Regional Week is here! All week, I’ll be posting these projections. I’ll start with the Friday regions, then work on the Saturday ones.
These are a bit different than they’ve been in the past. I’m trying to project as always, but changing it up a bit in terms of the methodology. In the past, I’ve just sorted out by the ratings, then scored as if they’re a cross country meet. Here, I’m doing the same thing, but running simulations. Essentially, given a runner’s ratings and the variance in their ratings throughout the season, I’m able to generate a large sample of “races” and not just one scenario. The thought is that this will give a better depiction of what may occur. If an athlete is consistent, that will be represented. If an athlete is polarizing, but has a chance at greatness, that will be represented. Both aren’t possible when just throwing out one race. The Macomb County projections were done in the same manner and TullyRunners explains it here.
By simulating many races, I can hand out odds and percentages as well. The team projections have odds of winning and odds of placing in the top three. The individual ones have an athlete’s average place. I do have odds on an athlete winning and getting top-15, but I’m not gonna post that ahead of time. Maybe after and we can see who came through in a huge way on the huge stage.
What I’ve learned from the past is that I write way too much in the previews early in the week, then worry I’ll run out of time mid-week, running on fumes and puttering out by Thursday. The writing is the time-consuming part, the maths are actually easy. I’ll save the writing chops for next week, where penning four previews is easier on the brain than thirty-six.
Always remember, races aren’t run on paper or on a computer. These are meant to better our sport, give more attention where little is given. Although I take pride in their accuracy, this endeavor would be quite boring if it were always perfect. I look forward to these being proven wrong and I expect to hear about it as well.
Regional Week is here! All week, I’ll be posting these projections. I’ll start with the Friday regions, then work on the Saturday ones.
These are a bit different than they’ve been in the past. I’m trying to project as always, but changing it up a bit in terms of the methodology. In the past, I’ve just sorted out by the ratings, then scored as if they’re a cross country meet. Here, I’m doing the same thing, but running simulations. Essentially, given a runner’s ratings and the variance in their ratings throughout the season, I’m able to generate a large sample of “races” and not just one scenario. The thought is that this will give a better depiction of what may occur. If an athlete is consistent, that will be represented. If an athlete is polarizing, but has a chance at greatness, that will be represented. Both aren’t possible when just throwing out one race. The Macomb County projections were done in the same manner and TullyRunners explains it here.
By simulating many races, I can hand out odds and percentages as well. The team projections have odds of winning and odds of placing in the top three. The individual ones have an athlete’s average place. I do have odds on an athlete winning and getting top-15, but I’m not gonna post that ahead of time. Maybe after and we can see who came through in a huge way on the huge stage.
What I’ve learned from the past is that I write way too much in the previews early in the week, then worry I’ll run out of time mid-week, running on fumes and puttering out by Thursday. The writing is the time-consuming part, the maths are actually easy. I’ll save the writing chops for next week, where penning four previews is easier on the brain than thirty-six.
Always remember, races aren’t run on paper or on a computer. These are meant to better our sport, give more attention where little is given. Although I take pride in their accuracy, this endeavor would be quite boring if it were always perfect. I look forward to these being proven wrong and I expect to hear about it as well.
Regional Week is here! All week, I’ll be posting these projections. I’ll start with the Friday regions, then work on the Saturday ones.
These are a bit different than they’ve been in the past. I’m trying to project as always, but changing it up a bit in terms of the methodology. In the past, I’ve just sorted out by the ratings, then scored as if they’re a cross country meet. Here, I’m doing the same thing, but running simulations. Essentially, given a runner’s ratings and the variance in their ratings throughout the season, I’m able to generate a large sample of “races” and not just one scenario. The thought is that this will give a better depiction of what may occur. If an athlete is consistent, that will be represented. If an athlete is polarizing, but has a chance at greatness, that will be represented. Both aren’t possible when just throwing out one race. The Macomb County projections were done in the same manner and TullyRunners explains it here.
By simulating many races, I can hand out odds and percentages as well. The team projections have odds of winning and odds of placing in the top three. The individual ones have an athlete’s average place. I do have odds on an athlete winning and getting top-15, but I’m not gonna post that ahead of time. Maybe after and we can see who came through in a huge way on the huge stage.
What I’ve learned from the past is that I write way too much in the previews early in the week, then worry I’ll run out of time mid-week, running on fumes and puttering out by Thursday. The writing is the time-consuming part, the maths are actually easy. I’ll save the writing chops for next week, where penning four previews is easier on the brain than thirty-six.
Always remember, races aren’t run on paper or on a computer. These are meant to better our sport, give more attention where little is given. Although I take pride in their accuracy, this endeavor would be quite boring if it were always perfect. I look forward to these being proven wrong and I expect to hear about it as well.
Michigan XC Speed Ratings 