people are exactly saying/thinking the same now about the nx. And they will get dissapointed again. Never ending story.

so funny 

Might as well have said below one billion, the wii u will never get there all the same.

Is the word ''Less'' so difficult to comprehend? 

The fact that i need to deal with this shit just because i said: ''Less'' then 35 million life-time is quite disturbing. 
People are so damn emotional over a fucking lame prediction that was intended to be lame, wtf? SMH

now that we know the sales its clear that it will stay under 20Million lifetime. bump this in 5 years.

Either emotional or they used to sleep during maths lessons.   

I can't blame them for sleeping during classes though, we all when we was young loved playing games late at night completely uintentional and sometimes intentional because 5 hours feels like 1 hour

I remember this one. I was always one of those who remained skeptical of the Wii U, but I could never have imagined it would do as badly as it has, probably not even close.
I believe my first semi-realistic estimation with figures was 18-24 million, made some time in late 2013 or early 2014.

Here it is; http://gamrconnect.vgchartz.com/thread.php?id=171400&page=3

Late 2013, not nearly as impressive as the OP here though, even I had much higher expectations even if I was skeptical after learning more about the concept at E3 2012 (which is when I first mentioned properly that I thought the whole machine was a potentially terrible idea).

Kudos to the OP.

Or maybe we just didn't sleep during the chapter about statistics, where you realize that saying "below 1 billion" and "below 15 million" don't mean the same thing. Actually, you probably could sleep during statistics and still realize that.

But okay then.

Every console and game ever will sell less than infinity.

Bam, 100% accurate prediction right there, amazing, right?

Even in statistics he made a correct statement, basically he used a percentile. Maybe you could argue that saying it will sell less than 1 billion has a lower information content, but it also has higher probability of being true.   
But basically he did an even different thing: he subtly and justly trolled those that misunderstood predictions of selling "less than X" treating them like prediction of selling "more than Y". They aren't the same, they are similar but symmetrical, predicting "less than X" can have more probability of being true if X is large enough, while "more than Y" the opposite, if Y is small enough.
"Less than infinity" is a 100% accurate prediction also if the integral of probability reaches 1 asymptotically, but if a thing happens in a finite time interval, or, in this case, sales, it has most probably a finite, but not precise and well known interval until it's over, but it's bounded above by another function, in this case production, "less than infinity" is overshooting, you just need enough informations on the maximum number of units that will be produced and then you can make a 100% accurate prediction without exaggerating in prudence. If you can't know that precise number yet, you can choose as a quite safe bet lifetime sales of another device of the same kind that was more successful.
OTOH making a precise prediction of selling "more than Y" is very easy, you just have to put Y=number of units already sold.

In both cases who makes such predictions will be up to 100% accurate (easier to achieve with "more than Y" predictions), and in both cases many people will misunderstand and attack them, although for different reasons, but always for the wrong ones, having undestood a different thing from what they meant: if I predict that Wii U will sell "more than 12.9M", some Ninty fans not very good at maths (or just not very good at semantics) will attack me accusing me of being a pessimistic hater, if I predict it will sell "less than 100M", Ninty haters with the same flaws at maths or semantics will accuse me of being an overoptimistic Ninty fanboy.