Traditional Rankings Algorithm Restored
After making a Week #11 update to the rankings algorithm on December 4, 2013, we have reconsidered this change. The algorithm that has been the standard for the past five years, has been restored to the site. Week #11 rankings have been updated to reflect this return to the traditional algorithm.
Does this mean that we think the update was wrong? No. It means we agree that changing the algorithm mid-season was a poor decision. The reasons that drove the change are valid. The change does address most of these issues and is truly an improvement in general. It does have a few shortcomings. And like any complex issue, there are a solid list of pros and cons. We will not go through each one in detail. They are numerous and complex.
MYHockey surveyed it's top volunteers regarding the change. Over 100 volunteers responded with 65% in favor, 20% neutral, 15% against. If this were a democratic process, the change would have stayed in place. I am sure I will receive a bunch of email now telling us that we have no backbone. That we should have stuck to our guns. I get it. All opinions are valid.
We are reverting back to the traditional algorithm because we agree that changing the rules mid-season is not "right". Secondly, there is valid concern that the tradition of MYHockey rankings improving in their "accuracy" over the course of the season as more and more interplay between teams from around North America improves the overall rankings. By limiting the rankings to the last 20 games played, this annual improvement in the overall rankings is jeopardized.
The concept of understanding how teams are "trending" is very valuable. Instead of basing the rankings only upon the last 20 games recorded, MYHockey is now working on a series of enhancements to allow users to view the rankings in multiple ways. A "trending" view will be included in this view. We anticipate having these enhancements available within the month. In this way, if you agree that the rankings don't put enough emphasis on recent performance, you can see this view. We have some additional ideas that we have been mulling over for months (in some cases years), that this new platform will give us the opportunity to provide you as we finalize these additional information views.
We appreciate all the feedback. We listen and we care. While passionate pleas for one change or another come across our email on a regular basis, passionate pleas are not typically that persuasive. Analytical arguments get our attention and ultimately it was the analytical arguments that have lead to the original change and then the reversal of this decision.
Does this mean that we think the update was wrong? No. It means we agree that changing the algorithm mid-season was a poor decision. The reasons that drove the change are valid. The change does address most of these issues and is truly an improvement in general. It does have a few shortcomings. And like any complex issue, there are a solid list of pros and cons. We will not go through each one in detail. They are numerous and complex.
MYHockey surveyed it's top volunteers regarding the change. Over 100 volunteers responded with 65% in favor, 20% neutral, 15% against. If this were a democratic process, the change would have stayed in place. I am sure I will receive a bunch of email now telling us that we have no backbone. That we should have stuck to our guns. I get it. All opinions are valid.
We are reverting back to the traditional algorithm because we agree that changing the rules mid-season is not "right". Secondly, there is valid concern that the tradition of MYHockey rankings improving in their "accuracy" over the course of the season as more and more interplay between teams from around North America improves the overall rankings. By limiting the rankings to the last 20 games played, this annual improvement in the overall rankings is jeopardized.
The concept of understanding how teams are "trending" is very valuable. Instead of basing the rankings only upon the last 20 games recorded, MYHockey is now working on a series of enhancements to allow users to view the rankings in multiple ways. A "trending" view will be included in this view. We anticipate having these enhancements available within the month. In this way, if you agree that the rankings don't put enough emphasis on recent performance, you can see this view. We have some additional ideas that we have been mulling over for months (in some cases years), that this new platform will give us the opportunity to provide you as we finalize these additional information views.
We appreciate all the feedback. We listen and we care. While passionate pleas for one change or another come across our email on a regular basis, passionate pleas are not typically that persuasive. Analytical arguments get our attention and ultimately it was the analytical arguments that have lead to the original change and then the reversal of this decision.
