Languages : English, Français

Extensions of the Syracuse sequence (3n+1, 3n+b, 5n+1,...) : Proofs of the different conjectures
Treat yourself to (or simply offer) revolutionary numbers
It's magic


  1. Proofs of the different conjectures
  2. Revolutionary numbers

    Proofs of the different conjectures for the extensions of the Syracuse sequence (3n+1, 3n+b, 5n+1,...)

    • 2024/10/17, Version 3.1 of the proofs of the theorems for the extensions of the Syracuse sequence (3n+1, 3n+b, 5n+1,...) is available :

      • Download :
        • PDF in French : JB_syr_Th_FR_V3_1.pdf (2.2 MB)
          SHA256 : 2809A0F11AE1534FA020209885232422EEB1706242C3FD0E9D682EE04B874306
        • PDF in English : JB_syr_Th_EN_V3_1.pdf (2.2 MB)
          SHA256 : 630BAD6B8B17FA748767ACFB96776BC5DFAAAFAE130F9B25F2F6CD6D1F6CECBE
        • PDF of results (appendix common to both previous files) : JB_syr_V3_1_results_1024.pdf (15.2 MB)
          SHA256 : 4245BAA2E6D5CFBC4F6EEBA956013F59EC554AE3B057A948D1FA9A93065683AD

        If a patron or donor wanted to acquire the original work that they could keep for themselves (DHTML file + Javascript + mathematical expressions written with the JTMath sub-library) :
          SHA256 : 7EE4920F1D6323FBBC0F55FD34DC8583DAE2264173E31328D7FC42E41F5AB61A (without the minimized JTMath library)
          SHA256 : 51FCE61F35C63041F53FAA24BBCFC9696D84EAF1FBFD70B4327FFCD7B2A842B8 (with the minimized JTMath library)

        Thus, this patron or donor could offer the ZIP archives with the JavaScript code for tests and the static HTML (JTMath converted), in both French and English,
        • for the proofs of the theorems for the extension of the Syracuse sequence
          SHA256 : 8EAB32BB1F39422F41BBC493C73692D65CE77B51B0C667085988806C3D82D788
        • for the proof of the theorem for the standard Syracuse sequence (without the extensions)
          SHA256 : A5F84143AFFC9C3A7939315AAE82B269FFA66CA0600C093C92BBCC3478F05199

        This would allow for quick validation of the theorem proofs (potentially by other interested parties).

        Of course, this patron or donor would be indirectly associated with this mathematical result

        A similar investment is possible for the use of the symmetric encryption program with a secret key… but also for "version 1" of the "Theorem 0" or the base encryption function (214 bytes, symbolically referencing the Syracuse sequence).

        All these files have been uploaded (and deposited) to the website https://www.copyright.eu

      • Abstract :

        First, we will list a set of definitions, a practical summary to facilitate the reading of the document.

        We will state the different theorems that we will prove in this document.

        We will first study the extensions of the Syracuse sequence defined as follows : if un is odd then un+1=3un+b (with b odd, b=1 for the standard case)
        If, at first, we focus on the classical Syracuse conjecture, we can ignore the texts written in green which globally correspond to the extension b≠1.

        We will retrieve Shalom Eliahou's result (the link between the minimum length of a non-trivial cycle and the maximum value for which the conjecture has been verified) in another way, using a JGL boundary transition list, which will serve us for the following reasoning.
        The section V, the most delicate and the longest, which mathematically proves this result observed by calculations, can be ignored at a first reading.

        Thanks to a new Theorem, we will be able to prove that the elements of a non-trivial cycle are necessarily in an interval that we will specify.

        We can then easily conclude that there are no non-trivial cycles, the lower bound of this interval being greater than the maximum possible value for having a cycle.
        A similar reasoning will allow us to conclude that there is no divergence towards infinity.
        The theoretical results of the method are consistent with the tests and also with the Eric Roosendaal's list of records found up to that date.

