CPJ in One-Score Games

[CuseJacket]

Actually, I don't think that they're even close to being the same.

If you go back and look at my threads on ppd vs pwr5, I've both acknowledged the limitations you mention and given reasons for showing that over a whole season they do seem to largely average out. I also try and overcome these limitations to some extent by only including teams that have played at least 3 pwr5 opponents. You may notice that I have not posted any PPDvsPWR5 rankings yet for this year. It's because even though I recognize that strength of schedule can vary wildly even with 3-4 PWR5 opponents, the stat begins to become more reasonable at this point.

The fact that you don't appreciate these differences, well, I don't know what that means.

I appreciate that and understood it prior to my reply. Thank you for doing those stats and for making a meaningful attempt to make them valuable.

I also appreciate that the stats in this thread include a large period of time.

I agree there are flaws with both. The degree to which there are flaws doesn't bother me. I'm not "insulted" by the OPs efforts or what the post says.

 

[Deleted member 2897]

How about Duke in 2016. That shows up as a win for CPJ, but we were blowing them out of the water before we collapsed and had to rally to win at the end.

Boston College and Georgia were both legitimate come backs.

Interesing CPJ went 3-0 in close games last year.

2015 was 1-6. That 1 was a blocked kick return against Florida State. Thats totally random, yet counts as winning a close game. Notre Dame wasn't close until some fluke plays at the end.

2014 Georgia Southern took a miracle comeback. Virginia Tech basically handed us that one.

So many different stories that its hard to construct a trend. I think over the very long run (like 10 years) it would all even out.

I think we are an underdog in many of our games. Are we just barely losing to a lot of tough teams or are we barely losing to cupcakes? That to me matters. And I think that may be the story. Because CPJ wins like 60% of his games. So to me if we lose the majority of the close games but we win 60% of our games overall, we must be beating the easier teams more decisively but just barely losing in the tougher games.

 

[AE 87]

I appreciate that and understood it prior to my reply. Thank you for doing those stats and for making a meaningful attempt to make them valuable.

I also appreciate that the stats in this thread include a large period of time.

I agree there are flaws with both. The degree to which there are flaws doesn't bother me. I'm not "insulted" by the OPs efforts or what the post says.

OK, then I am missing something. Please explain to me the logic of the OP's stat that makes it interesting and comparable.

I'll begin by explaining the logic of the PPDvsPWR5 stat so you know what I'm looking for.

a) The Ultimate Standard of the quality of a football team is its wins vs losses.
b) Wins and Losses are a matter of points scored vs points allowed.
c) While Special Teams (besides FGs) also play a role, and defenses can score points, Team Strength can be approximated by Offensive Pts Scored and Defensive Pts allowed
d) Since different teams play at different rates, calculating these stats per drive as opposed to per game allows better comparison between teams, of offense and defense independently
e) While there is variability among Pwr5 opponents, it is less variability than when non-PWR5 opponents are included; this variability is also mollified by requiring at least 3 pwr5 opponents
f) Ranking Teams by the Difference of their OPPD and their DPPD correlates well with National Rankings like AP, thereby giving some validity to the two stats.
​

So, now, you will explain why you consider it equally meaningful to draw conclusions about the skill of a coach in close games when you

a) don't address issues of team injuries, strength of opposition, etc -- you can use the difference between 2015 and 2016 raised by bwelbo to make your point
b) don't compare CPJ to all the other coaches to know where he would be ranked
c) don't provide a justification for why someone should consider the coach an specialized and independent variable for this particular conclusion
d) don't explain what "length of time" means when the number of 1-score games can vary wildly from 1 year to the next​

Generally, just explain the logic of the stat that makes it meaningful

Thanks

 

[CuseJacket]

OK, then I am missing something. Please explain to me the logic of the OP's stat that makes it interesting and comparable.

I'll begin by explaining the logic of the PPDvsPWR5 stat so you know what I'm looking for.

a) The Ultimate Standard of the quality of a football team is its wins vs losses.
b) Wins and Losses are a matter of points scored vs points allowed.
c) While Special Teams (besides FGs) also play a role, and defenses can score points, Team Strength can be approximated by Offensive Pts Scored and Defensive Pts allowed
d) Since different teams play at different rates, calculating these stats per drive as opposed to per game allows better comparison between teams, of offense and defense independently
e) While there is variability among Pwr5 opponents, it is less variability than when non-PWR5 opponents are included; this variability is also mollified by requiring at least 3 pwr5 opponents
f) Ranking Teams by the Difference of their OPPD and their DPPD correlates well with National Rankings like AP, thereby giving some validity to the two stats.
​

So, now, you will explain why you consider it equally meaningful to draw conclusions about the skill of a coach in close games when you

Actually, I didn't use it to drawing any conclusions about the coach. I recognize that was part of the original intent of the OP and apologies if my posts led you to believe that's why I thought it was interesting. The reason I like data is I can draw my own conclusions and don't need the narratives that sometimes come with it, even if I disagree with the premise or conclusion.

