Ok, flexiforce recommends a standard inverting amplifier design. With such an amplifier you have input and output voltages of several volts, but the potential directly at the inverting input is close to that at the non-inverting input, which is signal ground. This is no surprise of course, because a modern OPamp shows a very high open loop gain, so it needs only a very small potential change between the both inputs to produce a big output voltage.

But that means at the same time, that the inverting input is highly sensitive to noise and EMI: A very small portion of interference is enough to produce a big result at the output. For that reason, the distance between this pin and the components in the feedback loop of OPamp must be as short as ever possible, usually a few millimeters. But such a flexiforce sensor presents connections which are way too long, even if an additional cable is omitted.

Another aspect is stray capacitance. Extending the connections at the inverting input introduces stray capacitance from the inverting input to ground. This results in an increase of gain at high frequencies, so much that the amplifier can immediately start to oscillate. Also, the stray capacitance results in an phase shift in the feedback loop (phase lag) which makes the design instable. An absolute must is to provide an additional phase lead by connecting a suited cap across the inverting input and output of OPamp.

Such an "open" inverting amplifier isn't protected against ESD at its inverting input at all as well. Whatever hits the flexiforce sensor is directly injected into the inverting input. Such a crap might work on the laboratory bench, but will miserably fail in any more rough environment.

Kai