I agree Vaikiss. It won't be as accurate, but they should try something.
If they added a factor to include the Normal/High/Very High bracket you could weed out a lot of the influence of smurfs. Something has to be done or that will lead to the score being heavily dominated by smurf stackers, because the smurfs appear as no-stats players.
But if you valued the no-stats players based on the overall bracket it could still work. Games where you stuck several fake new players on your team would be in Normal and would not count for much. And once they got up into Very High the games would be relatively fair again and the value of smurf stacking would be minimized.
Of course when this was done in dotacash there were some people dedicated enough to establish smurf accounts where they played to lose lots of games with random people, then played a couple games with the account they wanted to boost...just a few so they would not lose their low ranking. Then they would go back and grind some more losses before boosting again. Some people would buy low rated accounts, some created them and sold them for boosting. You would not believe how far people will go to get fake stats. There were accounts with 500+ games and 90% loss. Think of all the time that took.
Since intentional losing carries a high risk of going to the LPQ they could make any game played with LPQ players count for no points.
You can't stop all forms of cheating but the system can be designed to minimize the impact of such shenanigans.
发表评论请登录
Since DBR has gone.. it would be nice to see each games fall in which bracket. Relevant and informative, right?