advice and that you just want to see whether the predictions are accurate or not, there's nothing wrong with that. I'm glad we can follow you through this process and I'm sure that if you had posted this in the Python sub forum people would've replied differently.

This sub forum can be a bit harsh. A lot of it has to do with the history of low quality posts from when people come here during crypto hype. Looking at the replies on this thread it looks like people think you're saying saying that you have a crystal ball, but that's not the case here.

[Grimeta]

Your predictions are right on point so far 💯 please post more 🙂

[SIM-3867235]

What's backtesting results like with this?
Also what inputs are you using for your model?

The model reaches a validation loss of around 3-4% and same goes for training loss. The inputs of the model is something that I cannot disclose.

[Grim]

The model reaches a validation loss of around 3-4% and same goes for training loss. The inputs of the model is something that I cannot disclose.

Still helpful either way thanks for sharing it

[SIM-3867235]

I like that you're sharing your way to training your model with us so we can see how your model improves. You state that it's not financial advice and that you just want to see whether the predictions are accurate or not, there's nothing wrong with that. I'm glad we can follow you through this process and I'm sure that if you had posted this in the Python sub forum people would've replied differently.

This sub forum can be a bit harsh. A lot of it has to do with the history of low quality posts from when people come here during crypto hype. Looking at the replies on this thread it looks like people think you're saying saying that you have a crystal ball, but that's not the case here.

Yeah I understand. The predictions that the model makes can never be taken as financial advice since there is always a slight chance that the prediction fails. The model, at this stage cannot predict a sudden dump for example. I think the Python subforum would not be fit for this thread since I am not that open to sharing the source code for this model. It's just a thread where you can follow along on the progress.

---

Your predictions are right on point so far 💯 please post more 🙂

Thanks man! Yes, I am preparing new predictions for the next 2 days. I am very happy with the results so far. I am currently improving the model so that I will be able to disclose price predictions.

---

Results


Originally the predictions were:

2021-07-13 to 2021-07-14 DOWN

2021-07-14 to 2021-07-15 UP

Actual price movements were:

2021-07-13 to 2021-07-14 DOWN 5.3%

2021-07-14 to now UP 3.88%

This makes the second round of predictions correct. I might finish the updated version to give a bigger outlook as well as on a longer timespan. If not, you can expect new predictions for the coming two days in a couple of hours.

[Don Vito Corleone]

Yeah I understand. The predictions that the model makes can never be taken as financial advice since there is always a slight chance that the prediction fails. The model, at this stage cannot predict a sudden dump for example. I think the Python subforum would not be fit for this thread since I am not that open to sharing the source code for this model. It's just a thread where you can follow along on the progress.

Yeah exactly, I meant it in a positive way ✌ I'll keep checking out the progress you make, all the best my man.

[SIM-3867235]

New Predictions

2021-07-15 to 2021-07-16 DOWN

2021-07-16 to 2021-07-17 UP (almost neutral movement)

---

Yeah exactly, I meant it in a positive way ✌ I'll keep checking out the progress you make, all the best my man.

Thank you. I appreciate it. 🤩🤩🤩

---

Should I make a website for all the predictions?

[Joca300]

Great post, had the same prospect for the markets, so yeah, money was made, short ratio has also declined, lets make some satoshis guys

[Neal Caffrey]

Well done!  To compare this with coincidence, the chance of guessing this outcome correctly at random is 12.5%.

I'm reading this as sarcastic but it looks like OP read it as congratulatory, please confirm ✌

I like that you're sharing your way to training your model with us so we can see how your model improves. You state that it's not financial advice and that you just want to see whether the predictions are accurate or not, there's nothing wrong with that. I'm glad we can follow you through this process and I'm sure that if you had posted this in the Python sub forum people would've replied differently.

He hasn't shared anything - I'm guessing he's just using some kind of Keras LTSM model or something that doesn't consider the longer time-frame momentum trades that have been the real moneymakers. There simply isn't enough testing data that reflects other changes and if you modeled waves off historical waves, there's an underlying assumption that there are an infinite number of people that can buy BTC, which obviously isn't the case.

[SIM-3867235]

I'm reading this as sarcastic but it looks like OP read it as congratulatory, please confirm ✌


He hasn't shared anything - I'm guessing he's just using some kind of Keras LTSM model or something that doesn't consider the longer time-frame momentum trades that have been the real moneymakers. There simply isn't enough testing data that reflects other changes and if you modeled waves off historical waves, there's an underlying assumption that there are an infinite number of people that can buy BTC, which obviously isn't the case.

"longer time-frame momentum trades that have been the real moneymakers" Please read the entire thread before "guessing". The motive of making a model that is able to predict market direction on a day-to-day basis is for traders that engage in futures/options and leverage trading.

---

Great post, had the same prospect for the markets, so yeah, money was made, short ratio has also declined, lets make some satoshis guys

Thanks for your kind remarks. I will be trying out futures trading on these predictions for the next round of predictions.

[luv.]

Mercury is in cancer so I can only guess that BTC will hit $27k within a week

[Sixen+]

Market is so boring that people is making threads of price predictions for a range of 2k,
Nevermind now you edited the thread and it only shows the direction it will go for the next 3 days without price prediction? Ok

This thread is irrelevant honestly but what makes you think a 2k range on an asset is boring? That’s sub 10 percent move in a day. You know how profitable this is? This is lovely. Just came here to say this that’s all