Are Rhythmic Catches (RC) and IMBO any good? In this article I go over the definitions of different aspects of juggling patterns, the goals of RC and IMBO, desiderata of the Notational Engineering Laboratory, some metrics from my previous posts and some usage anecdotes.

I also made some comparisons with Harmonic Throws, and of course there is a conclusion to this whole evaluation.

Patterns, throws, orientations, poses

In the process of this research I have come to think of all toss juggling tricks as consisting of 4 different elements. I call these “patterns”, “throws”, “orientations” and “poses”. I will explain my definitions of these terms:

Patterns deal with the timings of throws and catches, and the order of manipulation sites. Patterns describe whether a prop goes from one hand to the other, or whether it returns to the same hand, or anywhere else. Siteswap would be a form of pattern notation, and so would the top layer of Harmonic Throws.

Throws are about individual flight paths of objects. Describing a throw may involve describing a pathway relative to the jugglers body or to other objects. BTN would be a form of throw notation, and so would the middle layer of Harmonic Throws.

Orientations deal with the rotation of props. For example a club may be pointed in various directions, and may spin around various axis. For small juggling balls their orientation and rotation is normally irrelevant, but for less symmetrical props their rotation can be the defining feature of a juggling trick. Rotation Notation is an example of a, well, rotation notation.

Poses are about the movement and poses of the body. It may not always be clear when a movement is part of a juggling trick, or when it is part of dance. For example, a pirouette with props in the air, or lifting your legs during an under the leg throw, tend to be regarded as part of the juggling trick. But walking forward whilst juggling a cascade, does that make it a different trick than without the walking? There are a lot of dance notation systems such as labanotation, and also the bottom layer of Harmonic Throws would be considered pose notation.

This project has focused on patterns and throws, orientations and poses have mostly been disregarded.


When this project started, I had 6 of goals in mind of what the systems should be capable of, but since I never published these, let me write about them now.

  1. A new system should describe patterns and throws for the solo juggler
    I disregard orientations and poses and multi person patterns, but want to describe as much of the imaginable patterns and throws as possible.

  2. The system should for body interaction
    It must be possible to describe patterns that deal with props caught and thrown and hit with body parts other than hands.

  3. It should be able to describe paths around the body
    It must be possible to describe body throws such as backcrosses and under the leg throws without using names for these moves.

  4. It should do so elegantly
    The system should not provide redundant information, and be as easy to read, write and learn as possible.

  5. It has to improve upon existing notation systems
    If an existing system can do all that a new system can do, then what’s the point of a new one?

  6. To be suitable for simulation
    I want software to be able to parse this system, so that we can move beyond just siteswap simulators and have juggling simulators that can for example also simulate body throws.

Did I achieve these goals?

I’d say that yes, very clearly Rythmic Catches and IMBO together can describe patterns and throws, as proven in the Notation Comparison and in Notating tricky tricks. RC specifically allows for body interaction, and IMBO specifically describes paths around the body. This covers goal 1, 2 and 3.

As for goal 4, in the Notation comparison I make a strong argument for why RC and IMBO are among the more elegant systems. They don’t require the use of left and right, they don’t require setup information, they make clever use of default information and they are non graphical, which allows them to be relatively simple.

Although not all existing notation systems were tested, I believe that RC and IMBO indeed improve upon many if not all existing juggling notation systems in significant ways, which is goal 5. This is also the outcome of the Notation comparison where RC and IMBO were compared with 6 other systems. This does not mean that there is no place for the other systems, many of them fill specific niches or may be simpler to use in a context were less complexity is needed. Many of the features of RC and IMBO are similar to the features in Harmonic Throws, this will be discussed further in the dedicated section below.

Goal 6 may be the hardest to test, as there is not yet any simulation software available that can deal with RC, IMBO or both. Such software has been in development by Ilse Arwert and myself, but at the time of writing this is not yet finished enough to do some testing with, and will likely never be completed due to the scale of the project. However as RC and IMBO are text based and can therefor easily be parsed by computers, and because they attempt to take away a lot of ambiguity compared to existing systems, I believe they would be very good candidates for future simulators.

Desiderata for Revolutionary New Notations

The Notational Engineering Laboratory published an article in 1995 about desired qualities of new notation systems. I am not convinced this list is very good or very relevant, but it is arguably better to have at least a look at this than to only test my own systems by my own standards.

