Going into that game I had no idea how we would score much less move the ball. I remember commenting that night "this is what offense looks like against a top ten opponent".
To me it's disingenuous when I hear OSU was inexperienced...as if we weren't!
Log in or register to post comments about the Virginia Tech Hokies
Amen fellow Hokie. I hate that people always forget that we were just as young and as inexperienced as OSU that night. BTW, we were playin at OSU. On paper, they litreally had every advantage a team could ask for.
Log in or register to post comments about the Virginia Tech Hokies
They were a good team, they just weren't a great team. Over the rest of the season Barrett settled, the Oline gelled, and Elliot grew into a monster. We beat them pretty soundly at home in all aspects, but to say that the OSU team tonight was equal to whom we beat is just delusional. After that game we went separate ways.
Log in or register to post comments about the Virginia Tech Hokies
Who's saying that? They clearly are a great team, but we beat them on the night we played them and really that's all that matters in regards to Hokie football and OSU. Just because they improved over the season doesn't mean our win is any less meaningful. To say that our win is devalued because they became a better team afterwards is also delusional. We became a better team and beat Cincy in the bowl, does that mean ECU's win over us means less? nope.
Log in or register to post comments about the Virginia Tech Hokies
Do you think for a second that any OSU fan or player actually gives a rat's ass they lost to us in September? I talked to one after the game since I was cheering for whichever team did something good because I didn't care and he thanked me for cheering with him. We told him we were Hokies and he basically said "that's cool, you beat us but we still won it all." We beat them, but that loss was like our JMU loss in 2010 and rallied them to finish the season in perfection while honing every aspect of their team's game.
We coulda, shoulda, woulda. But we had injuries, bad games, and finished just above .500. They won 13 straight.
Log in or register to post comments about the Virginia Tech Hokies
How is this your conclusion from what he stated? He said nothing to the effect of how an OSU fan's opinion really mattered, just to the fact that our win shouldn't be devalued.
Log in or register to post comments about the Virginia Tech Hokies
It's not being devalued. It's just everyone here is overvaluing it and has been since they beat Alabama last week. His post and others in threads were always a sarcastic we should ignore our victory over OSU because they weren't a good team then. Which holds up decently to the stats at hand. They were a team that had a new QB play against a legend at DC who regularly eats new QBs for breakfast, a new Oline, and a beast RB who just got better as the season went on.
OSU undoubtedly got better as the season went on. The team that played the beginning of September was not the same caliber as the one that played the past 4-6 weeks. Every aspect of their game got better because they rallied together and overcame adversity. We beat them in week 2, great. What did we do with that? We lost to ECU, GT, Pitt, Miami, BC, and Wake.
We lost games because of injuries, they didn't. Don't undervalue their win tonight because we beat them months ago.
Log in or register to post comments about the Virginia Tech Hokies
You are missing the point wildly. I have not seen any poster on here claim we coulda, shoulda, woulda been national champions, especially not me. But that doesn't mean that our win doesn't mean a lot for the program and it's course correction. I am not even talking about their win tonight. In fact, I specifically stated I was commenting on our win the night we played them.
OSU got better over the course of the season. Great. So does almost every team. So what? We beat them when we were meant to, and that's the only thing that matters between the two teams. But certainly that something for the Hokies to be proud of. Why shouldn't they be? Because OSU got better? that's not how sports work. You don't think ECU is proud of beating us even though we got better? of course they are, because they brought it to our house and took us down. That's the point.
Log in or register to post comments about the Virginia Tech Hokies
Excuses, excuses....You sound like an OSU apologist.
We kicked their ass, fair and square on the day we played them.
Thats it....thats all. We kicked there ass in the one chance we had to play them.
I don't care who they are now or that they won the national championship.
We have bragging rights until Labor Day.
I don't care if they got better or if they were inexperienced...thats just an excuse. On the day we played we were undoubtedly the better team.
Log in or register to post comments about the Virginia Tech Hokies
We'll have bragging rights long after Labor Day as well.
We'll forever be their blemish on the 2014 season. And it happened in their house on opening night. Nothing will ever take that away from us, including a loss to open next year.
Log in or register to post comments about the Virginia Tech Hokies
I think it's more delusional when Ohio State fans say they weren't very good and got better, because they lost to us...I've watched their games. They look like the same team that lost to us when I watch them. They play the same way IMO. The biggest difference in them that I see is that their O-Line got better. We're the best defense they saw this year and we played our best offensively that game. It really just shows how close we are.
Log in or register to post comments about the Virginia Tech Hokies
I don't think anyone has ever said OSU wasn't a good team, just that they got much better as the season progressed. Apparently thinking that a team got better intrinsically means that at one point they were bad?
Good teams can improve, and that's what OSU did. No one is saying they were bad then.
