R-Precision of documents with same relevance degree

Let’s assume I have a collection of 10 documents,so |D| = 10 and a collection of relevant documents R , where |R| = 3 & R = {d1,d3,d8}. Then I make a query,q, to a random retrieval system and it returns this. Ans(S1,q) = < d1,d3,{d2,d5,d8},d4} Then d2,d5,d8 are in the same position.

Pos.       | Relevant d1         | 1 d3         | 1 d2,d5,d8   | 1 d4         | 0 

So R-Precision in the first 3 positions is R-Precision = 1, right? Or do I count them as:

Pos.       | Relevant d1         | 1 d3         | 1 d2         | 1 d5         | 0 d8         | 0 d5         | 0