Originally Posted by
somerandomnamee
The real problems are : 1) Some people have messed up elos dues to bugs that were never fixed 2) Sometime the game pops even when no balanced match is possible ( mostly happens with strong premades )
1) This is not true. Or at least there's no evidence of it, bugs are always possible (I've personally investigated a number of claims of this, and always found the ELO to be performing accurately). Some players were impacted by a bug with the ELO system a ways back. That system was tossed (and the scores removed from all players) and the current ELO system was used since then (with the current ELO system's scores). Since both ELO systems were operating simultaneously and independently the entire time, there were no data issues. 2) This isn't a problem for ELO. Let's say the 5 best PvPers in RIFT form a group together and queue up. Now the matchmaker is probably not gonna be able to match them 100% appropriately, but the ELO system will know that and compensate. This presents a challenge for the Matchmaker, but not the ranking system itself. The question becomes how much should the Matchmaker weight match times compared with potentially lopsided matches. Is it more fun to be waiting longer for matches and have them be more even, or is it better to play more even though there will be occasional mismatches. We try to strike a balance here, because drifting too far in either direction isn't as much fun. It gets even more complicated because RIFT PvP doesn't lend itself particularly well to score parity. It's much more likely that the scores will be more lopsided than the skill levels of the teams indicate. Small differences in performance in a given match tend to produce outsized results.
Let's block ads! (Why?)
Ocho: Why use Elo system in warfronts
Aucun commentaire:
Enregistrer un commentaire