0 percent to 100 percent. It's built on a standard distribution. so basically it means that if you have a 50% score youre better than half the teams out there. If its 75% it means youre better than 3 out of 4 teams, etc. It's all rough estimates based on PECOTA, ZiPS and Bill James with predicted playing time shares. Yes, it's a h2h league so I have one "swing" roster position (currently Broxton) which means if for whatever reason i need a hitter I can drop him for it. My hitting is strong enough i dont need them all playing 7 days a week, six is enough. Since it is h2h theres no ip limit, so having a lot of sp to rack up Ws and Ks is useful. ive also managed to do it with guys who are good in ERA and WHIP so my staff is dominant. as i said above the system is based off three of my leagues. I calculated all of the stats for their teams using an average of PECOTA, James, and ZiPS with an adjustment for an arbitrary expected playing time that I chose (ie not Lee with 440 PA). I then ranked them as if it was roto, so 12 points to the winner of a stat, 11 for second and so on. I ran a regression on all the leagues (36 teams) to get a stat to points estimate. In the system, 218 HRs gets you 6.5 points based on the regression from those 36 teams. So that puts you around 7th in an average league. I did this for all twelve stats. I then summed up your expected points in a nuetral league to get your teams expected score. I then took all of these teams and got their expected scores. I then found the standard deviation of these scores and the average so I can then find the distribution. Since there is a set number of Rs, W, etc in a league we can expect the distribution to be normal. That means we can then find out the percentile each team is in (roughly of course). It's very hard to move up once your good. For example one team has an expected point score of 92.6, which puts there percentile ranking at 98.3%, Another team has an expected point score of 103.8, a difference of a whopping 11.5 points (basically going from worst to first in a category), but there percentile rank went up just 1.5% to 99.8%. If you make that same 11 point jump from 69 to 80 points, your percentile ranking increases 24%!. So in otherwords I am saying the system isn't linear, and it shouldnt be. There should be a lot more teams that the system spits out as a 50/100 (average) than it does a 70/100, above average. Also you can use this to estimate where you rank in the league. There are 12 teams in each league, so that means youre probably the best if your score is above 92%. Yeah, I probably analyze this stuff way too much.