Judging WC26

I specifically have been given the opportunity to judge the Pen Spinning World Cup 2026. Probably as much a surprise to the organisers as it was to me, since it was decently well known that I was not exactly the type of person who regularly accepted requests to judge large events. The reason for that is simple: it takes a lot of effort and if you want to do it decent, it takes more effort. To do it well? An absurd amount of effort.

This time around I have some new motivations that required me to accept. In short, pen spinning judging is a complete unstructured mess and it hasn’t been improving for decades. Judges that understand both the wide range of the hobby that is required to even critique pen spinning of the highest level are quite rare, and those who actually try to build a comprehensive system for even their own judging are like unicorns and fairies, they just don’t exist.

Until now.

Instead of handwaving issues away as “we’ll fix it later”, I’m going to be handling judging this event quite a bit differently than we did in the past. First, I have been at work on a large scale research document (it’s 22 A4 pages currently) that explores and expands upon the nature of judging pen spinning as a whole. This will not be done in time for the release of the World Cup rulebook, too bad! I have, however, passed it on in the world cup judging channels (which have been silent as a crypt, perhaps not the best sign), so that some parts of it can be used.

I’ve pushed for two main things off the bat in the document that I have shared: better aggregation methods and actual descriptions of score mappings that judges should be using. The former is rather obvious, ranked choice aggregration (ranked choice voting) is almost a completely solved problem, but we consistently use systems like ranked sum and ranked mean (mathematically equivalent of one another, by the way) that produce unsatisfying, easily distorted results. The history is plain to see because I wrote it all out in depth in my advisory document, we have been doing this forever and it still sucks. Instead I’ve suggested we use pairwise-majority methods, we shall see if my suggestion is heeded.

The matter of score mapping is to do with how judges score a criterion on its range for a given performance. Currently judges don’t seem to have any consistency, if at all with it, and this leads to bias at worse and distorted results with a skewed meta at the worse. Did you know that many judges instinctively taper off their scores aggressively after the 4-6 middle range? This leads to each additional half point being increasingly difficult to obtain. The net result is that the best way to gain an advantage is by being decisively mediocre or slightly above mediocre across all the scoring categories, since if you pour exceptional effort into one criterion you may only gain a point or two, but lose more than that elsewhere due to this effect.

This is an absurdity to me, I do not want to see combos that express mediocrity in a criteria, I want to see combos that express exceptionalism, so I proposed a cubic centered power mapping that causes scores to accelerate towards the maximums as you pass beyond the middle hump of average. How shall we implement this? Simple, I shall simply judge all the scores as linearly as I can manage, then apply the power mapping calculation to get the final scores, shrinking the cluster of average towards ‘5’ while punishing far below average and rewarding far above average spinning. Does this mean we are going to see more 9s and 10s on scoresheets? Yes, and I don’t care. It’ll probably only be on mine anyway, I do know that convincing certain individuals use even a 0-10 scale is difficult.

In addition, I’m applying a divide and conquer approach to judging all criteria from now on. All of your scores will be balanced against the totality of all spinners, individual matchups are not being compared (in fact, I would prefer to not even know who is battling who, I do not care). With divide and conquer, the consistency of each score relative to all other spinners will be absolute in order if not in proportion (where errors can still be made). Gone are the days of saying “How in the hell did I get less difficulty than that?”, and having to assume the judge just got confused while judging 100 combos. If I score your difficulty below someone, it means I deliberately and specifically compared you to either them or one who performed worse than them, and you were found wanting.

The rest of the document will come out in time, it’s going to end up serving as a judging advisory to all pen spinning. It’s a lot more in depth and detailed than we’ve had on the topic in probably 10+ years, I can say this without even the slightest hesitation for we have not at all been taking the conversation seriously. It’s a matter of breaking pen spinning down into its measurable, observable components with clean boundaries between and justifying every one. We haven’t done any of that, we’ve kind of just picked some wide themes we like without a lot of consideration and made them a touch more complicated over time. I’ve started from them as a baseline but I think they might have to be thrown out or the definitions greatly tightened, it’s just not good enough.

Ah, and one last word. I have some words of warning for the niche technique (cough multipen) users among you: I’m also working on ways to grade you properly against actual pen spinning. Personally, it should probably be its own discipline soon if it keeps on the way it has, with not as much overlap with pen spinning as I would like for something in pen spinning competitions, but I am also not the organiser, so I will judge it as though it is allowed. Multipen in particular has had a very fun run lately where it seems to be softened in terms of execution penalties and massively buffed in originality and difficulty.

My approach for judging multipen is going to be as follows: The fact you are using the technique is irrelevant. When you show off original elements, the fact that the element is new is going to be under a lot more scrutiny. A lot of the tricks that are being shown off are the equivalent of creating a sonic variation in the 2000s, and unfortunately these are not going to be showered with difficulty and originality points, so be warned. If you want the equivalent of something that is scored in difficulty like a high level power sequence, then that sequence you show off had better take just as many years to train. Also, I am not lowering the bar for execution or structure. You should note that with my centered power mapping that performing middling in a criteria gets middling scores, and doing worse than middling will be so punishing that I am certain my scoresheets will have a reputation for tyranny by the end of the tournament.

Obviously, if you do something that shows real depth, difficulty etc etc in multipen you shall be awarded ample points, just make sure you don’t do anything silly like full hand grab -> put pen into slot x3 in the middle of your combo because I will display levels of authoritarianism usually reserved for use by for your favourite military dictatorship of choice.