GT should be ranked in the top 25

ibeattetris

ibeattetris

Helluva Engineer
Messages
3,604
http://www.espn.com/college-football/statistics/teamratings/_/tab/efficiency
Team efficiencies are based on the point contributions of each unit to the team's scoring margin, on a per-play basis. The values are adjusted for strength of schedule and down-weighted for "garbage time" (based on win probability). The scale goes from 0 to 100; higher numbers are better and the average is roughly 50 for all categories. Efficiencies update daily during the season.
Click to expand...
38 - GT
42 - UVA
 
ibeattetris

ibeattetris

Helluva Engineer
Messages
3,604
AE 87 said:
Obviously, I'm using a raw stat and not a sophisticated calculation of individual strength of schedule. I don't see any easy way to use it to calculate conference strength.
Click to expand...
Have you ever looked into "average opponent of opponent ppddiff"? The assumption being, the higher your opponent of opponent ppddiff is, the higher you overall strength of schedule is. You then average that across a conference and you may be able to determine strength of a conference. You can do things like only count OOC, or counting all.

Just food for thought.
 
AE 87

AE 87

Helluva Engineer
Messages
13,026
ibeattetris said:
Have you ever looked into "average opponent of opponent ppddiff"? The assumption being, the higher your opponent of opponent ppddiff is, the higher you overall strength of schedule is. You then average that across a conference and you may be able to determine strength of a conference. You can do things like only count OOC, or counting all.

Just food for thought.
Click to expand...

I just play with raw stats from cfbstats.com using a spreadsheet.

Doing what you suggest would require more sophistication and is beyond my interest.
 
bobongo

bobongo

Helluva Engineer
Messages
7,580
ibeattetris said:
Things just start to look really weird once you don't take any stats into consideration
http://www.cpiratings.com/top25.html
This only looks at:
Record
Opponent Record
Opponent Opponent Record

The top 4 is almost always easy due to straight record, but does anyone really think Georgia (the team that lost to LSU whom Alabama annihilated) is better than Alabama? Based on a static SOS approach where only W/L matter, then that certainly could be the case, since Alabama has beaten nobody and apparently UGA has. You could always try to manually craft these sorts of things, but that seems incredibly time consuming and open to the bias I mention. I certainly value looking at opponent record and opponent opponent record as part portion of the criteria, but I think it should really only be a part of the full story, and not the only story.
Click to expand...

The more I look at these rankings, the more they make sense to me. I don't think they look all that weird at all, really. There are some rankings that seem off to the casual subjective observer, but by and large they seem right and are based on results on the field using the metric that matters most in my opinion, winning and losing. But the weakness in the poll is that is doesn't seem to accurately reflect SOS. The winning percentage of G5 teams, for example, is achieved against other G5 teams for the most part, and the winning percentage of the teams they play will likewise be against inferior opponents, on average.

Nonetheless, what I would like to do is take this ranking at the end of the season and apply it to the bowl games. Let's see what kind of win percentage is achieved by the higher ranking team in every bowl game vs. the winning percentage of the higher ranking teams in another, stat-based poll. That may give us a better idea of the validity of the two polls, although I do believe that polls should reflect something earned on the field as opposed to simply "being the better team" in someone's arbitrary determination. Maybe that's why the poll looks off in some respects. Maybe Alabama is a better team than Georgia, but have they earned it? They've played a cushier schedule, to be sure. But by the end of the season and championship game this may be rectified. At any rate, I will concede that the poll may be off in some respects, but I will also offer that the other polls seem off in some respects as well. I notice that Mississippi State is ranked #50 in the poll. That seems about right to me, and seems to be a much more realistic ranking than the one they are given in the other, popular polls

So let's compare polls scientifically. After the season and all the championship games are played in each conference, I'll count how many of the bowl games are won by the team ranked higher in this poll, and you can count how many games are won by the team ranked higher in any stat-based poll of your choosing. Then we can see how that turns out. Deal?
 
MikeJackets1967

MikeJackets1967

Helluva Engineer
Messages
14,844
Location
Lovely Ducktown,Tennessee
FightWinDrink said:
for the people that care, we check in at 38th in the Coaches poll with 10 votes

they have us 6th in the ACC for an idea of how we're perceived vs our conference peers and maybe some insight into bowl season
Click to expand...
It looks like if GT upsets #5 Georgia they could be #20-#25 ranked going into their Bowl Game where ever it is.
 

Deleted member 2897

Guest
FightWinDrink said:
for the people that care, we check in at 38th in the Coaches poll with 10 votes

they have us 6th in the ACC for an idea of how we're perceived vs our conference peers and maybe some insight into bowl season
Click to expand...

That’s a big fat crock of crap.

Iowa State is #16 at 6-4. Boston College is a train wreck, but ranked #20 at 7-4.

Georgia Tech is the most underrated, overlooked, disrespected team in the country.
 

Heisman's Ghost

Helluva Engineer
Messages
4,867
Location
Albany Georgia
CINCYMETJACKET said:
I don't see Tech's schedule as being worthy of a top 25 if we had 8 wins at this point. Wins vs. Alcorn St., Bowling Green, Louisville, VT, UNC, and Miami. Losses against USF, Pitt, Clemson, and Duke. In a typical season, perhaps that gets it done with wins vs LOU/UNC, VT, and MIA, but they're all down this year. In hindsight, from what I've seen from the teams that we've played on our schedule so far this year, the only team that is relevant would be Clemson. Perhaps Pitt, who is receiving votes, and so far has lost to #13 PSU, #13 UCF, and #5 ND. But they also lost to UNC. If we beat UVA and uga, then I'd think it may be reasonable. At this time, no.
Click to expand...

It is written in Scripture, somewhere, that any Tech team that loses to Duke must beat UGA to atone for that sin in order to be ranked in the top 25. I know it is in the Bible somewhere, maybe Revelations.
 

Heisman's Ghost

Helluva Engineer
Messages
4,867
Location
Albany Georgia
smathis30 said:
As much **** as people give Alabama for their schedule, they have already beaten 5 already bowl eligible teams, which trails just Clemson out of college teams who has 6. GT has 0. Every ranked team currently has at least one.
Click to expand...

All you have to do is watch Alabama and you know that their SECOND team would beat about 70% of the teams in college football. I thought that they would slide at least a little bit when they lost Ridley off last year's team but damned if they did not get better. They should be playing in some kind of professional league with the other teams that are pay for play like UGA, the Falcons, Tennessee, Ohio State etc.
 

Heisman's Ghost

Helluva Engineer
Messages
4,867
Location
Albany Georgia
ibeattetris said:
Things just start to look really weird once you don't take any stats into consideration
http://www.cpiratings.com/top25.html
This only looks at:
Record
Opponent Record
Opponent Opponent Record

The top 4 is almost always easy due to straight record, but does anyone really think Georgia (the team that lost to LSU whom Alabama annihilated) is better than Alabama? Based on a static SOS approach where only W/L matter, then that certainly could be the case, since Alabama has beaten nobody and apparently UGA has. You could always try to manually craft these sorts of things, but that seems incredibly time consuming and open to the bias I mention. I certainly value looking at opponent record and opponent opponent record as part portion of the criteria, but I think it should really only be a part of the full story, and not the only story.
Click to expand...

Forget UGA, I am beginning to think there are some pro teams not much better than Alabama when they are healthy and clicking. This, is of course, meant for entertainment purposes only but still. If they win out, a case can be made for them to be the best team in college football history.
 
Top