New benchmark : Occam

Want to calculate something or share your results?
Locked
User avatar
abiu
Posts: 26
Joined: Mon Dec 01, 2008 3:01 am
Location: Ningbo, China
Contact:

New benchmark : Occam

Post by abiu »

Occam = 3bv/s * IOE
= 3bv * 3bv / Time /Total clicks

the vid that has more Occam means more perfect!!!
EWQMinesweeper
Posts: 419
Joined: Sun Nov 30, 2008 11:50 pm

Re: New benchmark : Occam

Post by EWQMinesweeper »

i strongly disagree that a lower ioe necessarily must mean that a game was played badly!

arbiter allows to distinguish ioe into throughput and correctness, correctness being the ratio of clicks that weren't wasted and throughput being the possible ioe if correctness was 100% (so to say throughput can be considered the actual ioe of the solve)

the way sweepers like eg ian, dion, dan or myself use double clicks, namely if the patterns allow it just throwing in some fast double clicks without waiting to see what is under the square we just opened, obviously causes some wasted clicks (in my case a quite a lot), but at the same time this technique brings a lot of speed. normal ioe decreases becauses of this clicking, but throughput is not affected by those clicks and allows a much better comparison of how well those games with this technique and eg kamil's are played.

i'd rather go for 2 different indices for perfection:

humanZiNi/s for speed
humanZiNithroughput (HZNT in new arbiter) for efficiency of the solve
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“
Locked