Live vs. Video

brianbear · Post by **brianbear** » 12 Apr 2009 14:24

I was discussing cleanliness of tricks (whirls especially) in another topic and was thinking...
should big ones and sick threes, and maybe even shred 30 be judged on video in slow motion instead of live, I understand it would take to long for routines (I also feel like if of one or two tricks looks "the" live in a routine, it shouldn't have a big impact on the overall performance of the 2 minute shred). Any thoughts on this?

Jeremy · Post by **Jeremy** » 12 Apr 2009 19:21

Shred 30 is always judged on video (or almost always, maybe there are exceptions).

I think it's fine to judge sick 1 and sick 3 live. In no other sport are judges perfect and so we shouldn't expect them to be perfect in footbag. It's much better for the crowd to give results quickly, and having an impressive performance for the crowd is more important to me than getting all the results *absolutely* right. That doesn't mean we shouldn't point out mistakes when they're made, because you can't learn from mistakes if you don't know they exist.

Kimbe. · Post by **Kimbe.** » 14 Apr 2009 12:04

I agree with Jeremy about sick 1 and 3 judging. If there's no fuss about the cleanliness of a performance live, why would it have to be analyzed afterwards through video? If there is though, and the cleanliness is an important factor for the judging, then why not use highspeed cameras.

However, sometimes I feel like this is becoming too big an issue: if something is clearly "the", then there shouldn't even be an argument about it. As for a routine, I simply don't understand how one trick could affect the judging in a way that it would lose to another routine.

HackyRichard · Post by **HackyRichard** » 16 Apr 2009 16:19

Worlds in San Fran ('02, I think), Sick 3 Finals Sunil Jani - The Enforcer calling it as he sees it, that's how it should be done - live, direct, no BS.

If the judges can't make a firm decision on the spot, then I think you need to find you some new judges (or at least have systems in place that deal with "close call/split panel" situations when they arise)

I guess I'm thirding (is that even a word) Jeremy's position, keep it short and simple, after all the actual footbag action is short why make the judging disproportionately longer, don't make the ADD folk suffer any more than they need to

FlexThis · Post by **FlexThis** » 28 Apr 2009 12:34

Worlds in San Fran ('02, I think), Sick 3 Finals Sunil Jani - The Enforcer calling it as he sees it, that's how it should be done - live, direct, no BS.

I disagree with that statement. That would be completely subjective and would be BS.

I have been saying for over 10 years now that if you want to get technical with any comp and still present results in a timely fashion the answer is simple:

- Film the event in real time
- Use a TIVO or DVR to watch the performance in slow mo as the comp is taking place and being filmed.
- Judges marking scores for "dexes", cleanliness, etc... can watch the slow mo version and mark scores accordingly.

Since the performance is being judged as it is being performed, the results will only take as long as the slow mo playback. And since the slow mo playback is happening simultaneously, then there will be a time over lap between live film and slow mo. If the slow mo is 4x and the routine is 30 seconds, then the final score will take exactly 2 minutes from the start of the performance. That is fast IMO.

At a slower playback it will take a little longer. 4x should be fast enough. At this rate a 2 minute routine would take 6 minutes after the routine has finished, or 8 minutes from the start of the routine. I still find this to be incredibly fast and NOW accurate as well.

comments?

Tsiangkun · Post by **Tsiangkun** » 28 Apr 2009 12:42

It's true, for nearly 10 years we have been saying this, but nobody has shown themselves to be interested in having it done that way.

Jeremy · Post by **Jeremy** » 28 Apr 2009 21:19

Given how few moves there are in a routine that are 'the' and how little impact it has on the rankings for the event (I can't recall a single time where the cleanliness of moves has impacted the results), I think it's completely needless over kill.

The reason shred 30 is judged via video isn't to check if the moves are clean, but because it's difficult to write out each move in real time, getting the side of the body correct, especially with drops - or more difficult to keep track of what moves are not unique if you're just counting the scoring components.

Sick 1/sick 3 is perhaps the only event where 'the' moves can play a significant role, although again, I've never seen that actually happen in a serious competition.

Choosing to needlessly make competitions run for longer than necessary would be a bad move in my opinion.

Tsiangkun · Post by **Tsiangkun** » 29 Apr 2009 07:37

For tens years, there has been no need to check which side of the body , uniques, toe to clipper ratios, number of spins etc etc etc.

All that is needed to score shred thirty is a list of strings, and which side the first trick of the string starts on. Slower than live video makes it possible to score a shred 30 in about 42 seconds.

Jeremy · Post by **Jeremy** » 29 Apr 2009 15:27

I'd love to see you do that.

If you have two people doing it, you can have one call out each trick and the side it starts on for each run - but then you have to write out all the add values, count the uniques and do the calculation. If you can do all that, after already watched the 30 second video in slower than normal time, in 42 seconds, I'll be very impressed.

I'm not sure what your first point is. I agree there has been no need to know those exact numbers, and as far as I'm aware they haven't been counted either. It's easy to get quick idea of those things just by watching.

Tsiangkun · Post by **Tsiangkun** » 29 Apr 2009 16:21

my code can count anything that I have ever thought of counting. It just needs the name of tricks.

I spent a lot of time in the 90s scoring videos in slower motion. Two people is exactly what makes this possible. One person calls out the tricks, while another types them in. Aliases for tricks can be defined on the fly, so instead of typing butterfly, 'b' will suffice. New moves that the code doesn't know pops up windows once all the tricks are entered, and adds them to the database it uses before scoring the person.

I type very fast, it may take other people a lot longer.

The first part was just saying that my code provides a much more comprehensive list of what was done by a player than we currently track with paper and pencils. It does this because anything that can be counted, can be used as part of a scoring equation. Somebody might care about how balanced tricks are between the two toe surfaces, or how many spins were done on each side, etc etc etc

42 seconds is about the maximum number of tricks in a shred 30. I type 75 words per minute. Because the program will prompt for mis-spelled words, and short hand aliases can be defined in advance, and on the fly, the program makes entering moves very fast. Most common tricks can be given a one , two, or three letter abbreviation. Further more, by taking advantage of case, <P>ixie can be distinguised from <p>aradox, <S>pinning from <s>ymposium, and <D>ucking from <d>iving.

Scoring shred in 42 second is doable while being able to type at a rate significantly less than one word per second. The vocabulary being typed is very small, and the words can be redefined to be very short.

One last advantage of scoring shred by the list of tricks that was hit, is that it preserves that record of what was hit, and comparisons can be made across events by more than just a number. Results can be recomputed as scoring algorithms evolve to allow a comparison of past and current results.

crazylegs32 · Post by **crazylegs32** » 12 Aug 2009 21:25

the newcomers in footbag arent too concerned about events? Nobody does anything about anything?

Cameron- I believe you so post up results and vids of some routines/