| |
BGonline.org Forums
What's your favorite formula?
Posted By: Taper_Mike In Response To: What's your favorite formula? (Stick)
Date: Thursday, 2 August 2012, at 8:54 a.m.
–
Depending on the length of the race you are able to double using the Keith Count when the difference is 5 or 6 or even 7. Now, this particular race isn't long so I would expect the Keith Count to remain close to the constant you quote. I do have my own length of race break downs for when it would be correct to cube exceeding by 5 or 6 or 7 pips. I did not know that. I called this conservative, because at 6, the standard Keith Count is not even close to recommending a double.
When using the epc you only need to reference the amount of rolls the player doubling (winning) has to go. You don't take an average of the two as you sort of implied whether you meant to or not. I just don't want that to confuse anyone.
Your example tonight is the first time I have seen an explicit EPC formula. I see that it is different from Trice62, or Nack57 for that matter. I am very interested in this, because I have always been unhappy with trying to use EPC with Trice. I could never figure out whether to calculate the doubling window using the leader’s EPC or raw pip count as a base. When I use EPC, the thresholds seem too hard to reach. In practice, I have been using estimates of EPC to get the difference between the two players, but then using the raw pip count as the base from which to calculate Trice62. I still don’t know the proper method.
As far as averaging EPCs goes, I must have miscommunicated. I wasn’t even thinking about average. It is just that the exact EPCs (stated as the expected number of rolls remaining) are both close to 8.5 rolls. White is at 8.7; Black at 8.4. So, I was loosely comparing the position to an n-roll position, except in this case n = 8.5. That’s a double, and a take.
Mike
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.