AIDA Results Register

[trux]

I think average value should come from best result per athlete and not all of results.

Yes, it would be better, but it would require an additional database parsing cycle , or an additional index which is in this state not easily possible. Without preparsing, it would slow down considerably the system. I'll keep it on the to-do list.

But in the same time, in this way it will make everyone feeling better, since it is easier to get above the average

 

[tim_a]

But in the same time, in this way it will make everyone feeling better, since it is easier to get above the average

And because freediving is a very mental sport you should just leave it as it is.

 

[Kattie]

Great work! I find the graph particularly useful. Even though I work with stats every single day it took me a while to understand the tables though. It would be useful to add a legend or a short explanation.

I think that both the tables and the graphs have a huge potential for personal monitoring of performances. Thanks for doing this!!!

Eventually, if you feel bored and have nothing else to do, you may want to add something that would track the performances of athletes over time. It would be really interesting for newer athletes like me, for instance, to be able to see how the performances of athletes that have been in the sport for a long time have evolved over time.

Once again, well done! The stats on the AIDA international website exclude some competitors. Mistake or intentional?? In any case, I am sure that your site will remain the reference for all those really interested in the sport .

 

[trux]

Great work! I find the graph particularly useful. Even though I work with stats every single day it took me a while to understand the tables though. It would be useful to add a legend or a short explanation.

Thanks! Yes, I am aware some more legends and a documentation are needed, but that's a boring work, requiring a lot of time, and language skills (all texts need to be translated to all three language versions). So it is buried relatively deep on my to-do list. But if there is some volunteer willing to work on it, I am always ready to accept a helping hand

Do use, though, the mouse on the page as much as you can - there is a lot of additional information about labels and values hidden in the mouse-over tips. Not everything is described, but often you find some explanations. I find describing elements in this way much more intuitive and quicker to find, then cluttering the page with legends, or writing a separate documentation.

Eventually, if you feel bored and have nothing else to do, you may want to add something that would track the performances of athletes over time.

Yes, this is just under testing, and will be available soon.

'PS: I've got you PM too, and will look at the cards later

 

[trux]

OK, so the evolution graph is now enabled too. It is just the first test version. There are some more changes and improvements planned, but basically it should work. By default it shows all six disciplines combined in the same chart, so it is not always very well readable. In such case I recommend selecting a discipline in the top menu, and looking at the discipline evolution charts one at a time. I'll try incorporating hiding/displaying of individual disciplines directly into the graph, but am not yet completely sure whether it will be possible.

 

[osusim]

Excellent!
Fault: In case of only 1 performance per discipline the point not presents at combined graph. If the the discipline selected then both images (the cube and evolution graph) are gone. I guess a crash of a script happens.

 

[ericvrp]

Hi Ivo,

The spiderweb looks really nice with lots of information presented in a clear way.

One thing that clutters the picture are the outer rings of the web. Perhaps the best performance should be positioned on the outer edges instead of a fixed number of points (150?)?

cheers
Eric

 

[Kattie]

Thanks for the evolution graphs! Really interesting!

 

[trux]

Fault: In case of only 1 performance per discipline the point not presents at combined graph. If the the discipline selected then both images (the cube and evolution graph) are gone. I guess a crash of a script happens.

Well, the graph is suppressed on purpose in case of only one recorded performance (no evolution to show, hence no point in displaying it). The suppressing of the radar graph in such case is now fixed. Thanks.

One thing that clutters the picture are the outer rings of the web. Perhaps the best performance should be positioned on the outer edges instead of a fixed number of points (150?)?

Not sure if I understand what is disturbing. Do you mean the light gray radar web? I made it explicitly almost invisible, exactly for that reason, but perhaps on some screens with higher contrast it still shows up too strong. I can make it even lighter, but fear that most people will not see it at all then.

