FEI Off/Def/SpT Rankings 2018

iceeater1969

Helluva Engineer
Messages
8,916
I love this thread on stats as we head into last third of the season with the fcs games in the past. We have the " 5000 - 7000 posts guys gathering data from their favorite sites. Then the 10000 and above guys reply with their info.
I used to comment with the count of wins by teams we beat (i.e before vt the 2 power 5 teams we beat were 1-13) or some common sense metric. Then when I ran out of fingers and toes to count this type win and that type lose , I learned to relax and watch the experts duel it out.


.
 

ibeattetris

Helluva Engineer
Messages
3,545
Now, why doesn't Tech schedule like that? Instead we line up home and homes with UCF and Notre Dame. That's insane, imho. A large part of winning is a schedule that allows you to, you know, win.
I understand your point. But there is a very vocal population that thinks we should never schedule FCS and that we should only even have one G5 team a year. A lot of these posters are fine losing games against better teams than winning against lesser.
 

bobongo

Helluva Engineer
Messages
7,043
I understand your point. But there is a very vocal population that thinks we should never schedule FCS and that we should only even have one G5 team a year. A lot of these posters are fine losing games against better teams than winning against lesser.

Personally, I'd rather PLAY better teams than lesser. Who knows, we might actually win.
 

ibeattetris

Helluva Engineer
Messages
3,545
The only stats that matter are who you beat and who you lost to. All the rest is just blog fluff.
People were wrong about both the importance and which stats mattered in baseball too, so I understand your misunderstanding (y)

The goal of advanced statistics isn't "blog fluff", it's to determine key factors of success for a team. According to FEI, our offense is good and our defense is bad. This leads us to being around the 45th best team. Is anyone going to argue that that number is wrong given how we have played? At least there is internally consistent logic to determining the strength of teams via advanced analytics. The same cannot be said towards the "eye test" or head to head comparisons or whatever other garbage leads to these biased polls we have now.

Now I look at GT and see we have an FEI and SP+ offense in the top 15, defense below the bottom 90, and special teams in the bottom 115. Based on that, should improve our offense or our defense?
 

GT_05

Helluva Engineer
Messages
2,370
People were wrong about both the importance and which stats mattered in baseball too, so I understand your misunderstanding (y)

The goal of advanced statistics isn't "blog fluff", it's to determine key factors of success for a team. According to FEI, our offense is good and our defense is bad. This leads us to being around the 45th best team. Is anyone going to argue that that number is wrong given how we have played? At least there is internally consistent logic to determining the strength of teams via advanced analytics. The same cannot be said towards the "eye test" or head to head comparisons or whatever other garbage leads to these biased polls we have now.

Now I look at GT and see we have an FEI and SP+ offense in the top 15, defense below the bottom 90, and special teams in the bottom 115. Based on that, should improve our offense or our defense?

He was talking in an anecdotal sense, I believe.


Sent from my iPhone using Tapatalk
 

Cam

Helluva Engineer
Messages
1,591
Location
Atlanta, Georgia
The only issue I have with these statistics is that they do not take into account things like injuries or new systems. Our defense has gotten progressively better this season, but they're weighed down by some poor performances in our early schedule when we were figuring things out. I think our defense as of the last few weeks has been better than their rating. Alternatively, a great team could lose a key player and their ratings won't (immediately) be adjusted to account for it. That's where the eyeball test comes in.
 

ibeattetris

Helluva Engineer
Messages
3,545
The only issue I have with these statistics is that they do not take into account things like injuries or new systems. Our defense has gotten progressively better this season, but they're weighed down by some poor performances in our early schedule when we were figuring things out. I think our defense as of the last few weeks has been better than their rating. Alternatively, a great team could lose a key player and their ratings won't (immediately) be adjusted to account for it. That's where the eyeball test comes in.
That’s definitely true. These are meant to be season indicators not a measure of team velocity. It would definitely be cool if they provided the ability to look at only certain periods, but then the opponent adjust stats start to get harder to judge in general as well.

With only 12 games, football struggles from high variance more than other sports. But that is also why record does not always reflect actual performance.
 

bobongo

Helluva Engineer
Messages
7,043
That’s definitely true. These are meant to be season indicators not a measure of team velocity. It would definitely be cool if they provided the ability to look at only certain periods, but then the opponent adjust stats start to get harder to judge in general as well.

With only 12 games, football struggles from high variance more than other sports. But that is also why record does not always reflect actual performance.

