Quote:
Originally Posted by YouAreAwesome
Is this what you mean? Trying to group them further?
1. 3s3dKs
2. 7s7d6s
3. QsQd7s
4. 2d3sAs
5. 2s4d8c
6. 2s5dQc
2s6dQc the most similar in the list
7. 2d9sKs
2sQsKd
pretty different but still more similar than others
8. 3s5d8c
9. 3sTdJc
10. 3sJsAd
11. 4s9dTc
5s6dTc
can these be grouped?
12. 4s6sJd
4sTsJd Kind of close.
13. 4dTsKs
5s9sKd
14. 5d7s9s
15. 7d8sTs
7d8sJs
16. 7sQsAs
17. 7sKsAd
18.5s6dAc
8s9dAc
I did this super quick just to check if this is what you mean. I'm sure this could be easily improved.
Yea, this is one method of what I was talking about. Just highlighting similar boards like this is def one way to go about prioritizing boards, but it seemed one of the less effective methods (perhaps it isn't?). If I could devise a way to first categorize these boards into a few groups, I could then select the one's (out of each group) which I feel would be most strategically different according to a solver. I will for sure take your findings into account, and likely remove for example one of the 2s5dQc & 2s6dQc boards you highlighted.
As mentioned in the OP, my original idea was if I could define high, medium and low boards, I could then categorize using this method & remove flops accordingly. Without thinking too hard, other methods of categorization could work such as paired boards, A high boards, BW boards etc. but they seem less clean almost.
Honestly, perhaps this approach is OTT & unnecessary, reasonably new to well devised solver study, thus the post. It's also probably not a terrible idea to just put in the extra hours & use all 25 flops suggested in the subset, it just seems this process could be streamlined. Cheers