|
|
|
|
|
| I assure you there is much less of a smile. |
|
|
|
Appears our tie game today with Sexy Flanders results in no up or down for either team in the ELO. The lower team should go up and the upper team down. Any other result just doesn't make sense...unless the two teams are statistically even...which in this case they weren't.
It is my opinion (I figured I would just skip "humble") that without this portion of the ladder system working, you will never have proper rankings. As Jamie previously said, unless you start beating Tele' badly, and everyone else does also, you just can't catch them even if your team is just as good.
 |
|
|
|
| #1 team beat #3 by 6 points last week and they didn't move up at all, #2 team won and got 6 points putting them much closer to the top. #4 team earned 5 points and leapfrogged two other teams. Seems like there is lots of movement and people can catch anyone. Agree that the starting points probably seperated the teams unfairly but without it the first couple weeks might have been very rough for some of the lower teams(which is usually how things go in a ladder system)so I understand the reasons to avoid that. |
|
|
|
I think the fact the third place team is 1-6 highlights the starting points problem. The fourth and fifth place teams are in the middle division so they are playing "weaker" teams, making their ranking not comparative to the first three teams.
I think my point is, the current model doesn't have a proper over/under...not that nobody is being assigned any points.
But I guess even more so, my point is if this is the way of the future, we should be playing with the formula now when it doesn't matter. |
|
|
|
Bonk is where they need to be, just look at their points difference assuming they played the top teams.
This ladder doesn't consider wins/losses, just points relative to their opponent. If you want wins/losses we can go back to the previous ladder formula.
Ratings tend to converge on a team's true strength relative to its competitors after about 30 matches. (https://en.wikipedia.org/wiki/World_Football_Elo_Ratings)
With only 7 games played, the rating don't seem that far off if 30 games is the magic number.
The link I provided above offers some additional insight on how the world football uses ladders. We might be able to apply some concepts here.
Also, since the standings only shows the current points, we could perhaps consider the following alternative ways to rank teams:
- Rank based on Highest Point Score achieved in the Season
- Rank based on Average Point Score
|
|
|
|
| I grew up playing snakes and ladders. This is not a ladder. Maybe a snake dressed up as a ladder, but not a ladder. |
|
|
|
| Marc wrote: |
Ratings tend to converge on a team's true strength relative to its competitors after about 30 matches. (https://en.wikipedia.org/wiki/World_Football_Elo_Ratings)
|
I think I missed this earlier, but our indoor season is 24 weeks long. If it takes 30 matches for this system to find a team's true strength, then.... right? |
|