[reggieband]

I used to do well at these pattern matching tests with consistent scores in the range 125 - 135. This one I hit 118 but I ran out of time around question 32. I got stun locked on a couple of the earlier easier questions. I wonder if this is due to mental decline as I get a bit older (I'm in my early 40s now).

I took it again with the benefit of knowing most of the earlier question patterns and got 133 on the second attempt. Pretty sure it wouldn't be too hard to bump that higher if I cared. In that sense, these tests are heavily biased towards people who take these tests. Once you know the tricks these tests are built on you do a second-order pattern match (which trick is the question using) which helps you find the answer. I would guess you could grind these tests leet-code style to get arbitrarily high scores.

People here report getting bored by these tests and I don't feel that at all. I actually get totally stress-focused and the time flies by almost too fast. Seeing the pattern is a dopamine hit that my brain craves.

[password54321]

I mean you did the same test twice and gave yourself extra time. Of course you are going to perform better. This doesn't mean you will perform better for a totally new test.

[reggieband]

I argue that patterns learned on individual tests would apply to new tests. There are a set of patterns which have already been called out here. Translation, rotations, exclusions, logical and/xor on line segments and dots, etc. IME, most IQ tests sample from this set of patterns. Once you are familiar with that set you aren't really doing a first-order matching - instead you are going through your list of expected patterns and seeing if they apply to the particular question. This means you can breeze through the early questions drawn from that set using your memory and then save your cognitive time for the later harder patterns.