Why (Advanced) Stats Matter in College Football

[anon_7tbtqcx308nxh]

This is a good read for those of us who like to use advanced stats to bet or think about college football games. It's also a nice introduction for the unaware and uninitiated:

http://www.sbnation.com/college-football/2015/3/6/8157709/college-football-advanced-stats

 

[CatDaddy4daWin]

Special teams is one area I hope we can really improve on. If you are able to get some touchdowns or stops on special teams it wins games. period. We haven't had a kickoff/punt return threat since Dicky Lyons Jr. Boom could definitely be that type of guy, or Timmons if he could get steady on punt returns.

Good article though. Some of Mumme's 4th down shenanigans are probably backed up by good statistics. I believe our staff is doing this stuff in many ways. For instance, have definitely noticed a difference in injuries since we started Korem's program. I don't know what the numbers are but we have been fortunate. Some of that's luck and some of it is training the right way.

 

[Larry Skaggs]

How many wins is it predicting for the Cats?

 

[Deeeefense]

Wow - that's enough data to make Doctor Bob's head spin.

 

[WildCard]

For the most part interesting but a number of his points make little sense (at least as explained). Also note the first sentence:

I began tinkering with college football play-by-play data in 2007 because nobody else was.

Most of today's stats are either recorded stats or what I call "manufactured" stats, i.e., stats that are literally "manufactured" from recorded stats (e.g., yards or plays per point; sacks per attempt, etc.). these are aggregate stats based on the game or on the season. Seldom do such "analytics" delve into "play by play" stats.

However, coaches do that every week when breaking down film of multiple games trying to gain insight on "what" somebody does on 3rd and 3 between the 30s. Good coaches (especially guys that have a rep of doing "more with less" literally write a book on the opponents.

Some of his points are of great interest (e.g., points 1 and 2). I'm not sure how you can quantitatively separate the performance of "passer and receivers". I think you can probably separate "blockers and runners" by looking at short yardage plays and so called "big plays". A team with much success on short yardage plays probably has pretty good run blocking. OTOH, a team that has more "explosive" plays probably has better runners. Consider Boom Williams; made a number of explosive run plays for good yardage and good yards/attempt but UK was only #79 in rushing yards/game and #105 in YPR on 3rd and 3 or less to go. (Source: cfb.stats.com)

As I look through his list I see a number of things that are "charted" by staffs and in some stat sites (mostly NFL sites). But I see some things (e.g., strategies and tactics) that I think could not be accurately quantified. JMO.

Peace

 

[anon_7tbtqcx308nxh]

For the most part interesting but a number of his points make little sense (at least as explained). Also note the first sentence:

I began tinkering with college football play-by-play data in 2007 because nobody else was.

Most of today's stats are either recorded stats or what I call "manufactured" stats, i.e., stats that are literally "manufactured" from recorded stats (e.g., yards or plays per point; sacks per attempt, etc.). these are aggregate stats based on the game or on the season. Seldom do such "analytics" delve into "play by play" stats.

However, coaches do that every week when breaking down film of multiple games trying to gain insight on "what" somebody does on 3rd and 3 between the 30s. Good coaches (especially guys that have a rep of doing "more with less" literally write a book on the opponents.

Some of his points are of great interest (e.g., points 1 and 2). I'm not sure how you can quantitatively separate the performance of "passer and receivers". I think you can probably separate "blockers and runners" by looking at short yardage plays and so called "big plays". A team with much success on short yardage plays probably has pretty good run blocking. OTOH, a team that has more "explosive" plays probably has better runners. Consider Boom Williams; made a number of explosive run plays for good yardage and good yards/attempt but UK was only #79 in rushing yards/game and #105 in YPR on 3rd and 3 or less to go. (Source: cfb.stats.com)

As I look through his list I see a number of things that are "charted" by staffs and in some stat sites (mostly NFL sites). But I see some things (e.g., strategies and tactics) that I think could not be accurately quantified. JMO.

Peace

I think what he was trying to get at in terms of "tinkering" was working on a more up-to-date box score that goes beyond total yards, points, etc because those variables don't really get that deep. People were doing it for the NFL but not CFB. That's an assumption on my part based on some of his other posts.

No doubt, staffs are already doing a lot of this, but it's definitely not common knowledge on message boards or blogs where there remains an overabundance of gut assumptions (i.e. "Neal Brown never goes deep") or anecdotal evidence (i.e. "Fred Tiller always gets beat") as opposed to what an entire seasons worth of data actually says.

You're right that a lot of these things will be tough to quantify, and he admits he's grappling with most of these. I find these issues endlessly interesting.

 

[anon_7tbtqcx308nxh]

How many wins is it predicting for the Cats?

His preview for UK should come in July. He does all 120 FBS teams starting in February. I think he's working through either the ACC or Big 12 now.

 

[anon_7tbtqcx308nxh]

Special teams is one area I hope we can really improve on. If you are able to get some touchdowns or stops on special teams it wins games. period. We haven't had a kickoff/punt return threat since Dicky Lyons Jr. Boom could definitely be that type of guy, or Timmons if he could get steady on punt returns.

Good article though. Some of Mumme's 4th down shenanigans are probably backed up by good statistics. I believe our staff is doing this stuff in many ways. For instance, have definitely noticed a difference in injuries since we started Korem's program. I don't know what the numbers are but we have been fortunate. Some of that's luck and some of it is training the right way.

Going for it on 4th down typically is the right decision. Mumme was on to something. Ha.

I've noticed the same about Korem's program, and I draw the same conclusions.

 

[Deeeefense]

Going for it on 4th down typically is the right decision. Mumme was on to something. Ha.

I've noticed the same about Korem's program, and I draw the same conclusions.

I'm glad someone finally pointed out that not only are you statistically better off by going for it strategically on 4th down (under most circumstances) but also, the reason nearly no coach does at - they can't stand the heat when it fails, and also don't want the stats messed up to make them look worse than they really are.