Log in or register to post comments about the Virginia Tech Hokies
Ohio State's Schedule (all stats current to National Championship Game):
Opponent______
Total Defense__
Scoring Defense__
Yards Allowed__
Points Allowed
Navy
403.8 (70th)
27.3 (70th)
420
34
Virginia Tech
343.8 (21st)
20.2 (14th)
327
21
Kent State
430.3 (88th)
29.0 (81st)
628
66
Cincinnati
439.1 (96th)
27.2 (69th)
710
50
Maryland
436.2 (95th)
30.2 (87th)
533
52
Rutgers
442.8 (98th)
30.2 (87th)
585
56
Penn State
278.7 (2nd)
18.6 (7th)
293
31
Illinois
456.4 (109th)
34.0 (107th)
545
55
Michigan State
315.8 (8th)
21.5 (22nd)
568
49
Minnesota
368.2 (39th)
24.2 (33rd)
489
31
Indiana
433.8 (93rd)
32.8 (101st)
527
42
Michigan
311.3 (7th)
22.4 (28th)
416
42
Wisconsin
294.1 (4th)
20.8 (17th)
558
59
Alabama
328.4 (12th)
18.4 (6th)
537
42
Looking at the first two games, their offense scored about right at the defense's average, and then they exploded over the next four games, all of which were against bottom 40 (for all intents and purposes, bottom 30) Total Defenses. However, two of these games were against G5 teams who were blown out in all of their games against P5 teams (with the exception of VT's victory over Cincy).
Kent State gave up 520 yards and 45 points to UVA the week after the OSU game, and Cincinnati gave up 621 yards and 55 points to Miami two weeks after losing to OSU(they also gave up 610 yards and 41 points to Memphis the week after losing to OSU).
Yardage wise, they weren't demolishing P5 teams until the Michigan State game. Points wise, they were doing very well with at least 13 points above the defense's average.
Something to consider is that the Big 10's defensive statistics were also inflated by the abysmal quality of offenses in the league:
Team______
Total Offense__
Scoring Offense
Ohio State
509.7 (9th)
45.0 (5th)
Michigan State
500.8 (11th)
43.0 (7th)
Wisconsin
468.9 (21st)
34.6 (27th)
Nebraska
452.3 (31st)
37.8 (12th)
Indiana
405.0 (61st)
25.1 (87th)
Iowa
400.1 (63rd)
28.2 (68th)
Rutgers
390.0 (73rd)
26.7 (80th)
Illinois
367.2 (94th)
25.9 (84th)
Minnesota
357.3 (103rd)
28.4 (66th)
Northwestern
353.1 (104th)
23.0 (98th)
Purdue
344.6 (108th)
23.8 (95th)
Maryland
342.1 (109th)
28.5 (65th)
Penn State
335.3 (111th)
20.6 (110th)
Michigan
333.0 (112th)
20.9 (109th)
The average offense in the Big Ten gains 375.7 ypg and 29.4 ppg. To compare with the other conferences:
The top four B10 offenses average 482.9 yards (18th) and 40.1 points per game (9th), and the remaining ten teams average 360.7 yards (101st) and 25.2 points per game (86th). This shows how top heavy the league is, further demonstrating that the defensive rankings in the B10 are inflated by the low quality of the offenses.
So here is the crux of my argument: any improvement by the Ohio State offense immediately following the Virginia Tech game was masked by the fact that the defenses they played over the next four games were flat out terrible. Furthermore, the Big Ten's defensive rankings as a whole are inflated by the offensive ineptitude of the league (to some degree, this has also happened in the ACC, but the ACC is extremely well balanced and not very top heavy). The argument could actually be made that the Ohio State offense was relatively unimproved until the Michigan State game, where they turned the corner and started thrashing elite defenses, and that their gaudy numbers up until that point, excluding the Penn State game, were a direct result of playing very weak teams.
TLDR: While I'm not a statistician, I've done the best I can to give you data that I think demonstrates my argument that Ohio State's offense did improve over the season. Admittedly, you could argue that the Ohio State offense was always good (excluding the Navy and Penn State games), but I think that if you look at the quality of the defenses they faced, it is apparent that they were not the type of teams that would challenge any competent offense, and led to a statistical inflation that masked the improvements the offense made.
Log in or register to post comments about the Virginia Tech Hokies
This is their scoring offense per game during the regular season, normalized against the opponent's average scoring defense for the season. A value of 1.0 means they scored exactly what the opponent gave up on average.
Seriously. Look at this graph and tell me where the improvent over the season occurred?
Log in or register to post comments about the Virginia Tech Hokies
that doesn't make sense at all though. If they were merely "good" and became "great" then it would be an upward trend. But that didn't happen. I would say they were a great team all along and our win was their wake up call. That doesn't mean they weren't a great team playing us we just came up and punched them in the mouth. This is more evident as they immediately became beast after losing to us and kept playing that way all season.
Log in or register to post comments about the Virginia Tech Hokies
This is the "flipped switch" argument that states, essentially, immediately after playing us OSU underwent a dramatic offensive improvement that then held relatively steady throughput the rest of the season. That argument is dubious in my mind. And yet the flipped switch argument is better supported by the data than the argument that is more frequently trotted out on this board that OSU demonstrated gradual offensive improvement over the course of the season. There is zero statistical evidence to support that assertion.
