Walter Lewin vids-why ± 0.5 cm uncertainty why not ± 0.1? here is the  link to walter lewin video lecture please jump to  13:13
https://www.youtube.com/watch?v=q9IWoQ199_o 
8.01x - Lect 2 - 1D Kinematics - Speed, Velocity, Acceleration
i thought that all meter ruler/ meter stick  use ± 0.1 cm as uncertainty .how did he get ± 0.5 cm? 
 A: The answer lies in the lecture itself. At 12:49 Lewin himself gives; "I cannot do that any better, really, than maybe even half a centimetre... I cannot guarantee it better than half a centimetre." What he is talking about is an error analysis concept of stating a best estimate ($148.5cm$ in this case) followed by a range in which we are absolutely certain the measured quantity lies ($\pm 0.5cm$ in this case). In other words, the wires are certainly not more than $149cm$ apart and certainly no less than $148cm$ apart.
You will learn as you go on in error analysis that the range to fix onto the end of a measured quantity is not so straightforward as taking the smallest possible uncertainty of the instrument used ($\pm0.1cm$ in this case). There are many cases in which you will have to think and change the uncertainty of your measurement to suit certain circumstances. For instance, there is a special case where if the leading digit in the uncertainty is a $1$, then keeping two significant figures instead of one in your uncertainty may be better (and often is). Let's say you round something from $\pm0.14$ to $\pm0.1$, this is a significant and substantial proportionate reduction because the $0.04$ is a big part of $0.14$. 
In this case, I think Professor Lewin is using a half of a centimetre as a safe bet. Sure, you could very plausibly estimate a better uncertainty like a third of a centimetre but for the demonstration he is giving, it is not a big deal.
