For Loop Ranker: a complete ranker that runs as a twiddler
The For Loop Ranker has outputs per query and per result plus web-specific settings. RESONEO reads it as a full ranker working inside the twiddler layer. What it scores is not known.
01What is known
Three schema objects belong to the For Loop Ranker, or FLR. FLROutputQuerySignals and FLROutputResultSignals, in the twiddler.for_loop namespace, hold what it produces for the query and for each result. QueryParamSignals, in ngs.web.for_loop, holds settings for web search; what the prefix ngs stands for is not known.
A for loop is the programming construct that goes through a list one item at a time. RESONEO takes the name to mean that the FLR scores the results in turn, and stresses that this is a full ranker running as a twiddler, not a simple adjustment. Like every twiddler, it only recommends; the framework in Superroot, the layer that fans a query out to Google's indexes and merges what comes back, decides.
02What is not known
The name describes how the ranker runs, not what it judges. Which features it uses, which model sits behind it, for which queries it fires and how far it can move a result are all unknown.
The finding that matters is structural: if RESONEO is right, the twiddler layer holds not only small rules but at least one ranker of its own. On our map the FLR belongs to the twiddler layer and to no E-E-A-T system.