Home
Articles
Photos
Interviews
Forums
New posts
Search forums
Georgia Tech Recruiting
Dashboard
What's new
New posts
New profile posts
Latest activity
Chat
Log in
Register
What's new
Search
Search
Search titles only
By:
New posts
Search forums
Menu
Log in
Register
Install the app
Install
Home
Forums
Georgia Tech Athletics
Georgia Tech Football
Tech vs. Louisville
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser
.
Reply to thread
Message
<blockquote data-quote="cpf2001" data-source="post: 957678" data-attributes="member: 6459"><p>Further than that even, right? Most programs don't change that drastically year over year. Coach stays the same, roster evolves, lose some production, gain some production. Those are all fairly "modeling friendly" events.</p><p></p><p>Then there are the Deion's or the "wow maybe Geoff Collins really brought the whole thing down that much" rare events.</p><p></p><p>They're tougher to figure out how to quantify <strong>and</strong> they might not even make a big impact on your overall model quality evaluation. Are you testing more for "how accurate did this do on the largest number of teams" or "how accurate did this do on the most random one-off changes"? Or trying to weigh both somehow? From what I've seen, the majority of systems built out there right now are squarely in the first camp: predict well for the standard cases. So even if some of them do better on the wildcard cases, that doesn't mean that anyone necessarily can tell you <em>which ones</em> those are or has a good way of ranking for that.</p></blockquote><p></p>
[QUOTE="cpf2001, post: 957678, member: 6459"] Further than that even, right? Most programs don't change that drastically year over year. Coach stays the same, roster evolves, lose some production, gain some production. Those are all fairly "modeling friendly" events. Then there are the Deion's or the "wow maybe Geoff Collins really brought the whole thing down that much" rare events. They're tougher to figure out how to quantify [B]and[/B] they might not even make a big impact on your overall model quality evaluation. Are you testing more for "how accurate did this do on the largest number of teams" or "how accurate did this do on the most random one-off changes"? Or trying to weigh both somehow? From what I've seen, the majority of systems built out there right now are squarely in the first camp: predict well for the standard cases. So even if some of them do better on the wildcard cases, that doesn't mean that anyone necessarily can tell you [I]which ones[/I] those are or has a good way of ranking for that. [/QUOTE]
Insert quotes…
Verification
What is the last name of the current Head Football Coach?
Post reply
Home
Forums
Georgia Tech Athletics
Georgia Tech Football
Tech vs. Louisville
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
Accept
Learn more…
Top