At the simplest level what I thought was interesting is the correlation of winning close games and the years we had success. Another way I looked at that is to say, were we really that much better in our up years than in our down years? Or, another question it made me think about is, what is the difference in those years '08, '09, '14, '16? Talent, chemistry, etc? I don't really have all the answers, but the data is interesting food for thought.

 

[g0lftime]

The old saying is that players win games and coaches lose them. What is important is whether the coaches put the players in situations to win games. The players then have to execute. We were in position to win both the UT and the Miami games. We failed to execute. Our coaches had us in a position to win. A drop here or there, a slip, a tipped pass that happened to fall into the receiver was not the coaches fault. One could argue the bubble screens were not defensed well but the offense made some mistakes too. There were no really blown coverages and that fourth down catch was into double coverage. Its too bad we didn't knock it down.

 

[Sideways]

So after basically not sleeping last night I got to thinking that I've been watching the same story unfold in close games a LOT recently...Tech gets ahead by multiple scores...offense slows from scoring, but also can't chew clock, defense lets game slip away on final drive. It's not just a meltdown on one side or the other, but it's basically a team meltdown every time it happens.

So I got to thinking about what @ilovetheoption said, and what I've said several times myself over the past few years...we just didn't get the 50-50 balls and that's been the difference in wins and losses and good seasons and bad seasons. What I started to think to myself in particular was that CPJ has just not been that good a coach in close games. What's I found funny is that I remember early his tenure thinking he was an incredible in-game coach and that could be the difference between average seasons and good seasons. So I decided to pull the wins and losses in one score games (8 points or less) to see if there was anything to this...and here it goes...

Just for kicks I decided to pull Gailey's numbers too to see how they compare

The top chart shows games, wins, and losses by season in games decided by 8 points, the season win %, cumulative win % (from first season to that given season for the coach), and Win % since (from that season to final season for the coach).

What's crazy is that Gailey never had a season where he had a losing record in one score games. He played two less in his tenure than Johnson did in his first six seasons, so it wasn't all that different. I'll spare you my other thoughts on his coaching abilities since that's not what this is about.

Johnson on the other hand has been pretty erratic in close games. Honestly he was great in his first two seasons and has been either average or just plain bad since with 2016 being the exception. Since 2010 we are 13-24 in games decided by one score or less...35%...that's pretty freakin' bad. Even if you throw out 2015 where we were 1-6 in these games he's only at 40% since 2010.

So what does it mean? Well, as we've all said before, if winning close games is the difference between good seasons and average seasons then we've got some work to do...and it seems like our head coach in particular has had a problem with this since he's been the only constant over those 10 years.

I love the style of football, but CPJ needs to find a way to win these games. I'm not sure why it happens...many could point to not being able to score late, but I think more often than not it's not being able to hang on for a win as I described at the start of the post.

View attachment 2984

Only a Tech fan would put this information in a chart. It is interesting but I am not sure that it means that much. Gailey may have done better in one score games because that defense was so hard to move on in late game situations. They just kept coming and eventually they would get to the quarterback and it was over. One game they played Miami and that guy was sacked like 7 times in the game. It's a wonder he didn't end up in the hospital. Kai Michael Hall hit him so hard one time I thought he would't get up. The answer for CPJ is obvious: we need better athletes on defense especially on the perimeter. Team speed is lacking at certain positions which is exploited by some teams. Better defense and more depth will result in more one score games going our way.

 

[LibertyTurns]

At the simplest level what I thought was interesting is the correlation of winning close games and the years we had success. Another way I looked at that is to say, were we really that much better in our up years than in our down years? Or, another question it made me think about is, what is the difference in those years '08, '09, '14, '16? Talent, chemistry, etc? I don't really have all the answers, but the data is interesting food for thought.

Hard to figure out GT’s psyche.

Take the Orange Bowl for example. At the game you just knew we were going to ram the ball down Miss St’s throat every time we got it. You could feel it. Last year at the cesspool you could see it coming during the 94yd drive. Mutts were worried they were letting us back in the game and in the end we gave it to them in the end in glorious fashion. I mean how did Searcy get into the end zone????