These desiderata are:

  • Reify a New Ontological Dimension
  • Represent Processes Well
  • Discover New Relationships
  • Provide a New Integration
  • Improve Ease of Use
  • Assign Meaning by Place Not Neighbors
  • Fully Utilize a New Media

From the little text and examples that are available I am not completely sure I understand all of these fully, but I will do my best attempt at briefly discussing each of these in regards to RC and IMBO:

Reify a New Ontological Dimension

IMBO can represent when a throw is made in relation to to the pathways concept as introduced by BTN.

RC can represent the timing of actions in relationship to the moment of catching.

Both are new concepts which the notations reify.

Represent Processes Well

RC builds upon many other numerical juggling representations such as siteswap, which have proven that they can represent juggling very well, can work to prescribe complex patterns and can work to generate new ones. It is expected that with the higher amount of possible detail with RC than with siteswap an even more realistic representation can be made, when needed.

IMBO builds upon BTN which was such a useful representation that it has made its way into newer systems such as Graphic Juggling Notation and Harmonic Throws. Again, as there is more detail and less ambiguity in IMBO than in BTN an even more realistic and useful representation can be made.

Discover New Relationships

Amongst other things, RC helps solidify the understanding of timing in juggling. I for one discovered how the throw timing is influenced by zips in a pattern such as in siteswap 441 or RC 2s(2s,0)- only when I tried to understand RC. I believe there are a lot of misconceptions among jugglers about timings of throws and catches among jugglers, and I think RC can help clear these up.

IMBO clarifies how the position of throwing and catching on a given ball path changes what body throw we perceive it as. This understanding was not yet discovered by me until I conceived IMBO and tried to understand its rules. IMBO also made be understand the relationships between penguin catches, overhead throws and backcrosses (Mp, Mo, M) in a way I never had before.

Provide a New Integration

RC and IMBO are succinct representations of the essences of juggling tricks.

RC builds upon the existing grammars in numerical juggling notation systems. IMBO creates a new grammar (using existing text characters) for it’s new abstraction of body throws.

Improve Ease of Use

Compared to the long descriptions needed in analyzing ball tricks, the transcription of those tricks using RC & IMBO was considerably easier to write and read, and of course a lot shorter.

A stronger comparison with other systems could be made, but I have detailed why I consider RC & IMBO much more elegant or simple than many of its competitors.

Assign Meaning by Place Not Neighbors

The meanings of symbols in RC and IMBO is not influenced by their neighbors, there are no symbols with more than one meaning.

Fully Utilize a New Media

A lot of communication about juggling is done through computers. Therefor only easily accessible text characters were used in IMBO and RC.


Lets summarize some of the “metrics” from the evaluation of RC and IMBO here. I put “metrics” in quotation marks, as despite my rigorous research, the tests that I have put RC and IMBO through were all developed by myself and their results are to some degree opinion based. More details on that can be found in the posts explaining those tests themselves.


Despite the short existence of the new systems I have already found use for them

Using RC to rehearse a show

In April this year I was invited to substitute a juggler in a juggling performance. I had to learn a lot of juggling choreography in very little time. To make this easier, I wrote down all of the juggling routines in an early version of Rhythmic Catches.

Not only did the process of writing help me to understand what happened in the sequences, also it proved to be very useful in rehearsals to be able to quickly check what I was supposed to do at any given time. This was easier than having to find the video, scroll to the right moment in time and then figure out what the original cast was doing. It was also easier than asking the other cast members, as they often did not know exactly my part throw by throw. I could not have learnt the routines as fast as I did without Rhythmic Catches.

As I wrote the Rhythmic Catches notation on paper I could easily compliment it with some small notes about choreography or repetition of sections. Also, it was easy to edit the notation quickly with a pencil during the rehearsals.

substitution score
Juggling routines in RC to learn a show. Image has been edited to remove names.

Using IMBO in workshops

I have given two workshops to jugglers about IMBO. The first was less specific, I did not teach any of the notation but have used the concepts from it such as how the release point influences the kind of body throw. IMBO gave me a very clear structure of how to teach a workshop on body throws in general, without limiting myself to the few that I know names of. It also made it easier to give a definition to the concept of “body throws”.

