Primary competition visual

UNICEF Arm 2030 Vision #1: Flood Prediction in Malawi

Helping Malawi
$10 000 USD
Completed (almost 6 years ago)
Forecast
1992 joined
484 active
Starti
Dec 02, 19
Closei
May 17, 20
Reveali
May 17, 20
Well done, Zindi :) (a small brief how our team got disqualified)
Connect · 30 Jun 2020, 09:48 · 6

Hello, Zindians! Well, finally after 5 weeks of a competition we got to know that our team is disqualified :) To say that I'm shocked by the explanation of Zindi is to say nothing :)

To cut long short - below you can see the explanation why we got disqualified:

"We have reviewed your submission and your code, and we, unfortunately, will be disqualifying your team. The review committee has come to this decision because the use of flood_100 violates the rule that does not permit the use of 2019 data. The ban of 2019 data is as much about the ground-truth nature of such data as it is about the ability for the government of Malawi to use these models to predict flooding in the future."

And finally, why we are shocked by that: flood_100 layer is basically a theoretical estimate of previous floods in the region. This layer is listed here (http://www.masdap.mw/layers/geonode:q100_atkins_flood_outline) again, this is the officially allowed data source. The layer is posted to MASDAP in 2013, and it is a flood Probabilistic Flood Frequency up to the year 2012. So, @ZINDI disqualifies our team for the violation of the "rule that does not permit the use of 2019 data" that we haven't used! Haha, this is really ridiculous.

From the discussion, I know that a lot of teams used floor frequency data, and this is even more frustrating! Please feel free to comment below, it would be nice to hear your opinion on this topic. Remember, next time it can be you :)

Discussion 6 answers

This is really strange Zindi. I think they made this decision under the pressure of some teams who wanted to disqualify your team which is absolutely disgusting. It probably would be logical if they have a really strong proofs against your team and not contradicting to their own rules. I understand your concerns as you spend a lot of your time solving this hack but finally it was a waste of time. So Zindi we are waiting a proper explanation from you else you will lose a lot of participants on your platform.

30 Jun 2020, 09:59
Upvotes 0

Thank you! I really appreciate your comment above. And yes, I suspected that we'll be disqualified, even though, we haven't violated any rules. But the explanation is just ridiculous, this is like a shotgun blast to the face :)

Absolutely agree with you yurii, the rules says that you can use any layer from masdap as long as it is previous to 2019 and you obviously respected the conditions.

Zindi contradicted their own rules and i think anyone of us will think twice before investing time in any coming competitions (who knows whos next on the list !).

Great job yurii and your team and best of luck for your coming challenges.

30 Jun 2020, 11:42
Upvotes 0

Thank you for your support! Congrats on taking the 5th place in this competition!

@ZINDI could you please comment?

30 Jun 2020, 15:12
Upvotes 0

That’s really disappointing to hear, but I appreciate the clarification. I understand the reasoning behind excluding 2019 data, especially if it compromises the integrity or future usability of the model for real-world applications. Still, it's a tough lesson—definitely something we’ll keep in mind more carefully for future competitions. Thanks for the feedback. sprunki phase 5

16 Sep 2025, 02:07
Upvotes 0