Around twenty years ago, two methods relevant for systematic innovation and improvement - TRIZ and Value Analysis - have been merged in a specific way, which was then incorporated in software packages about innovation and became part of certified TRIZ education as well. An intended key purpose of this nowadays established method of Value Analysis in TRIZ is to identify system components of low ideality - or value, respectively -, i.e. parts of the system, that do not give a satisfactory functional contribution in relation to their cost. Hence, Value Analysis should point out the sweet spots for subsequent improvement, innovation or even patent circumvention activities. Unfortunately, the commonly used standard approach for Value Analyis in TRIZ, which is based on a function ranking algorithm, leads to results, that are inconsistent and not trustworthy in general. This work illustrates these shortcomings and explains their origins. Derived from key requirements necessary for a meaningful concept, VA++, a new advanced approach for Value Analysis in TRIZ, is presented and validated.