search

Eve-ning / vsrgtools

R Library to mangle with vsrg files
MIT License
2 stars 0 forks source link

Successive Jacks are not considered in calculation #10

Open Eve-ning opened 5 years ago

Eve-ning commented 5 years ago

Repetitive but Slow patterns are inflating the difficulty calculation too much as depicted by

> runComparisons()
[1] "TRUE | 8.1 | YST - The Lost Dedic  >  ZUN remix by 44teru-"   
[2] "TRUE | 4.05 | Feryquitous - Vend (  >  ZUN remix by 44teru-"  
[3] "FALSE | -0.35 | BanYa & Warak - Beet  >  penoreri - Everlasti"
[4] "TRUE | 1.75 | Wotamin - Gigantic O  >  Wotamin - Gigantic O"  
[5] "FALSE | -5.15 | xi - ANiMA (Kuo Kyok  >  Umeboshi Chazuke - P"
[6] "TRUE | 0.6 | Igorrr & Ruby My Dea  >  Kaneko Chiharu - Kai"   
[7] "TRUE | 4.3 | UNDEAD CORPORATION -  >  Camellia - Light It "   
[8] "TRUE | 0.95 | Camellia - Bassline   >  ginkiha - Anemoi ([ "  
[9] "FALSE | -2 | Betwixt & Between -   >  typeMARS - Triumph &"

Probably need to use another model to either squash the value down or run non-repetitive values up. Latter is preferred.

Eve-ning commented 4 years ago

With #25, this will be important as it does seem like minijacks are underrated

Eve-ning commented 4 years ago

Personally cannot confirm this for now, model.jack seems to be doing okay with this, will KIV