Sending Messages faster than with Speed of Light using Quantum Entaglement?

A warning in the beginning: I am not a Quantum Physicist, I am merely an interested observer. Thus, this text is a question rather than a statement, and I would be happy to receive some critical comments. But let's dive in:

Quantum particles can be entangled, meaning, that pairs of them can be seen as one physical system. As soon as the quantum state collapsed for one particle, e.g. due to measurement, the other particle also "collapses" to a specific quantum state. This usually happens much faster than light would need to overcome the distance between the particles – as far as I understand – nearly immediately after the first measurement. Even over large distances.

An electron can take spins of 1/2 or -1/2. If two entagled electrons are produced, one is measured and the value taken is -1/2 we know that the state of the other electron is 1/2. As mentioned, the surprising fact is, that this process can happen over large distances and the corresponding collapse of the state happens immediately, thus "faster than the speed of light". The common wisdom is, though, that this does not allow communication faster than the speed of light. One reason is, that we cannot foresee which value the first particle will take. Both collapse at the same time and with opposite, but not predictable values.

Now to my thought experiment: Assume, we can produce more than one entangled pairs of particles. Additionally, we define (for encoding purposes) that a spin of -1/2 corresponds with 0 and +1/2 with 1 (in the case of electrons).  Now we produce a number of entagled particles, wait until they are in a large distance, say one light year apart. 

The goal is to transmit the number 42 to the other side. 42 is in binary form 101010. We cannot define the outcome of the measurement of the next particle, but we know, that whatever the outcome might be, the value on the other side is deterministic the opposite. Then we collapse the series of particles on one side using the following strategy: 

• If the currently measured value corresponds with the desired number that should be sent, we measure the next particle within 1 second (or some other arbitrary, but previously defined duration). 
• If the currently measured value doe not correspond with the next bit to be sent, we wait 2 seconds. Thus, the receiver can judge by the interval of the collapses whether the current value should be considered or ignored. In my example:

current value | duration after measure | value recorded
  0:                2s                     ignored

  0:                2s                     ignored

  1:                1s                     --> 1

  1:                2s                     ignored

  0:                1s                     --> 0

  1:                1s                     --> 1

  0:                1s                     --> 0

  0:                2s                     ignored

  1:                1s                     --> 1

  0:                3s                     --> 0 
  1:                                       eof
  ...

And the binary number 1-0-1-0-1-0 (3 second is supposed to be a stop-signal indicating the end of the message) can be reconstructed.The whole duration of the sending process, in this simple example took 15 seconds. (Of course under the assumption of very slow intervals between measurements.) If the distance between the two particles is larger than 15 light seconds plus the time needed for the collapses to happen after the measurements, the message transmission would have been faster than the speed of light. The actual encoding of the message is done using the intervals of the measuring process with the help of the fact, that the quantum systems produce some sort of a stream of random numbers.

Now my question to professional physicists: Assuming that something is wrong with my idea: Why would this thought experiment not work? Which wrong assumption am I making.