True for the offense as well - vast swings from running at will to being stuffed - from the sizzle at VT to the fizzle against Duke. But the overall stats just grind it all up into one uniform number. The real story is what's behind the stats, and that's especially true in our case because we go from the penthouse to the outhouse and back again from one week to the next.
 

takethepoints

Helluva Engineer
Messages
5,877
That’s what Duke does. It helps them go 6-6 even though they go 2-6 in ACC play.
I suppose the real problem here is that for those who would like to get rid of the present staff, nothing would do. Suppose we schedule like Duke and win 8 games (easy, imho). Then they'll say that we got into a bowl game and a winning season because we played a schedule of cream puffs. "We need to be better! We shouldn't accept mediocrity!" So we don't. We schedule UCF and ND instead, two very successful programs that produce teams that are hard to beat. And Tech goes (maybe) 6 - 6 and loses both games (likely). Then they'll say that our offense limits recruiting and we need to loosen admission standards for football players. "We need to be better! We shouldn't accept mediocrity!"

Me, I'm pretty straightforward on this: I want to win more games and I don't want the schedule to stand in the way of doing so. And 10 years later, nobody will say anything about the teams who won more except "Gee, we were pretty good back then." Because it's your won - lose record that counts.
 

bobongo

Helluva Engineer
Messages
7,043
I suppose the real problem here is that for those who would like to get rid of the present staff, nothing would do. Suppose we schedule like Duke and win 8 games (easy, imho). Then they'll say that we got into a bowl game and a winning season because we played a schedule of cream puffs. "We need to be better! We shouldn't accept mediocrity!" So we don't. We schedule UCF and ND instead, two very successful programs that produce teams that are hard to beat. And Tech goes (maybe) 6 - 6 and loses both games (likely). Then they'll say that our offense limits recruiting and we need to loosen admission standards for football players. "We need to be better! We shouldn't accept mediocrity!"

Me, I'm pretty straightforward on this: I want to win more games and I don't want the schedule to stand in the way of doing so. And 10 years later, nobody will say anything about the teams who won more except "Gee, we were pretty good back then." Because it's your won - lose record that counts.

If we schedule a bunch of creampuffs and people say our winning season doesn't mean anything because we didn't play anybody, they'll be telling the truth. The biggest reason I don't want to see us beat up on a schedule of creampuffs is because it's boring and meaningless.
 

GT_05

Helluva Engineer
Messages
2,370
If we schedule a bunch of creampuffs and people say our winning season doesn't mean anything because we didn't play anybody, they'll be telling the truth. The biggest reason I don't want to see us beat up on a schedule of creampuffs is because it's boring and meaningless.

UCF


Sent from my iPhone using Tapatalk
 

ibeattetris

Helluva Engineer
Messages
3,545
True for the offense as well - vast swings from running at will to being stuffed - from the sizzle at VT to the fizzle against Duke. But the overall stats just grind it all up into one uniform number. The real story is what's behind the stats, and that's especially true in our case because we go from the penthouse to the outhouse and back again from one week to the next.
We keep bringing up the offense, but we could say the same thing about the defense. Outside of Duke, the defense hasn't exactly helped much in winning any of our games. Sometimes you have to win a game 14-7. We don't do that.

Let's say you score on 80% of your drives throughout a season, but in one of the games you didn't score (maybe your star QB was out and things didn't work out). Just because you lost that game because of the offense doesn't mean the offense is bad. You still have an amazing offense.

Season based stats like this are about telling a comprehensive story, not about necessarily justifying wins or losses. I would never take an advanced stat and use it to predict games, because as @Cam mentioned, it misses the context of situations at times. That doesn't invalidate them though.
 

bobongo

Helluva Engineer
Messages
7,043
UCF


Sent from my iPhone using Tapatalk

But for an AAC team, they had a pretty good OOC schedule lined up, including Pitt, UNC, and Navy (although I see UNC ducked their game - something about a hurricane ;=). Now as far as them being number 12 in the country, maybe not really.
 

ibeattetris

Helluva Engineer
Messages
3,545
Previous games have updated in FEI
GT offense coming in at 11th place and only 0.02 away from WVU for 10th.
Defense still struggling at 95th.

S&P+ has GT at 17 for offense and 104 on defense.

Combined rank of 52 (fifth best 4+ loss team..)

Turnovers continue to be the make or break aspect of our team.
 
Top