Millisecond Forums

I am struggling with implementing randomization in my experiment. 
The layout of the study is the following: 
My participants must be randomly assigned to start with one sequence of trials. After that, they still have to complete the sequence that hasn't been chosen to ensure counterbalancing in the long run. 
Currently, my script looks like this for the randomization:

<expt>
/ blocks =
if(values.selector == 0){
    return [1 = Practice_EmotionalStroop; 2 = EmotionalStroopEnglish; 3 = EmotionalStroopGerman; 4 = finish;]
}    
else{
    return [1 = Practice_EmotionalStroop;2 = EmotionalStroopGerman;3 = EmotionalStroopEnglish; 4 = finish;]
}
</expt>

I also changed something in the value section to ensure the randomization of the selector:

/selector = round(rand(0, 1))

At the moment my experiment always starts in English.

In general, all participants have to go through the practice trial first, then they should be randomly allocated to start the experiment in English or German first. Afterward, they still have to go through the trial/block that they haven't been allocated to in the first place.
I have no experience in programming and I chose an already-written script from the millisecond library and changed it to my needs. Except for this part, all is working fine.

Thanks for any help!