yes, it's having the most false results even, this is because they can't predict correct search engines based on market share of just 1-2% if you check the keyword "money" you will see how few searches that keyword has, then check that keyword on google adwords keyword tool and see the high volume, then take a good other keyword such as cars and check their ratio on google trends, you will see 1:1.28 ratio, however the search volume that is displayed on WT is only 2930/day, which means 87,900 searches for money/month. i doubt that's the actually number on google when they state the keyword has high competition and is having high search volumes. now if you check cars in WT there are 39962 searches / day. there is no 1:1.28 ratio here. but more like 1:13 ratio. WT is far far away from accurate, and even if you keep the correct ratio for 2,3,4 keywords, their numbers could be tribally wrong.
I cannot say Wordtracker is not accurrate because it is not based solely on google alone but in combination with other search engines. In most cases if you compare the results from wordtracker with other keyword tools it looks very similar. Just don't rely on one keyword tool since there are so many of them, both paid and free.
It's true - WT and Google Keyword Tool are on two completely different planets. One thing I know for sure - when Google Keyword Tool says that something has a "low" search volume (the short green bar), that could mean one search per day or less...I've looked into this with local searches and saw what people in my town were searching for.
I would say if you were going to use two, use WT and Google Keyword Tool in conjunction together. While Google keyword tool is very basic, I like its graphical representation of volume and competition.
Every tool has it's own algorithm, definitely they have different results. And besides, wordtracker does not based their result on google.