Give what the numbers show, normalized against opposing defenses, the real argumemt is to explain the two outliers, Navy and VT. I think two arguments can be made: the flipped switch and the return to baseline. I've already explained the flipped switch.
The return to baseline would argue that OSU has demonstrated the same consistent level of offensive performance throughout the regular season but some extenuating factor, such as looking past Navy and getting shut down by an unexpected defense against VT, caused lower than expected output in those two games.
You can like the flipped switch or you can like the return to baseline. But by what the data shows, there is essentially no evidence to support the theory of gradual, season-long improvement.
Log in or register to post comments about the Virginia Tech Hokies
That's why I also presented the argument that the four teams they played after Virginia Tech had bad defenses that inflated the statistics, which was further compounded by the poor offenses of the Big Ten artificially bolstering defense's stats.
Since charts would take up too much space for this, I'll simply list out those four teams and the respective Scoring Offense of their opponents:
Kent State went 2-10 in the MAC, playing Ohio (20.5), South Alabama (22.5), Ohio State (45.0), UVA (25.8), Northern Illinois (31.5), Massachusetts (27.3), Army (24.9), Miami (OH) (22.3), Toledo (36.6), Bowling Green (30.0), and Akron (22.6).
Cincinnati went 9-4 in the AAC, playing Toledo (36.6), Miami (OH) (22.3), Ohio State (45.0), Memphis (36.2), Miami (FL) (29.2), SMU (11.1), USF (17.2), Tulane (16.0), ECU (35.8), UConn (15.5), Temple (23.1), Houston (29.8), and Virginia Tech (24.1).
Maryland went 7-6 in the B10, playing JMU (36.0), USF (17.2), WVU (33.5), Syracuse (17.1), Indiana (25.1), Ohio State (45.0), Iowa (28.2), Wisconsin (34.6), Penn State (20.6), Michigan State (43.0), Michigan (20.9), Rutgers (26.7), and Stanford (27.2).
Rutgers went 8-5 in the B10, playing WSU (31.8), Howard (25.4), Penn State (20.6), Navy (31.8), Tulane (16.0), Michigan (20.9), Ohio State (45.0), Nebraska (37.8), Wisconsin (34.6), Indiana (25.1), Michigan State (43.0), Maryland (28.5), and UNC (33.2).
And just for fun, but mostly because I like our defense so much:
Virginia Tech went 7-6 in the ACC, playing William & Mary (25.3), Ohio State (45.0), ECU (35.8), GT (37.9), WMU (33.8), UNC (33.2), Pitt (31.8), Miami (FL) (29.2), BC (26.2), Duke (32.4), Wake Forest (14.8), UVA (25.8), and Cincinnati (34.0).
Represented as a table:
Team_______
Average Offense Faced__
Scoring Defense__
Difference
Kent State
25.75 (86th)
29.0 (81st)
+3.25
Cincinnati
26.30 (83rd)
27.2 (69th)
+0.9
Maryland
28.83 (63rd)
30.2 (87th)
+1.37
Rutgers
30.26 (54th)
30.2 (87th)
-0.06
Virginia Tech
31.17 (51st)
20.2 (14th)
-10.97
Basically, Kent State typically gave up a field goal more than the average offense they faced; Cincinnati, Maryland, and Rutgers had marginal differences; and Virginia Tech held opponents around 11 points lower than their average as a whole.
However, take note of the average offense each team played: both Kent State and Cincinnati played teams with an average offense in the bottom 40. The median scoring offense in the country was Rice, scoring 28.8 points per game, good for 63rd in the country.
Basically what I am trying to point out is that the defenses Ohio State played in the four games following Virginia Tech were bad, which is made even worse by the fact that these defenses were typically playing against sub-par offenses. Rutgers is the only team who played an average above average offense, Maryland played an average average offense, and Cincinnati and Kent State played well below average average offenses.
It is my belief that this would lead to statistical inflation of the defenses, portraying them as better than they actually are. Kent State and Cincinnati absolutely crumbled against P5 teams, and Maryland and Rutgers were poor defenses in general. Furthermore, this inflation would hide the learning curve and would also result in the spike that you see in the graph you posted.
Of course, if we all want to leave this happy we can just take away that Virginia Tech held our opponents eleven points below their average.
Log in or register to post comments about the Virginia Tech Hokies
Ok, here are some charts that show their defense normalized against the median scoring Defense (which is 26.8 ppg by Boise State). I have inverted that line to better demonstrate this in relation to the graph you presented, so that the lower the orange line is the better the defense against the average. The blue line represents the data in the graph you posted earlier.
As you can see, this would mostly support your argument that Ohio State's offense was always good and just lulled through the first two games. However, as I pointed out before, both Kent State and Cincinnati were dismantled by every P5 team they played, with the exception of Cincinnati vs. Virginia Tech.
