| |
BGonline.org Forums
Edge
Posted By: rambiz In Response To: Jacobs/Trice Fish book - a proposed computer experiment (Bob Koca)
Date: Tuesday, 30 December 2014, at 7:04 p.m.
Bob Koca: "I would call A being 60% and B thus having 40% a 20% edge but you seem to be calling it a 10% edge."
Yes, indeed I thought of "edge" as whatever a player got over the 50% mark.
Bob Koca: "Using your definition of edge note that A beating B 75% and B beating C 75% predicts that it is impossible for C to ever beat A."
True. Moreover, another example would be A beats B 80% of the time. In this case, my model predicts that there is no way B could be more than a 70% favorite over any C. Apparently this model has some limitations. I do agree.
Let's do some comparison anyways:
Example 1: A is 80% favorite against B and B is 60% favorite against C.
My model: A is a 50%+(80%-50%)+(60%-50%)=90% favorite against C.
Your model: A=8, B=2,C=4/3 ----> A/(A+C)=85.7%
Example 2: A has 75% over B and B has 75% over C.
My model: A beats C 100% of the times.
Your model: A=3, B=1, C=1/3 ---->A/(A+C)=90%
Example 3: A has 55% over B and B has 65% over C.
My model: A has 70% over C.
Your model: A=55, B=45, C=hmm... 24.23 -----> A/(A+C)=69.4%
I'd very much appreciate, if you could please elaborate more on the underlying model, you are deploying.
Should I ever get the necessary skills ,I'd like to find out how much of a favorite XG is against a randomly checker-shuffling opponent at DMP.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.