I saved 5 identical documents to my Azure Search Index, with a weight of 1 applied to a name field (below).
var fieldWeights = new Dictionary<string, double>
{
{"name", 1},
};
As all documents saved are identical i was expecting that all documents would be returned with the same search score. From the image below you can see that the first two are the same but the last three are a bit lower.
You might find the following article useful: How full-text search works in Azure Search, especially the section about Scoring in a distributed index. It explains that because Azure Search indexes are sharded to facilitate efficient scaling operations, relevance score of documents that are mapped to different shards could differ slightly as term statistics are computed at the shard level. In general, we don't recommend developing any programmatic dependency on the value of the relevance score as it is not stable and consistent for different reasons. Accurate relative order of documents in the results set is what we're optimizing for.