The next chart takes that into account, removing all non-P5 teams from the average scoring defense and normalizing each team's scoring defense against P5 teams against the average of 26.8.
The giant spike following the Virginia Tech game disappears, and over the next five games the gap between what the defense is expected to give up and what Ohio State scores increases. There were still lulls vs Illinois, @ Minnesota, and vs Indiana, but the overall trend of the offense is upwards.
Of course, we're all Hokies here so I would like to point out that, after only taking into account P5 competition, Virginia Tech was the second best defense they faced, edged out narrowly by Penn State and slightly edging out Alabama. Against all competition, we would have been fourth. We were the only elite defense they did not thrash all season.
If our defense can learn to contain the big play better than they did this season, then they have a chance to return to that 2004-2007 form, when Tech finished no worse than 3rd in scoring defense.
Log in or register to post comments about the Virginia Tech Hokies
Are you seriously using FBS-wide and P5-wide scoring defense averages to try to flatten the slope of OSU's offensive improvement post-VT?
I mean... wow. Just wow.
There's lies, damn lies and statistics, and then there's whatever you're trying to do here.
The only normalizing that needs be done is to compare the points OSU scored against an opponent vs the average number of points that specific opponent gave up on average per game. There is no need to introduce a hypothetical median defense which has no statistical relevance to what OSU should score against a specific opponent. That's just playing with the numbers to try to get the graph to look how you want.
Log in or register to post comments about the Virginia Tech Hokies
Duly noted. Here is the graph that takes what you said into account, using the average points allowed.
The line from your graph is against all competition, the one below is only taking P5 competition into account. My argument has always been that Kent St and Cincinnati benefit statistically from playing weaker competition; they got blown out by all of the P5 teams they played.
If you go purely based on the point averages against all competition, then both of these teams defenses were objectively better than Notre Dame's, which I don't believe at all.
Again, by focusing on points allowed to P5 conferences, the spike in the games following Virginia Tech disappears. There is a clear upward trend over the course of the graph.
I'm no statistician, so cut me a little slack.
Log in or register to post comments about the Virginia Tech Hokies
First and most importantly, I consider this a spirited debate rather than an argument, so if I come off as harsh I apologize. I'm just enjoying the discourse.
Second, the problem I have with only including P5 opponents in normalizing the scoring defense of OSU's opponents is, in the case of Kent State and Cincy, it reduces your sample size to the point that it's too small to be statistically meaningful.
Kent State played Ohio State and LOLUVA. Cincy played Ohio State, Miami and us. With samples that small, the result against OSU has too great an effect on the mean. Kent State's scoring defense against P5 competition is determined in half by the 66 points OSU hung on them. With a sample size of two, you're strongarming the slope of the graph to be lower. Of course the Kent State result is normalized closer to 1.0, because the Ohio State game is responsible for half of the normalized value! It's self-fulfilling prophecy.
However, I also agree with you that it's bunk to assert that Kent State and Cincy have better defenses than Notre Dame, even if the statistics show it, because the results against Kent State and Cincy's opponents and the results against Notre Dame's opponents doesn't share enough data points in common to draw meaningful conclusions. But like I've said, we can't look at Kent State and Cincy's results against P5 opponents because the sample size is too small. So let's just throw all non-P5 competition out the window enentirely
Using all of the above graphs, Ohio State did not show consistent offensive improvement in the regular season from the Maryland game through the Michigan game. Their performance was up and down. So the options we face when looking only at P5 competition are either:
Ohio State's offensive improvement took place entirely against Kent State and Cincinnati.
Ohio State's offensive remained constant, and a mitigating factor affected the VT result.
I'll call 1 the modified flipped switch and 2 is the same return to baseline argument I've been presenting.
The data still doesn't support the idea that Ohio State got better as the season went on. Once they hit conference play, they were as good as they were going to get on offense, from a statistical standpoint. All we really accomplish by disregarding the non-P5 data is to give the flipped switch argument a little more wiggle room, because we're willfully ignoring the Kent State and Cincinnati games.
Log in or register to post comments about the Virginia Tech Hokies
Adjusted for P5, I would say the upward trend continues relatively closely through the Michigan State game before it gets really wonky. However, I do agree that there really aren't enough data points to completely judge Kent State and Cincinnati. Like you said, it does end up being somewhat of a self-fulfilling prophecy, but with Kent State and Cincinnati playing in non-P5 conferences in becomes very difficult to figure out how good their defenses actually are.
In the end, it ends up like most things with college football where there simply isn't enough data to really dig deep and compare something that doesn't happen in conference.
Log in or register to post comments about the Virginia Tech Hokies
In the end, it ends up like most things with college football where there simply isn't enough data to really dig deep and compare something that doesn't happen in conference.
But trying to do so is what keeps the hot stove burning during the offseason.
