Mahikoa
Footballguy
Statistically, team defense seems like it should be something that varies week to week drastically, but should be relatively easy to predict (on an average basis over many samples) how a defense should do against a particular offense. I've never actually seen this done statistically - meaning, a defense that averages 250 yards against going against an offense averaging 310 yards p/g will give up 270 yards in actuality on average.
This seems like it could have a ton of use in a lot of leagues. So, I'd like to lead a project to figure this out and at least give decent estimates. Some ground rules:
1. Yes, there are other ways of doing this, and this method doesn't account for injuries, trades, signings, etc. I'm assuming you'll adjust for that after the numbers.
2. If you don't like lots of number crunching, this isn't the thread for you. Sorry.
3. I'm more than happy to adjust any methods I come up with based on feedback. I won't do it unless I agree with it. However, I'm more than willing to admit that I know nothing about this, and just have a vague idea of how this is supposed to work.
Now that those are established (and more might be added later), let's talk about next steps. We need to find metrics to study and evaluate to come upon . What stats are going to be useful to 80% or more of leagues -- meaning, which stats do most leagues count for points? These are the ones I can think of off the top of my head:
1. Yards against
2. Points against
3. Interceptions
4. Fumbles recovered
5. Defensive touchdowns
6. Special teams touchdowns
There are many next steps in my mind after we set ground-rules on stats we're looking for - including how to evaluate these on a week by week basis, a season by season basis, etc. If you're willing to help collect/analyze/critique results, please step forward. If you're not interested, or think this isn't the method you'd like to use, please use your creative energies elsewhere.
Thanks!
This seems like it could have a ton of use in a lot of leagues. So, I'd like to lead a project to figure this out and at least give decent estimates. Some ground rules:
1. Yes, there are other ways of doing this, and this method doesn't account for injuries, trades, signings, etc. I'm assuming you'll adjust for that after the numbers.
2. If you don't like lots of number crunching, this isn't the thread for you. Sorry.
3. I'm more than happy to adjust any methods I come up with based on feedback. I won't do it unless I agree with it. However, I'm more than willing to admit that I know nothing about this, and just have a vague idea of how this is supposed to work.
Now that those are established (and more might be added later), let's talk about next steps. We need to find metrics to study and evaluate to come upon . What stats are going to be useful to 80% or more of leagues -- meaning, which stats do most leagues count for points? These are the ones I can think of off the top of my head:
1. Yards against
2. Points against
3. Interceptions
4. Fumbles recovered
5. Defensive touchdowns
6. Special teams touchdowns
There are many next steps in my mind after we set ground-rules on stats we're looking for - including how to evaluate these on a week by week basis, a season by season basis, etc. If you're willing to help collect/analyze/critique results, please step forward. If you're not interested, or think this isn't the method you'd like to use, please use your creative energies elsewhere.
Thanks!
