Posted by obituaryconc on 2/26/2014 2:04:00 PM (view original):
Just wondering how you calculate RWOPR.....more for my current project that I am working on for programming. I am attempting to create a version of Hoops Dynasty that is different but has the same overall idea. Any input would be great and anyone interested in helping out with my project would be appreciated as well.
Disclaimer: I am not using any of Hoops Dynasty lingo or anything of that nature. Just working on a similar idea since I have time and just want something to do. I am not going to use any of Hoops Dynasty information or trademarks or copyright.
Pasted from my previous (surprised by how far back) posts:
Ok, I ran something I'm calling WOPR on everyone as well as the 2 season conference wins average thingy... I don't remember what WOPR was supposed to stand for (weighted overall something or other). It takes the average OVR rating of JRs multiplied by 3 plus the average of SOs multiplied by 2 plus the average of FRs unmodified. This allows for an allowance for IQ and bases relative future strength on the rosters of the teams instead of past performance. However, it has numerous flaws of its own. It doesn't account for incoming freshmen. It includes walk-ons (it doesn't have to, I just didn't think to exclude them until after some had gotten through and I didn't feel like re-doing them all). It is based on OVR, which is not exactly a correlation with real performance. It fails to account for possible future transfers or jucos. But it probably does a better job than the last 2 conference wins average.
I have modified the WOPR formula so that it now takes into account the number of each class returning, rather than just providing an average value for that class. I feel this will more accurately reflect team strength. This resulted in a new order for the WOPR. I'll post the new RWOPR rankings, and conferences can decide to use the new, the old or something else entirely. Starting next season I'll only be providing the new RWOPR. Keep in mind that the RWOPR isan average of the OVR values of current classes, multiplied by a value based upon class (3 for current juniors, 2 for current sophomores). In the past this value was then divided by the number of players in the class, so basically one 700 rated junior counted as heavily as if a team had 4 700 rated juniors. This new formula should help correct that, and it gives weight to teams that return more players. However, it is still based on OVR (not the best way to value a player), obviously gives no value to incoming freshmen (since I have no idea what they will be) and does not take into account the value of the coach of the team. I suspect that if a conference gets together and physically ranks next season's team's on their own by looking at them and factoring in those things my formula does not that they would achieve a more accurate prediction of relative power, but in the absence of that I think this revised formula will do a better job than the current one.