We could have this conversation until spring training, (We won't, for the sake of other TKP readers.) You've done a rather remarkable job in laying out your argument based on the facts, and though I disagree with your interpretation of the data, it's passionate yet civilized debates like this that make TKP such a great community.
I'm already gonna buy Koastal Kings a drink if we beat Ohio State in Lane. I'll buy you one too.
Log in or register to post comments about the Virginia Tech Hokies
The distribution of the ball through the air was huge in this game. I re-watched it today and the X-factor was Malleck on third downs. Brewer hit him at least three times throughout the game on third down to get the first. He was rock solid. He missed some time throughout the season but it is crucial that we utilize him more next year.
Log in or register to post comments about the Virginia Tech Hokies
I remember listening to Bill and Mike call the game, and when Brewer threw that interception on the 3rd play of the game or whenever it was, I hate to admit it now, but my reaction was something along the lines of "and so it begins." Little did I know what had actually begun. Thus the Brewer INT=win equation was solidified.
Log in or register to post comments about the Virginia Tech Hokies
Comments
Right on, Andy.
It wasn't always pretty but enough guys made plays and we were ready for that game.
True Dat!
Still got my ticket stub. What a great win and a thrilling game.
Perfect game plan on both sides of the ball, with O and D complementing each other all night.
Amen Andy
I was there- it was obvious that Brewer and the offense could move the ball and score when we needed it.
Going into that game I had no idea how we would score much less move the ball. I remember commenting that night "this is what offense looks like against a top ten opponent".
To me it's disingenuous when I hear OSU was inexperienced...as if we weren't!
Amen fellow Hokie. I hate that people always forget that we were just as young and as inexperienced as OSU that night. BTW, we were playin at OSU. On paper, they litreally had every advantage a team could ask for.
tsk, tsk ... facts Andy? come on. You clearly haven't done your research, OSU wasn't a good team then. That's the only reason we beat them.
They were a good team, they just weren't a great team. Over the rest of the season Barrett settled, the Oline gelled, and Elliot grew into a monster. We beat them pretty soundly at home in all aspects, but to say that the OSU team tonight was equal to whom we beat is just delusional. After that game we went separate ways.
Or maybe that we aren't exactly the bad team everyone thinks we are when we aren't dealing with a shitload of injuries....
Who's saying that? They clearly are a great team, but we beat them on the night we played them and really that's all that matters in regards to Hokie football and OSU. Just because they improved over the season doesn't mean our win is any less meaningful. To say that our win is devalued because they became a better team afterwards is also delusional. We became a better team and beat Cincy in the bowl, does that mean ECU's win over us means less? nope.
Do you think for a second that any OSU fan or player actually gives a rat's ass they lost to us in September? I talked to one after the game since I was cheering for whichever team did something good because I didn't care and he thanked me for cheering with him. We told him we were Hokies and he basically said "that's cool, you beat us but we still won it all." We beat them, but that loss was like our JMU loss in 2010 and rallied them to finish the season in perfection while honing every aspect of their team's game.
We coulda, shoulda, woulda. But we had injuries, bad games, and finished just above .500. They won 13 straight.
How is this your conclusion from what he stated? He said nothing to the effect of how an OSU fan's opinion really mattered, just to the fact that our win shouldn't be devalued.
It's not being devalued. It's just everyone here is overvaluing it and has been since they beat Alabama last week. His post and others in threads were always a sarcastic we should ignore our victory over OSU because they weren't a good team then. Which holds up decently to the stats at hand. They were a team that had a new QB play against a legend at DC who regularly eats new QBs for breakfast, a new Oline, and a beast RB who just got better as the season went on.
OSU undoubtedly got better as the season went on. The team that played the beginning of September was not the same caliber as the one that played the past 4-6 weeks. Every aspect of their game got better because they rallied together and overcame adversity. We beat them in week 2, great. What did we do with that? We lost to ECU, GT, Pitt, Miami, BC, and Wake.
We lost games because of injuries, they didn't. Don't undervalue their win tonight because we beat them months ago.
You are missing the point wildly. I have not seen any poster on here claim we coulda, shoulda, woulda been national champions, especially not me. But that doesn't mean that our win doesn't mean a lot for the program and it's course correction. I am not even talking about their win tonight. In fact, I specifically stated I was commenting on our win the night we played them.
OSU got better over the course of the season. Great. So does almost every team. So what? We beat them when we were meant to, and that's the only thing that matters between the two teams. But certainly that something for the Hokies to be proud of. Why shouldn't they be? Because OSU got better? that's not how sports work. You don't think ECU is proud of beating us even though we got better? of course they are, because they brought it to our house and took us down. That's the point.
Excuses, excuses....You sound like an OSU apologist.
We kicked their ass, fair and square on the day we played them.
Thats it....thats all. We kicked there ass in the one chance we had to play them.
I don't care who they are now or that they won the national championship.
We have bragging rights until Labor Day.
