So here's my take on it, expanding on my previous comment.
At any given time, you look at the roulette board. The chance of seeing those given 7 numbers, on the board, at that time, are 1 in a few billion.
If I were to stand at the entrance of the casino, and bet that 4,1,3,5,2,4,1 was going to show up in the casino that day, any person would be stupid not to bet against me. That is the same, whether those numbers are 4,1,3,... or 29,29,29,... In both cases, the odds would be very much against me.
Now what Aslan and the OP are noticing is something different. They already have a preclassified notion of sequences which would be considered "remarkable" to the human mind. These would probably consist of 2,2,2... 3,3,3..., 1,2,3,4,... 2,4,6,8,... and a slew of other preclassified "unbelievable" sequences.
Now they are correct in saying that the chance of seeing an "unbelievable" sequence at all is extremely small (since it's a small subset of all possible sequences) and at any given time, we have a much much higher chance of glancing up and seeing one of the "believable" sequences. What others' (read: London's and my) point is, is that your preclassified notion of what should go in the "unbelievable" set is completely arbitrary.
Now a separate issue is that with the hundreds of thousands of roulette spins per day, or week, or whatever, your odds of seeing one of your "unbelievable" sequences is much less than 1 in a few billion. But as long as you don't arbitrarily put too many sequences in the "unbelievable" category, it can still be considered remarkable.