Started By
Message
Massey Composite- Laat 5 Years
Posted on 6/7/23 at 9:52 pm
Posted on 6/7/23 at 9:52 pm
So I pulled the Massey Composite Rankings for the last 5 years. The results are below.
Anyone have a good idea on how to use this to compare the SOP’s (Strength of Permanent’s)?
And yes, I’m aware that these rankings are flawed but at least it is a bunch of computers averaged out.
Example Link: Massey Composite 2022 Final
Thoughts?
Anyone have a good idea on how to use this to compare the SOP’s (Strength of Permanent’s)?
And yes, I’m aware that these rankings are flawed but at least it is a bunch of computers averaged out.
Example Link: Massey Composite 2022 Final
Thoughts?
Posted on 6/7/23 at 10:01 pm to BigBro
Exclude the fake year 2020. It was a glorified spring game season for everyone
This post was edited on 6/7/23 at 10:02 pm
Posted on 6/7/23 at 10:05 pm to VermilionTiger
quote:
Exclude the fake year 2020. It was a glorified spring game season for everyone
While you’re at it, exclude the 2 years of chad morris for us. Thanks.
Posted on 6/7/23 at 10:08 pm to VermilionTiger
quote:
Exclude the fake year 2020. It was a glorified spring game season for everyone
2018, 2019, 2021, 2022
This post was edited on 6/7/23 at 10:21 pm
Posted on 6/7/23 at 10:28 pm to BigBro
quote:
2018, 2019, 2021, 2022
Now we need you to exclude the two years of Chad Morris.
Posted on 6/7/23 at 10:40 pm to Farmer1906
delete
This post was edited on 6/7/23 at 10:51 pm
Posted on 6/7/23 at 10:45 pm to TheCheshireHog
quote:
Now we need you to exclude the two years of Chad Morris.
how about 2020-2022 for you..
This post was edited on 6/7/23 at 10:52 pm
Posted on 6/8/23 at 6:32 am to BigBro
This is a really great thread. I'm so glad we invited these cerebral superstars to the conference. What an asset.
Posted on 6/8/23 at 8:16 am to BigBro
quote:
how to use this to compare the SOP’s
Assuming we know who those are, I think you could take some combination of a team's median ranking in this system over the last five (or three) years in this system [past], a [current] ranking, and perhaps some random factors like a team talent composite, etc. [future] and arrive at a comparative ranking by summing those overall metrics.
I love to do this type of stuff, and SOG has some really good personal ranking systems that he likes to post. (I hardly ever post mine, because quite frankly his logic is usually similar, if not better.)
As he and I have discussed, the issue will almost always be how you weight these metrics to arrive at your final rating. In other words, how much do you value past performance versus current or even future projected performance? How much has SOS affected past values? Should that be included? If so, how much is that actually worth? ... ad infinitum.
IMO, the weights assigned to each factor is what separates social media pontifications from real, analytical/predictive systems like the BCS (or Massey).
Anyway, just my two cents.
Posted on 6/8/23 at 8:45 am to paperwasp
Considering a little more, one thing I don't like about using systems like Massey (for this) is that it takes those complex algorithms and averages their rank straight-up.
So you end up with #1, #2, #3, etc.
But in reality there may be a huge tier between #1, #2, and #3 — like 100, 98, 76.
Assigning and averaging just the ranks removes these relative gaps and pulls any obvious outliers back together (where they're then only separated by one unit across the board).
So you end up with #1, #2, #3, etc.
But in reality there may be a huge tier between #1, #2, and #3 — like 100, 98, 76.
Assigning and averaging just the ranks removes these relative gaps and pulls any obvious outliers back together (where they're then only separated by one unit across the board).
Posted on 6/8/23 at 2:25 pm to paperwasp
quote:
Assigning and averaging just the ranks removes these relative gaps and pulls any obvious outliers back together (where they're then only separated by one unit across the board).
Good feedback, and I agree on everything you said. Here is one sticking point for me though..
At the end of the day, does it really matter?
From my POV as a Texas fan, I don’t care if Oklahoma, A&M and Arkansas all made the playoffs or all finished tied for last place in the conference. Those are the 3 games that I want.
(assuming 3 permanent)
Maybe that changes, or maybe other teams don’t feel that way. It’s also very possible that we don’t get the full story going on behind the scenes.
IE I assume Alabama wants Tennessee and vice versa, but maybe that isn’t true.
But I am 99% confident that Texas would schedule Oklahoma OOC if they didn’t show up on our schedule one year. I would have said the same thing about the A&M game before 2012.
Popular
Back to top
Follow SECRant for SEC Football News