Well, so much for that one, huh? Week 1 of the 2019 season has come and gone, and one of my initial projections is shot.
Given that it was that Calgary would go undefeated, there's no reason to be shocked. And no reason to panic! Even though my ratings got two games out of four wrong, I'd much rather see my ratings improve in their forecasting over time. Let's get the misses out of the way early on.
There's no reason to overreact, either. The Stampeders were, by far, the best team in the league last year. If they should end up collapsing, it'll take a long time to determine that—not just one game.
With that said, here are my post-week 1 ratings.
Notice that the Stamps are still on top. That shows how far out ahead they were to begin with, that they could lose a game by 4 and stay on top. Results here aren't completely based on the game results: I'm maintaining some constants in each team's ratings for a few weeks, until the teams have each played enough of the others that they become "connected" by their common opponents, and then I can really begin to compare them to one another.
Likewise, home field advantage is going to remain fixed at 3.00 for a while. I should expect it to arrive there on its own soon, but letting it move with too few results can cause it to be too high or too low. I've seen it go to double digits and go negative many more times than I should have.
Using these ratings, we arrive at my current standings projection.
Now that games have been played, we see how my setup begins to change with time. The first set of columns, after team names and rankings, is filled in with current records for each team. Everyone but Toronto has "1 GAME PLAYED" and "17 GAMES LEFT".
Perhaps most importantly, we see that, because of the constants in each team's rating, Ottawa and Winnipeg have each won a game they "shouldn't have" and, in turn, Calgary and BC have each lost a game that they "should have" won.
Even after we've established a hierarchy of our teams, no matter how rigid it may seem, we have to remember that games aren't played on paper or in computers. They're played on football fields, and on any given day, some pretty unlikely and downright weird things can happen on those fields. We'll know how good each team "should be" in the end, but records of actual wins and losses are what determine the end to a season.
Speaking of that end, here's the projected playoff bracket we get from these standings.
Winnipeg's big win in Vancouver has allowed them to jump the Eskimos. Additionally, BC's slide has allowed the East Division to prevent a crossover and place a third team in the postseason. Hamilton receives the tiebreaker over Ottawa (who replaces BC in this scenario) based on divisional record.
Be sure to leave any thoughts in the comments here, and don't forget to subscribe if you think you'll enjoy keeping track of the CFL this year with me.
Given that it was that Calgary would go undefeated, there's no reason to be shocked. And no reason to panic! Even though my ratings got two games out of four wrong, I'd much rather see my ratings improve in their forecasting over time. Let's get the misses out of the way early on.
There's no reason to overreact, either. The Stampeders were, by far, the best team in the league last year. If they should end up collapsing, it'll take a long time to determine that—not just one game.
With that said, here are my post-week 1 ratings.
 |
| CFL ratings following one week of games. |
Notice that the Stamps are still on top. That shows how far out ahead they were to begin with, that they could lose a game by 4 and stay on top. Results here aren't completely based on the game results: I'm maintaining some constants in each team's ratings for a few weeks, until the teams have each played enough of the others that they become "connected" by their common opponents, and then I can really begin to compare them to one another.
Likewise, home field advantage is going to remain fixed at 3.00 for a while. I should expect it to arrive there on its own soon, but letting it move with too few results can cause it to be too high or too low. I've seen it go to double digits and go negative many more times than I should have.
Using these ratings, we arrive at my current standings projection.
 |
| Projected standings based on one week of CFL action. |
Now that games have been played, we see how my setup begins to change with time. The first set of columns, after team names and rankings, is filled in with current records for each team. Everyone but Toronto has "1 GAME PLAYED" and "17 GAMES LEFT".
Perhaps most importantly, we see that, because of the constants in each team's rating, Ottawa and Winnipeg have each won a game they "shouldn't have" and, in turn, Calgary and BC have each lost a game that they "should have" won.
Even after we've established a hierarchy of our teams, no matter how rigid it may seem, we have to remember that games aren't played on paper or in computers. They're played on football fields, and on any given day, some pretty unlikely and downright weird things can happen on those fields. We'll know how good each team "should be" in the end, but records of actual wins and losses are what determine the end to a season.
Speaking of that end, here's the projected playoff bracket we get from these standings.
 |
| After just one week, our playoff outlook has changed quite a bit. |
Winnipeg's big win in Vancouver has allowed them to jump the Eskimos. Additionally, BC's slide has allowed the East Division to prevent a crossover and place a third team in the postseason. Hamilton receives the tiebreaker over Ottawa (who replaces BC in this scenario) based on divisional record.
Be sure to leave any thoughts in the comments here, and don't forget to subscribe if you think you'll enjoy keeping track of the CFL this year with me.
Comments
Post a Comment