Search content
Sort by

Showing 3 of 3 results by FKamil5711
Post
Topic
Board Bitcoin Discussion
Re: Bitcoin puzzle transaction ~32 BTC prize to who solves it
by
FKamil5711
on 06/01/2025, 11:59:19 UTC
It is so funny how some people think that because two points look alike then its a hot zone worth to investigate, while some others think the complete opposite. And they're both right: you can indeed argue that you can get closer to a solution little by little (see genetic algorithms), and you can also argue that statistically the solution should be as far away as possible from any other similar solution.

It's a very good example of why they are both actually wrong, and the winner here is simple, his name is Random and it seems no one wants to accept its definition or his highly entropic personality. Random is very hated and bullied, and no one believes him or takes him for granted, because he has a really big problem: every time you look at him, there are high chances he may look completely different. Or maybe not. Or maybe just a little? But random has a girlfriend: Prime. Prime understands him and the two share a lot of common ground, but they have a secret relationship.

The curve equation and parameters are completely and intentionally devised and chosen in such a way to ruin all of these assumptions, hopes, and "I know better"s. No, there is no magic pattern, formula, neural network, simplification, reduction, statistical analysis anomalies, or other similar BS. There is no pattern in either decimal, hex, binary, ASCII art, Photoshop horoscope, or any other visual mapping from bits to representation that you can ever think of.

Also if one does not have a clue about what hex numbers mean, they should really not waste time on discovering fundamental knowledge (for themselves only) while trying to break something called Elliptic Curve Cryptography. Now, this doesn't mean they are stupid, it just means that someone may be able to understand black holes equations, but it doesn't mean they should ever touch programming.

It’s clear you have a strong understanding of elliptic curve cryptography, and your points about randomness and the intentional design of ECC are well taken. I completely agree that the parameters of the curve were chosen to defy patterns, and my attempt to find hot zones or meaningful clustering was based on flawed assumptions.

I think there’s still value in experimenting, I am a layman in cryptography after all, even when the outcome is almost certainly failure. I went into this knowing that without GPU acceleration or distributed systems, my chances of finding the key were practically zero. But for me, this was less about cracking ECC and more about learning and  challenging myself.

Even though the process didn’t lead to a solution, it gave me a deeper appreciation for the strength of ECC and how robust these systems are. I recognize now that randomness really is the "winner" here, and any clustering I thought I saw was more a product of my own misinterpretation than a meaningful signal.

I share this not to "reinvent the wheel" or to claim expertise, but because I found the process rewarding and hope others might learn from my mistakes or even just my curiosity. While I’ll be stepping away from this puzzle, I’ve gained a lot from the experience, and for me, that’s what matters most.

Post
Topic
Board Bitcoin Discussion
Re: Bitcoin puzzle transaction ~32 BTC prize to who solves it
by
FKamil5711
on 06/01/2025, 07:50:49 UTC
By analyzing previously found private keys for earlier wallets, I noticed a pattern in the number of leading zeros. Based on this, I estimated that the private key for Wallet #145 might have 30–32 leading zeros.

A near miss occurs when part of the public key derived from a guessed private key matches part of the target public key. The more bytes that match, the closer the guessed private key is likely to be to the target key on the elliptic curve.

Unfortunately both of your assumptions you need to revise.
1. If looking for Puzzle 145 then it is certain to be contained in the bit range given by pow(2, 144) to pow(2, 145) - 1. Which means in the the Hex range of 0000000000000000000000000001000000000000000000000000000000000000   to  0000000000000000000000000001ffffffffffffffffffffffffffffffffffff

2. On Elliptic curve math, the distance is not measured like ordinary math. You can't say by looking even the public Key with the actual distance of 1. See there is no matching pattern between the bytes of these two.

Private_Key     Public_Key
1                    0279be667ef9dcbbac55a06295ce870b07029bfcdb2dce28d959f2815b16f81798
2                    02c6047f9441ed7d6d3045406e95c07cd85c778e4b8cef3ca7abac09b95c709ee5



Ah,at least now I can see why my efforts were doomed from the start, it is now glaringly obvious.
Post
Topic
Board Bitcoin Discussion
Re: Bitcoin puzzle transaction ~32 BTC prize to who solves it
by
FKamil5711
on 06/01/2025, 06:08:57 UTC
My Attempt at Finding the Private Key for Wallet #145

Hello everyone,

This is my attempt at finding the private key for Wallet #145. Please note that I’m not a programmer or mathematician, I’m just someone who became fascinated by cryptography and stumbled across this puzzle. I might be completely wrong here. I'm happy to learn and would love to hear where I went astray if that’s the case.


