| |
BGonline.org Forums
Live variance reduction tournament
Posted By: Timothy Chow In Response To: Live variance reduction tournament (Johan Segers)
Date: Tuesday, 6 September 2011, at 11:33 p.m.
That does make sense, which is why I advocate using MWC difference rather than error rate.
That is, at the end of a match, we consult the bot to figure out the total MWC I gave up over the course of the match, and we do the same for my opponent. The difference between these totals is what I call the "MWC difference." If I played better, then the MWC difference is my score. If I played worse, then my score is the negative of the MWC difference.
For example, say I gave up 8% MWC over the course of the match and my opponent gave up 13%. Then I score 5% and my opponent scores –5%.
The advantage of this method is that I am motivated to outplay my opponent, not just play as the bot tells me I should. If I think my opponent will drop my double even though it's technically a ND/T, then in an error-rate tournament I will be motivated to hold the cube, but in and MWC-difference tournament, I will be motivated to double in order to elicit the incorrect drop because I'll gain overall as the result of my opponent's error.
I don't know of anyone who has tried running such a tournament, however. Seems to me to be better than either a variance-reduction tournament or an error-rate tournament, but Casper van der Tak may be the only person I've convinced.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.