I don't care if they got better or if they were inexperienced...thats just an excuse. On the day we played we were undoubtedly the better team.
so much this. well said
I'm... I'm sorry... Were... Were you saying something?
We'll have bragging rights long after Labor Day as well.
We'll forever be their blemish on the 2014 season. And it happened in their house on opening night. Nothing will ever take that away from us, including a loss to open next year.
sorry to nit pic but wasn't it the second week of the year (and thus, not opening night)?
Home opener for the season.
oh
Their first game was at Navy. Thus, it was opening night in the shoe.
I think it's more delusional when Ohio State fans say they weren't very good and got better, because they lost to us...I've watched their games. They look like the same team that lost to us when I watch them. They play the same way IMO. The biggest difference in them that I see is that their O-Line got better. We're the best defense they saw this year and we played our best offensively that game. It really just shows how close we are.
Agree! We had the fastest DLine they ever saw this year coupled with the best D-coordinator in the country. Our O played a good game as well.
IMO we matched up well against OSU and will do so again Labor Day. I'm not guaranteeing a win, but it will be fun either way!
It's delusional to say the roster we would have fielded later in the year could have bested the OSU!
We're only left to wonder what kind of gelling or monsters we would have created with a healthy roster.
I don't think anyone has ever said OSU wasn't a good team, just that they got much better as the season progressed. Apparently thinking that a team got better intrinsically means that at one point they were bad?
Good teams can improve, and that's what OSU did. No one is saying they were bad then.
I'd still like to see some hard statistical data showing an improvement on offense. So far, none has been presented.
Ohio State's Schedule (all stats current to National Championship Game):
Looking at the first two games, their offense scored about right at the defense's average, and then they exploded over the next four games, all of which were against bottom 40 (for all intents and purposes, bottom 30) Total Defenses. However, two of these games were against G5 teams who were blown out in all of their games against P5 teams (with the exception of VT's victory over Cincy).
Kent State gave up 520 yards and 45 points to UVA the week after the OSU game, and Cincinnati gave up 621 yards and 55 points to Miami two weeks after losing to OSU(they also gave up 610 yards and 41 points to Memphis the week after losing to OSU).
Yardage wise, they weren't demolishing P5 teams until the Michigan State game. Points wise, they were doing very well with at least 13 points above the defense's average.
Something to consider is that the Big 10's defensive statistics were also inflated by the abysmal quality of offenses in the league:
The average offense in the Big Ten gains 375.7 ypg and 29.4 ppg. To compare with the other conferences:
P5 Average Offense:
ACC: 392.3 (73rd) 28.4 (66th)
B10: 375.7 (88th) 29.4 (59th)
B12: 441.8 (35th) 32.1 (42nd)
P12: 449.8 (32nd) 33.3 (36th)
SEC: 414.6 (54th) 31.4 (49th)
The top four B10 offenses average 482.9 yards (18th) and 40.1 points per game (9th), and the remaining ten teams average 360.7 yards (101st) and 25.2 points per game (86th). This shows how top heavy the league is, further demonstrating that the defensive rankings in the B10 are inflated by the low quality of the offenses.
So here is the crux of my argument: any improvement by the Ohio State offense immediately following the Virginia Tech game was masked by the fact that the defenses they played over the next four games were flat out terrible. Furthermore, the Big Ten's defensive rankings as a whole are inflated by the offensive ineptitude of the league (to some degree, this has also happened in the ACC, but the ACC is extremely well balanced and not very top heavy). The argument could actually be made that the Ohio State offense was relatively unimproved until the Michigan State game, where they turned the corner and started thrashing elite defenses, and that their gaudy numbers up until that point, excluding the Penn State game, were a direct result of playing very weak teams.
TLDR: While I'm not a statistician, I've done the best I can to give you data that I think demonstrates my argument that Ohio State's offense did improve over the season. Admittedly, you could argue that the Ohio State offense was always good (excluding the Navy and Penn State games), but I think that if you look at the quality of the defenses they faced, it is apparent that they were not the type of teams that would challenge any competent offense, and led to a statistical inflation that masked the improvements the offense made.
Here it is in graphical form:
This is their scoring offense per game during the regular season, normalized against the opponent's average scoring defense for the season. A value of 1.0 means they scored exactly what the opponent gave up on average.
Seriously. Look at this graph and tell me where the improvent over the season occurred?
Considering they spiked up after the VT game and stayed above 1.5 until the Minnesota and Indiana games says that's where it happened.
that doesn't make sense at all though. If they were merely "good" and became "great" then it would be an upward trend. But that didn't happen. I would say they were a great team all along and our win was their wake up call. That doesn't mean they weren't a great team playing us we just came up and punched them in the mouth. This is more evident as they immediately became beast after losing to us and kept playing that way all season.
This is the "flipped switch" argument that states, essentially, immediately after playing us OSU underwent a dramatic offensive improvement that then held relatively steady throughput the rest of the season. That argument is dubious in my mind. And yet the flipped switch argument is better supported by the data than the argument that is more frequently trotted out on this board that OSU demonstrated gradual offensive improvement over the course of the season. There is zero statistical evidence to support that assertion.
