You are viewing a single comment's thread from:
RE: STEEMMONSTER THEORY: You should face ANYONE in your league
k early evidence is that perhaps I was too committed to my proposition because my first data was looking at just a really unfortunate set of circumstances... data is still showing the overall principle I'm talking about is likely proving true but it's not as widespread an issue as i had thought. More people playing will make it worse... but it's not as bad as i expected. I'll admit that.
this is taking a bit too much time to analyse and i have my actual real life job to do. haha
But here's what the data says.
If you want upward mobility in elo then you want
- Not a lot of people playing
- Mostly people playing higher than you... if someone is playing close to your score then stop playing for a while and wait for them to stop.
- Less people playing means greater differences in scores which means greater potential for score mobility.
- You want to be in a league like diamond or gold where the scores are more spread out and there's not as many people.
Basically we have an incentive now for less people to play ... specially in our own league and specially when we're at the top of the league.
Here are some data sets I was gonna go through @tufkat game me this one
Anyway the gold player had a bigger winning percentage and actually did have 3 big elo divergent scores against the same exact player which kinda throws off the stats in this case. But even still the elo divergent average was about 25% higher for tufkat even with those 3 anomalies. About two times greater difference without those 3 anomalies aka 50% 100v200. Take out the top 3 differences on tufkat and it barely moves the dial so they're not really anomalies.
Again willing to listen and see different data. I'm just thinking out loud.