Solutions to “Just KNIME It!” Challenge 07 - Season 2

Hi all,
Here is my solution.

Line plot was used for visualization.
In order to draw the end points of Pass, data was added with coordinates slightly shifted from endY.
This process was also necessary to prevent Pass from being connected to each other.
I think there are better ways.


Hi all,
Here is my solution.

I have created my workflow for analysis at the following points.
・Back pass or forward pass?
・Time Zone
・Pass distance (plot size is the pass distance)

Back passes have a higher success rate.
ManchesterCity has a higher percentage of back passes than ManchesterUnited.


Interesting! I had never heard of a pass map!


Hi All- here is my solution for this week. Overall, i see some people had similar idea in mind as what i had. I built in a Filter-by-Team functionality and then added time intervals in order to add a bit of granularity… all of the data at once was not really informative. I did use a refresh button (instead of using the re-execution setting of the multiple selection widgets) to allow for changing multiple filter settings at once before the refresh occured… much smoother this way. When both teams are added to the visualization, i colored the markers by home jersey color. The markers are placed at the starting point of the pass attempt. The color indicates successful (green) or failed (red) passing attempt.

That one was fun! Thanks for this challenge.


Hello KNIMErs
This is my take on S2 Challenge 07. I thought at the beginning as a good idea to try it with Py, but it took me more time than expected. Besides I have the feeling that R allows more creativity and more flexible too in combining charts :thinking: Good lessons learnt and fun anyhow.

PS.- @l6fader I like the way that you use the Multiple Selection Widget. I’m sure that I will implemented it this way quite often from now on.



@HeatherPikairos this is really a cool solution. I learned a few tricks and so I am really grateful for this contribution. Very nice job!


Hello JKIers,
Just for the records. I’m still working in the workflow with some upgrades. Pass arrows layer has been added to the dashboard. As an example, here in the picture there’s a sample plot of Manchester City’s long distance passes [>40 pitch-units ] (be aware that units are provided in X and Y percentage, so units need to be transformed over pitch shape).

I’ve just realised fron @HeatherPikairos’ post, that I was using the same FCPython resource for the contour maps. So I could add easily the pass arrows’ map on top.

Enjoy coding. BR


Hi Crazy KNIMErs :crazy_face:.

No me quedé muy contento :face_with_diagonal_mouth:con mi solución, y la he mejorado un poquito este fin de semana. :smiling_face:




Ps: source of Python Code: Drawing a Pass Map in Python - FC Python :ok_hand:

1 Like

Hello Community,

I saw many exciting solutions and someone very interested supported by the use of Python programming to draw a Soccer Field.

I developed my solution based on the flows published for other members.

I share the following carpet: JKISeason2-7 – KNIME Community Hub, this contains the same solution showed in different presentations.

All have the following information based on both teams, full stats from the data source and only the information relevant to the passes, grouped by each halftime.


Thanks @l6fader! I’m glad you found it useful! I was learning too :slight_smile:


My solution to Challenge 7


Hi, KNIMEr, :partying_face: :partying_face: :partying_face:

Here is mine

Points worth mentioning

  • I chose a slider to control the time frame.
  • Well, in addition to the visualisation of Manchester City’s successful passes, I also showed the failed passes, as well as Manchester United’s passes. This is because I believe that the success or failure of one team’s pass is closely related to that of the other. Only the passes are shown here.

Another thoughts

  • Matplotlib is a relatively slow solution.
  • 2D Density Plot is a great visual solution in this scenario.
  • The sunburst chart is also great.
  • Obviously there are many factors that influence the success of the pass. In addition to the pass, there are many other types of data, but it is a pity that there is not so much time to analyse them.



I wrote a relatively detailed analysis post in Chinese. If you are interested, you can take a look(maybe use translator?).


Dear KNIMErs,

alright, here’s my solution. Looking at the submissions so far, it feels a little bit “lame” because I have now soccer field visualized etc :wink:

Learned again a lot about re-executing when making changes in this little dashboard and also used some Auto-Binner functionality to bin some “match times” and the Euclidean Distance (aka movements in the 2D space rather than just from A to B) :slight_smile:

Never thought I actually will like mathematics again hahaha :smiley:


1 Like

Hi Everyone :slight_smile:

I just want to share an updated version of my workflow (in case anyone enjoyed the video!) because I had an issue generating a full video in AVI format and I have now received a solution from @gab1one on the forum:

The NativeQT format in -Image Writer- successfully generates the full video :slight_smile:

I have also added a little bit extra at the end of the workflow to compress the video and delete the original version in the data folder:

The original link to the hub workflow should work but I’ll share again:

Just for fun, I made a video with the passes of both teams in different colours, you can watch it here:

Best wishes


:nerd_face: As always on Tuesdays, here’s our solution to last week’s challenge! :nerd_face:

:soccer: Since this was an easy challenge, we kept the solution very simple and didactic, relying on a heat map that is quick to build with the 2D Density Plot (Plotly) node. As you can see, most of Manchester City’s successful passes are initiated from the center of the field.

:boom: Once again your incredible solutions AMAZE us! :exploding_head: This year it seems like you folks are really going the extra mile and really exploring how KNIME can be used. We are so happy! This really means a lot to us. :blush:

See you tomorrow for a new challenge!


@HeatherPikairos @gab1one
I found this solution to be very cool (as i mentioned above) and immediately tried adapting it to a real use case in my regular job. I generate the PNGs by converting from SVGs generated by the generic javascript node. What i find is that when i use the PNG to ImgPlus node, the image loses resolution and the colors really get messed up. Do either of you know how to overcome this?


Hi @l6fader thanks again for the comment about my workflow :slight_smile:
Unfortunately as I am a newbie in this area I cannot help to answer your query, but hopefully @gab1one can :slight_smile:


This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.