Give what the numbers show, normalized against opposing defenses, the real argumemt is to explain the two outliers, Navy and VT. I think two arguments can be made: the flipped switch and the return to baseline. I've already explained the flipped switch.
The return to baseline would argue that OSU has demonstrated the same consistent level of offensive performance throughout the regular season but some extenuating factor, such as looking past Navy and getting shut down by an unexpected defense against VT, caused lower than expected output in those two games.
You can like the flipped switch or you can like the return to baseline. But by what the data shows, there is essentially no evidence to support the theory of gradual, season-long improvement.
That's why I also presented the argument that the four teams they played after Virginia Tech had bad defenses that inflated the statistics, which was further compounded by the poor offenses of the Big Ten artificially bolstering defense's stats.
Since charts would take up too much space for this, I'll simply list out those four teams and the respective Scoring Offense of their opponents:
Kent State went 2-10 in the MAC, playing Ohio (20.5), South Alabama (22.5), Ohio State (45.0), UVA (25.8), Northern Illinois (31.5), Massachusetts (27.3), Army (24.9), Miami (OH) (22.3), Toledo (36.6), Bowling Green (30.0), and Akron (22.6).
Cincinnati went 9-4 in the AAC, playing Toledo (36.6), Miami (OH) (22.3), Ohio State (45.0), Memphis (36.2), Miami (FL) (29.2), SMU (11.1), USF (17.2), Tulane (16.0), ECU (35.8), UConn (15.5), Temple (23.1), Houston (29.8), and Virginia Tech (24.1).
Maryland went 7-6 in the B10, playing JMU (36.0), USF (17.2), WVU (33.5), Syracuse (17.1), Indiana (25.1), Ohio State (45.0), Iowa (28.2), Wisconsin (34.6), Penn State (20.6), Michigan State (43.0), Michigan (20.9), Rutgers (26.7), and Stanford (27.2).
Rutgers went 8-5 in the B10, playing WSU (31.8), Howard (25.4), Penn State (20.6), Navy (31.8), Tulane (16.0), Michigan (20.9), Ohio State (45.0), Nebraska (37.8), Wisconsin (34.6), Indiana (25.1), Michigan State (43.0), Maryland (28.5), and UNC (33.2).
And just for fun, but mostly because I like our defense so much:
Virginia Tech went 7-6 in the ACC, playing William & Mary (25.3), Ohio State (45.0), ECU (35.8), GT (37.9), WMU (33.8), UNC (33.2), Pitt (31.8), Miami (FL) (29.2), BC (26.2), Duke (32.4), Wake Forest (14.8), UVA (25.8), and Cincinnati (34.0).
Represented as a table:
Basically, Kent State typically gave up a field goal more than the average offense they faced; Cincinnati, Maryland, and Rutgers had marginal differences; and Virginia Tech held opponents around 11 points lower than their average as a whole.
However, take note of the average offense each team played: both Kent State and Cincinnati played teams with an average offense in the bottom 40. The median scoring offense in the country was Rice, scoring 28.8 points per game, good for 63rd in the country.
Basically what I am trying to point out is that the defenses Ohio State played in the four games following Virginia Tech were bad, which is made even worse by the fact that these defenses were typically playing against sub-par offenses. Rutgers is the only team who played an average above average offense, Maryland played an average average offense, and Cincinnati and Kent State played well below average average offenses.
It is my belief that this would lead to statistical inflation of the defenses, portraying them as better than they actually are. Kent State and Cincinnati absolutely crumbled against P5 teams, and Maryland and Rutgers were poor defenses in general. Furthermore, this inflation would hide the learning curve and would also result in the spike that you see in the graph you posted.
Of course, if we all want to leave this happy we can just take away that Virginia Tech held our opponents eleven points below their average.
Normalization counteracts inflation.
Ok, here are some charts that show their defense normalized against the median scoring Defense (which is 26.8 ppg by Boise State). I have inverted that line to better demonstrate this in relation to the graph you presented, so that the lower the orange line is the better the defense against the average. The blue line represents the data in the graph you posted earlier.
As you can see, this would mostly support your argument that Ohio State's offense was always good and just lulled through the first two games. However, as I pointed out before, both Kent State and Cincinnati were dismantled by every P5 team they played, with the exception of Cincinnati vs. Virginia Tech.
The next chart takes that into account, removing all non-P5 teams from the average scoring defense and normalizing each team's scoring defense against P5 teams against the average of 26.8.
The giant spike following the Virginia Tech game disappears, and over the next five games the gap between what the defense is expected to give up and what Ohio State scores increases. There were still lulls vs Illinois, @ Minnesota, and vs Indiana, but the overall trend of the offense is upwards.
