Glitches in Laxpower rankings?
Posted: Sun Mar 13, 2005 4:38 amLaxpower is a great resource for getting scores that are organized by team. Very convenient. Also it's kinda fun to see what a computer would do just with numbers ala the BCS. The actual ranking doesn't have a vote in our poll's like a few years ago.
I'm not sure how much poll voters are putting stock into the actual rankings, but there are a few things that I have noticed scanning through, looking for scores which may effect the "calculations" and possibly indirectly relative perceptions. I was wondering if other people noticed them too, or am I the only one that spends too much time trolling for scores around the nation. Someone'd probably have to email laxpower directly to see if it's a just display error or if these things are actually affecting the rankings. We're right around the time when the laxpower ratings would start to "be more representative" because more games have been played, so I thought I'd throw this out there.
1) sometimes there are double listings in the quality win factor... so if a team beat another ranked team, they get "bonus points". sometimes that bonus is displayed duplicate for the next game they play, regardless of who it is and regardless of if it is a win or loss. Not sure if they are giving extra bonus when they shouldn't be or if this is just a display error.
2) Laxpower used to rank Div A and Div B teams on one big list, but now it looks like the Division B Laxpower Rating is one group and Div A laxpower rating is another, calculated independently. So a Div A team with a 70ish rating in Div A is completely different than a Div B team with a 70ish rating in the Div B pool (which would probably be rated around 50 if it was one big list). However several Div A teams play Div. B teams. In calculating what is supposed to happen in these games, Laxpower seems to be mixing 2 completely seperate pools. The effect of this is that a Div. A team (say rated 70ish in the A group) that whips a top rated Div B team (rated 70ish in the B pool) is getting rewarded like they had whipped a Div. A team (rated 70). It also looks like it's messing with the strength of schedule rankings where teams playing top Div B teams get a higher average.
I'm not sure how much poll voters are putting stock into the actual rankings, but there are a few things that I have noticed scanning through, looking for scores which may effect the "calculations" and possibly indirectly relative perceptions. I was wondering if other people noticed them too, or am I the only one that spends too much time trolling for scores around the nation. Someone'd probably have to email laxpower directly to see if it's a just display error or if these things are actually affecting the rankings. We're right around the time when the laxpower ratings would start to "be more representative" because more games have been played, so I thought I'd throw this out there.
1) sometimes there are double listings in the quality win factor... so if a team beat another ranked team, they get "bonus points". sometimes that bonus is displayed duplicate for the next game they play, regardless of who it is and regardless of if it is a win or loss. Not sure if they are giving extra bonus when they shouldn't be or if this is just a display error.
2) Laxpower used to rank Div A and Div B teams on one big list, but now it looks like the Division B Laxpower Rating is one group and Div A laxpower rating is another, calculated independently. So a Div A team with a 70ish rating in Div A is completely different than a Div B team with a 70ish rating in the Div B pool (which would probably be rated around 50 if it was one big list). However several Div A teams play Div. B teams. In calculating what is supposed to happen in these games, Laxpower seems to be mixing 2 completely seperate pools. The effect of this is that a Div. A team (say rated 70ish in the A group) that whips a top rated Div B team (rated 70ish in the B pool) is getting rewarded like they had whipped a Div. A team (rated 70). It also looks like it's messing with the strength of schedule rankings where teams playing top Div B teams get a higher average.