Last night I was working on my sequencer project and, looking at the sensors I recently bought from SparkFun, I had this nasty idea... so i wired up the MQ-3 alcohol sensor, hooked a 200k trimmer to it just to calibrate the range of the measures, and voilà! The sequencer global shuffle can now be controlled using your breath alcohol content...

After a little regulation on the trimmer I got values ranging from 30 (no alcohol sources near the sensor) to 860 (with the sensor over a glass of vodka).
In the next days I'll try to make some tests in order to get a relation between BAC tables (see the link to wikipedia page on the right) and the output value of the sensor.
I'll have to get some more vodka.