Of course, we're all Hokies here so I would like to point out that, after only taking into account P5 competition, Virginia Tech was the second best defense they faced, edged out narrowly by Penn State and slightly edging out Alabama. Against all competition, we would have been fourth. We were the only elite defense they did not thrash all season.
If our defense can learn to contain the big play better than they did this season, then they have a chance to return to that 2004-2007 form, when Tech finished no worse than 3rd in scoring defense.
Are you seriously using FBS-wide and P5-wide scoring defense averages to try to flatten the slope of OSU's offensive improvement post-VT?
I mean... wow. Just wow.
There's lies, damn lies and statistics, and then there's whatever you're trying to do here.
The only normalizing that needs be done is to compare the points OSU scored against an opponent vs the average number of points that specific opponent gave up on average per game. There is no need to introduce a hypothetical median defense which has no statistical relevance to what OSU should score against a specific opponent. That's just playing with the numbers to try to get the graph to look how you want.
Duly noted. Here is the graph that takes what you said into account, using the average points allowed.
The line from your graph is against all competition, the one below is only taking P5 competition into account. My argument has always been that Kent St and Cincinnati benefit statistically from playing weaker competition; they got blown out by all of the P5 teams they played.
If you go purely based on the point averages against all competition, then both of these teams defenses were objectively better than Notre Dame's, which I don't believe at all.
Again, by focusing on points allowed to P5 conferences, the spike in the games following Virginia Tech disappears. There is a clear upward trend over the course of the graph.
I'm no statistician, so cut me a little slack.
First and most importantly, I consider this a spirited debate rather than an argument, so if I come off as harsh I apologize. I'm just enjoying the discourse.
Second, the problem I have with only including P5 opponents in normalizing the scoring defense of OSU's opponents is, in the case of Kent State and Cincy, it reduces your sample size to the point that it's too small to be statistically meaningful.
Kent State played Ohio State and LOLUVA. Cincy played Ohio State, Miami and us. With samples that small, the result against OSU has too great an effect on the mean. Kent State's scoring defense against P5 competition is determined in half by the 66 points OSU hung on them. With a sample size of two, you're strongarming the slope of the graph to be lower. Of course the Kent State result is normalized closer to 1.0, because the Ohio State game is responsible for half of the normalized value! It's self-fulfilling prophecy.
However, I also agree with you that it's bunk to assert that Kent State and Cincy have better defenses than Notre Dame, even if the statistics show it, because the results against Kent State and Cincy's opponents and the results against Notre Dame's opponents doesn't share enough data points in common to draw meaningful conclusions. But like I've said, we can't look at Kent State and Cincy's results against P5 opponents because the sample size is too small. So let's just throw all non-P5 competition out the window enentirely
Using all of the above graphs, Ohio State did not show consistent offensive improvement in the regular season from the Maryland game through the Michigan game. Their performance was up and down. So the options we face when looking only at P5 competition are either:
I'll call 1 the modified flipped switch and 2 is the same return to baseline argument I've been presenting.
The data still doesn't support the idea that Ohio State got better as the season went on. Once they hit conference play, they were as good as they were going to get on offense, from a statistical standpoint. All we really accomplish by disregarding the non-P5 data is to give the flipped switch argument a little more wiggle room, because we're willfully ignoring the Kent State and Cincinnati games.
Adjusted for P5, I would say the upward trend continues relatively closely through the Michigan State game before it gets really wonky. However, I do agree that there really aren't enough data points to completely judge Kent State and Cincinnati. Like you said, it does end up being somewhat of a self-fulfilling prophecy, but with Kent State and Cincinnati playing in non-P5 conferences in becomes very difficult to figure out how good their defenses actually are.
In the end, it ends up like most things with college football where there simply isn't enough data to really dig deep and compare something that doesn't happen in conference.
But trying to do so is what keeps the hot stove burning during the offseason.
We could have this conversation until spring training, (We won't, for the sake of other TKP readers.) You've done a rather remarkable job in laying out your argument based on the facts, and though I disagree with your interpretation of the data, it's passionate yet civilized debates like this that make TKP such a great community.
I'm already gonna buy Koastal Kings a drink if we beat Ohio State in Lane. I'll buy you one too.
me 3
We're all gonna have to figure out a meeting place.
Lot 18.
Oh great, this argument again (sarc).
nice article Andy. Well-written, well-timed.
The distribution of the ball through the air was huge in this game. I re-watched it today and the X-factor was Malleck on third downs. Brewer hit him at least three times throughout the game on third down to get the first. He was rock solid. He missed some time throughout the season but it is crucial that we utilize him more next year.
The prospect of having Malleck, Cline and Bucky all healthy at the same time has me (and I'm certain Lefty) salivating.
If Ohio State loses tonight, I'm telling people it is because they were looking ahead to their next game.
I remember listening to Bill and Mike call the game, and when Brewer threw that interception on the 3rd play of the game or whenever it was, I hate to admit it now, but my reaction was something along the lines of "and so it begins." Little did I know what had actually begun. Thus the Brewer INT=win equation was solidified.