mido22, I didn't blame you to be a cheater; I can only judge from what I can see, and as I pointed out, that really has a bad taste.
What happened to my 253-Bytes AC Python solution I explained before and won't repeat that.
If I understand correctly, you spent some time in solving that problem before it was published and then, after it was published, submitted the shortest C++ solution among your friends (340 Bytes). After I worked out and submitted a 319 Bytes PHP solution it took less than two hours(!) and you submitted a 287 Bytes C++ solution. No one knows, if you get to know others solutions and no one knows if you know the expected output which consists of only 7 lines in case of problem NUMWORD. That really has a bad taste and I do not feel as to take part in a fair competition. And that's what I want. I will not be the one who will tell you to disqualify your submission - that had a bad taste as well because of my actual second place in ranking.
There are two things that could be done:
1. In case you are related or one of the problemsetters and decide to submit a solution though, you should submit only 1 solution before any other solutions are submitted (that's what kokosek said). So no one can blame you that you had any advantage from knowing other's solutions.
2. As I stated in my latest comment below the problem, problemsetters should increase the number of datasets so that even submitting the expected output (without a real algo) would lead to a "long" solution. Number of cases could be changed e.g. to 200 and then a rejudge could be done. All working AC solutions can easily be modified to solve 200 testcases instead of 7.