        To verify the robustness of the method, we will study the case un+1=5un+1 and prove that there is at least one value for which the sequence diverges.

        The proofs, which can be improved, only use fairly simple mathematical reasoning and can be understood in the smallest details by many interested people, with everyone certainly able to understand the main points.

        The purpose of the document will be achieved, but, to go further, I have stated my own conjecture !

    Revolutionary numbers

    • Definitions

      The Syracuse Conjecture is certainly one of the few open mathematical problems whose statement is understandable by all, that is why it is so fascinating.

      I call a revolutionary number, in the sense of the Syracuse sequence, a number such that at the end of a certain number of steps in the Syracuse sequence (standard or reduced), has again a value very close to the starting number.
      So after a very chaotic journey, we almost did a cycle, which is like a revolution.
      As a reminder, if there was a non-trivial cycle, different from (4,2,1), then the Syracuse conjecture would be false, which gives mathematical interest to these numbers.
      For purists, it is a number such that there exists N for which the relative difference between u(N) and u(0) is very small, which means that we have "almost" a cycle for u(0), let |u(N) / u(0) - 1| < ε, ε being small. u representing the standard Syracuse sequence, u(n+1) = u(n)/2 if u(n) is even, and u(n+1) = 3×u(n)+1 if u(n) is odd.

      We will call v, the reduced Syracuse sequence, defined as follows: v(n+1) = v(n)/2 if v(n) is even, and v(n+1) = (3×u(n)+1)/2 if v(n) is odd.
    • Example

      It's a revolutionary number with ε = 4.366×10-5 for the Syracuse sequence, so with a relative difference of a few hundred thousandths.
      It also has a signature including in particular a serial number (SN) which is equal to 0.
      With all of these properties, there is theoretically less than a 1 in 10100 chance of finding one randomly, which means that it would be easier to mine all the remaining bitcoins.
      This fits modestly into a much larger mathematical problem: Easy to check, easy to find? Or easy to verify, difficult to forge
      The jewel is here: Example of revolutionary number and digital ticket based on the result of Syracuse
      To check, click the "Compute" button and look at the result in the "List of steps written as string of ASCII characters" text box, where SN represents the serial number of the ticket.
      You will also have all the details on the precision, as well as the plot of the trajectory (be careful, the scale is logarithmic) even until it reaches the value 1 after 6957 steps.
      You can see the results in this PDF file: jb_syr_sn_0_en.pdf
    • Purchase of revolutionary numbers

      Please enter your e-mail address correctly since the numbers will be sent directly to it.
      e-mail:
      Number of numbers:  Unit price: 10.54€
      Comment : Non
      Amount proposal for S :
      Order amount:
      Would you like to be an affiliate member?
      More information below
      No Yes

      These numbers have the same properties as the one in the example, their Serial Number (SN) is greater than 1,000,000.
      The lower the SN serial number, the more searched it will be.
      Keep your numbers in a safe place, you will be able to marvel at the result, everything is vibration including these numbers ... it's magic!
      Whether you are passionate about the Syracuse conjecture, collector or speculator, thank you in advance for showing me a little sympathy
    • Verification

      To check a number, you can program the Syracuse sequence, it is very simple for small numbers, a little more complicated for "large" numbers (greater than 128 bits) but there are public libraries for these calculations.
      You will find here an implementation in Javascript which has the advantage of working in an Internet browser: https://www.bajaxe.com/jb421/jb421_verif.html?lg=en&type=red
      Check that the type of Syracuse sequence is reduced, which should be the case with the link
      You just have to specify your starting number, the number of desired transitions and click on the "Compute" button
      You will also have all the details on the precision, as well as the plot of the trajectory (be careful, the scale is logarithmic) even until it reaches the value 1 if you check the option.


In addition : Some old projects
Jacques BALLASI, ENSIMAG engineer 1989 (France)
Contact: You can contact me at the following address: