Bias play on smaller data sets...?

Below is a set of 600 numbers. It is a consecutive stream presented in the original order, i.e, they have not been doctored, just removed from a bigger 10,000 number data set.

So, who thinks the numbers are:-

1, bias or not?

2, playable, inconclusive, or not worth pursuing?

3, if you deem it playable, where would you place your chips?

19 15 31 26
31 19 7 15
21 12 22 13
9 11 21 24
18 35 7 31
34 18 34 0
26 30 20 1
21 23 9 14
26 6 21 12
34 33 2 18
15 3 14 29
7 1 24 28
9 11 33 10
4 26 21 15
24 32 29 1
23 20 13 35

19 15 31 26
31 19 7 15
21 12 22 13
9 11 21 24
18 35 7 31
34 18 34 0
26 30 20 1
21 23 9 14
26 6 21 12
34 33 2 18
15 3 14 29
7 1 24 28
9 11 33 10
4 26 21 15
24 32 29 1
23 20 13 35

You have numbers on the image it would take time to take data from it.
I tried copy from XL , it makes image and numbers, then on the bottom i copy from XL to notpad to remove XL codes then copy to forum and get clean numbers.

19	15	31	26
31	19	7	15
21	12	22	13
9	11	21	24

And here i copy form notepad then used performatted text from options and it look the best.

Yes, new forum since I last logged in. It was generated in the image form by the forum, not myself. That gave me the choice to have numbers displayed in a messy format, or clear and concise in the image produced. Maybe I chose the wrong format but neither were what I wanted so I plumped for presentation.