People have always been giving more points to their neighbors on average, but not always. In case you don't know, each country gives more than one "vote", ie they assign 12 points to one country, 10 to another, 9 to yet another etc. Usually the 12 go to the sister country (ie US-Canada type relationships), but the rest would be spend in a reasonably fair way.

Please note that while from 1998 to 2008 (the peak years of this contest popularity-wise imo) almost 100% of the vote was coming from the audience (via paid SMS and calls). From 2009 to 2012 audience only accounted for 50% of the votes, the rest was picked by the Eurovision company. From 2012 on, the audience had no real effect on the votes. This is also when things started getting seriously political (Conchita, Ukraine, Russia etc).

It's interesting because I just noticed that the Eurovision followed the same path as most social media during the same timeframe. Trying to make you think you have a choice while you don't, trying to fake popularity as a means of enforcing some agenda etc.

As I understand it currently 50% of the votes come from the public, 50% from judges. Is this not correct ?

That is correct.

There are some small exceptions, such as in San Marino, where their small population (~32,000) and sharing of phone infrastructure with Italy makes it difficult to calculate a valid independent televote. And televotes can sometimes be ruled invalid by the auditor (PwC) if their auditing suggests it has been manipulated.

https://eurovoix.com/2017/03/08/san-marino-ebu-rejects-broad...

And sometimes it's the jury votes that are ruled invalid. This happened in 2015 when both FYR Macedonia and Montenegro were 100% televote:

https://eurovision.tv/story/full-voting-results-revealed-by-...