| |
BGonline.org Forums
65R-21$-53P-42 (Casper variant)
Posted By: Nack Ballard In Response To: 65R-21$-55P-42 (Casper van der Tak)
Date: Sunday, 10 July 2011, at 6:17 p.m.
To test the theory of applying the .100 swing across similar position pairs (i.e., given that 21$-53P-42 favors 5pt by .074, one can predict 65R-21$-53P-42 favors 4pt, by very roughly .026), I just finished a Snowie truncated rollout (1296 trials), which IMO is more reliable than a 2592-trial full rollout (double the length to 5184 and the reliability is about the same). Result: the 4pt won by .017. Noting that all but one of the elements of the comparison were based on truncated rollouts, there is pretty fair consistency.
148
65R-21$-53P-42 164
The truth rates to be somewhere between your margin and mine (if not, then there is a significant trunc bias). That is, there is little point to my extending the stable 1k trunc, but in case you extend the (high-variance) 2k full, the 4pt is more likely to gain than lose vs the 5pt compared to where it (apparently) stands now in your rollout.
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.