The scale is set to a fix maximum of 150 points in all axes after considering and weighting different options, and I believe it is the best choice. Using variable scale by putting the world's best performances on the outer edge would 1) remove the information about which discipline has more weight in combinations, 2) the performance graph area would become wildly variable with time and depending on the selected scope (year, gender, federation,...), 3) Individual profiles and categories would be no more comparable to each other

 

[trux]

OK, Eric, to make you happy, I added another option that will turn the graph scale from absolute (0-150 points) to relative (0-100% of world's best performance in each discipline). Currently you can only access it by manually adding the tag "relative" to the permalink URL, but I'll integrate the toggling of the absolute/relative mode directly into the graph later.

This is an example URL how to get the relative radar chart:

http://apnea.cz/ranking.html?pid:Riet_Paap+relative

 

[trux]

There is now a link above the radar graph telling "show relative graphs" - when you click it, the coordiantes of both graphs will change from absolute scale in points, to relative scale in percent of the world's best performance in each discipline. Getting back the graphs to the absolute scale is then possible in the same way.

 

[trux]

I have some other graphs and statistics in plan, but would like to have some feedback to know what would people interest the most. So if you have any wishes or ideas, do not hesitate to post here (or send privately if you prefer).

 

[osusim]

May be add some option of smart filtering to the evolution graph. Nobody care about 20 sec STA from a person with average 6 min STA or 50m DYN from a person with average 200m DYN if that was not the first official result.

 

[trux]

Yes, I was thinking about that, and already filtered out DQ's and DNS. However, I am not sure whether filtering unusually low results would not be too much. Finally it gives some information too, especially if it happens more frequently. And then the question is when I filter out all bail-outs of say 5% of the PB, why shouldn't I filter out also performances when the competitors quits at 50 or 60% of his PB. It if finally the same phenomena. So this is a bit controversial issue, but I could perhaps add some option where the visitor chooses himself whether all performances are plotted to the graph, or only some of them. Another option would be some median smooth curve, but I would need different tools for that.

 

[trux]

The result files from competitions I receive (if lucky) often contain errors, or are in unsuitable formats, and more errors are introduced then during the conversion and import process. It is already pretty hard and time-consuming job to enter all the results as they are, but when I am then swamped with error reports, and with change requests from competitors, it is really frustrating, and taking more and more time.

My plan always was automating the ranking and results handling as much as possible. Today, I completed an important step - user interface for editing the results

​

It means that now anyone with an account at Apnea.cz, can edit the ranking results, or even add new still missing ones. You can fix errors, add missing data, assign national, continental, and world records if omitted, link photos, videos, or dive profiles to individual performances, etc.

If you are interested, have a look at the detailed description of the Result Editor here: APNEA.cz - Blog

If you do not have an account at Apnea.cz, and need to edit some results, you can register for free. And if you have problems passing the entry theory test, let me know. At competitors or certified freedivers, I can make an exception and let you in without the test.

 

[trux]

You can now also add missing results - at the bottom of any page with results, there is a link "add a missing result". Click on it, and a blank form will pop up. Fill the data, add a link to the evidence for verification, or comment the reasons, and submit.

Bulk editor, and file import is in works, but it will take some more time to get it working.

Please help making the ranking better, more complete and more accurate! And please give me some feedback. Sometimes I am not sure whether someone cares at all, and whether I am not just wasting my time for nothing.

 

[aris ioannidis]

Trux, i think what you are doing is great; pls keep it up!

 

[trux]

Thanks, Aris!

I just added another new feature: when editing the results, you can now also add a photo, video, or a dive profile to the performance. Just click the link "add/edit performance media" at the bottom of the Result Editor form, and enter the address of the image or video files. Currently they already have to be hosted somewhere on the web, but I'll add the possibility to upload them from your PC later too, if I see the feature is really used.

 

[Kars]

I like it Ivo, I would love to see the dive profiles of the deep diving competitions.

 

[trux]

Well, if someone enters them, they will be there for everyone The problem is everyone would love to see them there too, but no-one is interested in helping and completing the data, so I am afraid that unless I'll do it myself, there won't be much to see