"Want better data? We found a similar food" returns very different data

Lector

I was searching for "Butterball ground turkey 93 7" and saw the prompt for better data. The problem is that for the same portion size (1oz), the better data has,

• 60% more calories,
• 40% more protein
• 65% more fat

Here's a screencast:

How can we trust this better data? If it's worth sacrificing the macros counts for the micros? Or should we always double check the macro values? Can the software just not suggest "better data" if the discrepancies are beyond some thresholds?

Bruno

I never use CRDB. Always NCCDB and USDA. Sometimes I "reverse engineer" foods using the provided values on the label and the giving ingredients using a custom recipe. Works most of the time very well.

Karen_Cronometer

Hi there,

In this case, the manufacturer has provided nutrition values for the raw turkey, whereas NCCDB always gives nutrition values for cooked meats.

This is problematic if you have entered in your serving size raw, though we still want to offer you the option of adding a food to your diary that has more complete data.