There's a couple reasons for this. The first one to look for is when they were issued. Frequently the .50s are released after the even dollars, and the 1.00s are released after the 2.50s and so on. That's the biggest one, particularly on monthlies and LEAPs.
You can often find matching positions that are individual legs of spreads / straddles / strangles.
When you've eliminated those two possibilities, look for whales speculating. You can often see these in the underlying volume (last Thursday there was a speculative long put position on SQ, and I believe in the ask...so long; conversely yesterday on NKE, there was a speculative short strangle on the Jan $70s).
Once you've got all this figured out, you can make some inferences about the market. Specifically, in the SQ case above the position was immediately hedged 1-to-1, and the volume bar is obvious. I didn't catch any volume on the NKE hedge, so my read on this is someone with a lot of money believes that NKE will cross $70 before 1/19/18.
The value in figuring all this out is finding where speculative longs are on stocks. That means market makers are short--and thus will be delta hedging. And you chose a great day to ask this because the monthly expiry dates show this most strongly! With this information, you can infer sometimes how nearly 4-5% of average daily volume will change hands based on the price.
And, I know there's a bunch of naysayers who are going to say nay. So, I'll end this with two points. One, I said
infer, not
know. Two, I made this post on 10/26 when SQ was trading at $34, and had never traded above $35. It's gone within .23 of $50 during after hours at its peak:
One more edit: As far as practical application of this knowledge, 35 and 37 will show this pattern on SQ in some permutation before mid December (look at the chains!) and 40 will show it very strongly by mid January.
How well you think I've done on that stock?

(also, those predictions are stale at this point--so, don't read too much into the January predictions)