In the second workshop I did teach the notation. There were a lot of interesting questions about how to deal with exceptions, and I was excited that I could answer all of them with IMBO and RC combined. IMBO served as a great tool to give these participating jugglers a more systematic lens of looking at body throws, and they used this to develop their own short routines with throws and catches that they had never even considered before.

Comparison with Harmonic Throws

Harmonic Throws (HT) is a notation system that is currently actively being developed. It is a great inspiration for this research project, and I’ve had a lot of support from it’s originator Jonathan Lardillier for RC and IMBO.

Because there are many similarities between RC & IMBO and Harmonic Throws I wanted to briefly discuss the similarities and differences so they can be better understood.

Also, it is worth mentioning that when I started with this research I was only aware of Harmonic Throws v1 (grid system), as was published in 2019. What I did not know is that Lardillier and his team have been working on a improved version, Harmonic Throws v2 (portée system). This new version solves many of the issues that I had with the original system, and it was some of those issues that inspired me to try and design my own systems. One of the core solutions I have in IMBO is coincidentally very similar to what has appeared in Harmonic Throws v2.

Harmonic Throws v2 is still unpublished at the time of writing, however in this comparison I will discuss the differences and similarities to v2 as this will allow this article to be relevant for longer.


  • HT and RC both are a beat-based system, suitable for representing rhythmical juggling
  • HT and RC & IMBO are both generalist systems, trying to represent as much features of toss juggling as possible. This also shows in the outcome of the notation systems comparison, where HT and RC & IMBO scored very similarly.
  • HT and IMBO both use the BTN system, improved with an extra description for going through a hole before or after a throw or catch.
  • HT and RC & IMBO both disregard rotations and orientations of props, making it most suitable for ball juggling but not for clubs and rings.


  • HT is a graphical system, RC and IMBO are written with text characters.
  • HT allows you to write down body movements by writing on the Benesh score.
  • HT has a left and a right side, making it more suitable for writing down choreographies but less suitable for writing down single tricks.
  • The counts in HT are typically represented by one number lower than in RC, so what is a 1 in RC is a 0 in HT.
  • In HT, both throws and catches have to happen exactly on a beat. This creates a very specific form of rhythmical juggling, that is unlike how most jugglers first learn to juggle. In RC, only catches happen on a beat, and the exact throw time is unspecified. This allows things like RC 0 which would be a hypothetical HT -1, and impossible in the HT system.
  • In HT, TAC symbols can only represent a single two hole connection at a time, in IMBO multiple holes can be represented at once. This also allows IMBO to have more options with it’s throw or catch position in a string of holes, however this is rarely useful in practice

Could Harmonic Throws replace RC and IMBO?

Currently, Harmonic Throws does not meet all of the goals I set out for RC and IMBO, including goals such as describing single tricks elegantly (because of the left and right side), being suitable for computer simulation and communication (because it is graphical), or being able to describe juggling as it is performed by myself (because of its rigid rhythmical structure). However, one could arguably create a modified version of the system that uses slightly different rules and text symbol, which could become very similar to RC & IMBO.

Do RC & IMBO then have a reason to exist? Well, to begin with, that modified version of HT does not exist and may never exist. But more importantly, I hope that both continue on to develop and grow towards their own strengths and niches. The similar concepts present in both systems may hopefully help practitioners of one system to be able to learn the other system with relative ease. Also, the presence of similar systems is indirect proof that we may both be on the right path. Similar issues (describing toss juggling with notation) may have similar optimal solutions (describing juggling with a beat based system, using a BTN hole theory derived throw notation).

This research project has certainly made me more enthusiast about Harmonic Throws, and I am looking very much forward to the publication of Harmonic Throws v2.


Rhythmic Catches and IMBO have managed to achieve all their goals, meet the desiderata of the Notational Engineering Laboratory, outperform other notation systems in a comparison test, are meaningfully different from Harmonic Throws, and have already had practical uses.

They were only able to describe a bit more than half of the tricky tricks that I tried it on, but as a juggling expert I must say that these tricks and jugglers are pushing to stretch beyond the known boundaries of juggling and therefor have tricks that are currently hard to describe.

Overall, these systems do a great job of doing what they need to do, and I consider their development to be a success.