Background

This script ran on a Hetzner EX44 server and my old HP G4 Chromebook running Arch Linux. From the start, I knew brute-forcing or even attempting BSGS without GPU acceleration was an almost impossible task. However, it was all I could afford, and I didn't want to invest too much in building a dedicated rig.

Eventually, I had to stop after my Chromebook gave out.I learned a lot during the process, and I wanted to share my findings with you.


Leading Zeros

By analyzing previously found private keys for earlier wallets, I noticed a pattern in the number of leading zeros. Based on this, I estimated that the private key for Wallet #145 might have 30–32 leading zeros. Here’s how I came to that conclusion:

Known Keys

Code:
Key #110: 00000000000000000000000000000000000035c0... (38 leading zeros)
Key #115: 00000000000000000000000000000000000060f4... (38 leading zeros)
Key #120: 00000000000000000000000000000000000b10f... (37 leading zeros)
Key #125: 0000000000000000000000000000000001c533... (36 leading zeros)

Pattern Analysis

There is a consistent loss of approximately 1 leading zero every 5 positions.

Between Key #125 and Key #145, there are 20 positions, meaning about 4 leading zeros are likely lost in this interval.

Starting from 36 leading zeros at Key #125, I subtracted 4, leaving me with an estimate of 32 leading zeros for Key #145.


Near Misses and Hot Zones

What is a Near Miss?

A near miss occurs when part of the public key derived from a guessed private key matches part of the target public key. The more bytes that match, the closer the guessed private key is likely to be to the target key on the elliptic curve.

Here’s an example:

Code:
Target Public Key:
03afdda497369e219a2c1c369954a930e4d3740968e5e4352475bcffce3140dae5

Guessed Public Key:
03afdda497369e219a2c1c369954a930e4d3a4936821b5458756bf12afc3412c87

Matching Bytes:
03afdda497369e219a2c1c36 (5 bytes matched)


What Are Hot Zones?

Hot zones are specific subranges within the private key space that are more likely to contain the target private key. They are identified by analyzing clusters of near misses. If multiple near misses occur in a particular range, it becomes a hot zone worth focusing on.

Example Hot Zones

Based on my analysis, here are some identified hot zones:

Code:
Range 0x4: (0x4c000000, 0x4dffffff)

Range 0x5:

Primary: (0x52000000, 0x53ffffff)

Secondary: (0x59000000, 0x5affffff)


Range 0x7: (0x75000000, 0x76ffffff)

For instance, if near misses cluster in the range 0x52000000–0x53ffffff, this subrange becomes a primary hot zone within the larger 0x5 range.


How Does This Process Work?

1. Run a Search:
The script uses BSGS to search the keyspace and logs all near misses, including the guessed private key and match details.


2. Analyze Near Misses:
Clustering is analyzed to define hot zones, which are subranges where matches are more likely.


3. Refine the Search:
Resources are dynamically reallocated to focus on promising hot zones, while continuing some exploration in less-likely areas to avoid missing the target key.




Challenges and Limitations

1. Bias:
If early near misses disproportionately influence the weights or hot zones, the search might ignore other valid ranges, leading to a self-reinforcing bias.


2. Sparse Data:
With too few near misses, it’s difficult to establish meaningful clustering or confidently define hot zones.


3. Random Noise:
Some near misses occur randomly, creating false hotspots if not carefully analyzed.


Final Thoughts

While I ultimately didn’t succeed in finding the private key for Wallet #145, I learned a lot about cryptography, elliptic curve mechanics, and computational optimization. I’d love to hear your thoughts or suggestions on my approach. If you see any glaring mistakes or areas for improvement, please don’t hesitate to point them out—I’m here to learn.

Thank you for reading, and I hope this helps others interested in cryptography or keyspace searching!


Feel free to share your thoughts, corrections, or suggestions. If you have insights on hot zones, leading zeros, or better search strategies, I’d love to hear them. I also read RetiredCoder's Rckangaroo script, which is much better than my code. I haven't fully read the thread yet but it seems like his solution is the best for the time being. Here’s the file for Hexplorer, my attempt at finding Key #145. https://drive.google.com/drive/folders/1P-94tGIjwhbvItMPlLqZPzsHSLfbLMYi

> Note: This is an older save, so if you try to run it, double-check it first. I really don’t want to give myself a brain aneurysm trying to code on my phone, typing this fried my brain. Also to the mods let me know if I'm breaking any rules.