I made a plot from those 99 numbers differently:
90 1
400 1
1718 1
1900 1
2000 1
2100 2
2900 1
3000 2
3200 1
3300 1
3600 1
3800 3
3900 1
4500 1
4700 1
4800 1
5000 2
5100 1
5200 2
6000 2
6100 1
6500 1
7000 3
7100 1
7500 1
7832 1
8000 5
8300 1
8400 1
8700 2
9000 3
9100 2
9600 1
9700 2
9800 2
10000 3
10200 2
11000 8
12000 9
12100 1
12200 2
12400 1
12500 1
12600 1
12700 2
13000 12
13190 1
15000 2
99 total
That is not a uniform distrubution. But if we assume that 12 instances of 13000 is just the same winner listed by 12 people (that is an important high number, that is why 12 people listed that as the number of someone they know, or just of one person), and if iwe leave the following plot:
90 1
400 1
1718 1
1900 1
2000 1
2100 1
2900 1
3000 1
3200 1
3300 1
3600 1
3800 1
3900 1
4500 1
4700 1
4800 1
5000 1
5100 1
5200 1
6000 1
6100 1
6500 1
7000 1
7100 1
7500 1
7832 1
8000 1
8300 1
8400 1
8700 1
9000 1
9100 1
9600 1
9700 1
9800 1
10000 1
10200 1
11000 1
12000 1
12100 1
12200 1
12400 1
12500 1
12600 1
12700 1
13000 1
13190 1
15000 1
51 total
We see that that is an evenly distributed random variable (between 1 and 15000).
At the same time just one instance of 25000 fits into randomly distributed random variable between 15000 and 27000, with amount of values 1% to 3% of total (total is 52)
That is what I was trying to say before.