Other times it seems we can’t get out of our own way. We demolished UT. There were probably 8-10 plays had they gone our way were knives in UT’s back. We went o’fer. At the U yesterday we get another 1st down on our last drive & we win. Lucky catch goes the wrong way. Kind of like that Pitt game in 2015 where the deflection changed momentum. Then the Pitt FG guy who completely sucks balls nails the longest FG ever at Grant Field. There’s a ton of examples both ways.

I guess what I’m saying is we’re about 50-50 when it comes to nail biters. That makes us wildly susceptible to a wide range of season record outcomes. If we could figure “it” out and pick off 3 of 4, we’d be a perennial Top 25 team with an occasional chance at a really special season by our current standards. It’s probably no more than having 2-3 “better” plays per game to swing those marginal games in our favor. Could be enhanced pre-game scouting preparation, individual mental preparation, better conditioning, or improved recruiting/more quality depth to get us there or a combination of all. Fact of the matter is right now we are just missing that little extra which typically separates good teams from mediocre ones.

 

[Animal02]

Hard to figure out GT’s psyche.

Take the Orange Bowl for example. At the game you just knew we were going to ram the ball down Miss St’s throat every time we got it. You could feel it. Last year at the cesspool you could see it coming during the 94yd drive. Mutts were worried they were letting us back in the game and in the end we gave it to them in the end in glorious fashion. I mean how did Searcy get into the end zone????

Other times it seems we can’t get out of our own way. We demolished UT. There were probably 8-10 plays had they gone our way were knives in UT’s back. We went o’fer. At the U yesterday we get another 1st down on our last drive & we win. Lucky catch goes the wrong way. Kind of like that Pitt game in 2015 where the deflection changed momentum. Then the Pitt FG guy who completely sucks balls nails the longest FG ever at Grant Field. There’s a ton of examples both ways.

I guess what I’m saying is we’re about 50-50 when it comes to nail biters. That makes us wildly susceptible to a wide range of season record outcomes. If we could figure “it” out and pick off 3 of 4, we’d be a perennial Top 25 team with an occasional chance at a really special season by our current standards. It’s probably no more than having 2-3 “better” plays per game to swing those marginal games in our favor. Could be enhanced pre-game scouting preparation, individual mental preparation, better conditioning, or improved recruiting/more quality depth to get us there or a combination of all. Fact of the matter is right now we are just missing that little extra which typically separates good teams from mediocre ones.

Wasn't the deflection in 2016? We lost both games on last second FG. The 2016 bounced around the cross bars.

 

[Whiskey_Clear]

I expect Tech folks to be better able to find and present meaningful data than this. Why not compare winning percentages for night games, full moon games, games played on the third Saturday of October? About as worthwhile as this discussion. But it's "data" so surely it can tell us something right? Yes, just not anything relevant to coaching ability with these limited variables.

 

[smathis30]

OK, then I am missing something. Please explain to me the logic of the OP's stat that makes it interesting and comparable.

I'll begin by explaining the logic of the PPDvsPWR5 stat so you know what I'm looking for.

a) The Ultimate Standard of the quality of a football team is its wins vs losses.
b) Wins and Losses are a matter of points scored vs points allowed.
c) While Special Teams (besides FGs) also play a role, and defenses can score points, Team Strength can be approximated by Offensive Pts Scored and Defensive Pts allowed
d) Since different teams play at different rates, calculating these stats per drive as opposed to per game allows better comparison between teams, of offense and defense independently
e) While there is variability among Pwr5 opponents, it is less variability than when non-PWR5 opponents are included; this variability is also mollified by requiring at least 3 pwr5 opponents
f) Ranking Teams by the Difference of their OPPD and their DPPD correlates well with National Rankings like AP, thereby giving some validity to the two stats.
​

So, now, you will explain why you consider it equally meaningful to draw conclusions about the skill of a coach in close games when you

a) don't address issues of team injuries, strength of opposition, etc -- you can use the difference between 2015 and 2016 raised by bwelbo to make your point
b) don't compare CPJ to all the other coaches to know where he would be ranked
c) don't provide a justification for why someone should consider the coach an specialized and independent variable for this particular conclusion
d) don't explain what "length of time" means when the number of 1-score games can vary wildly from 1 year to the next​

Generally, just explain the logic of the stat that makes it meaningful

Thanks

I like PPD, but it never really goes into the whole story. Its good at retroactively stating how good teams are but doesn't really indicate how well teams will do in the future. UNC in 2015 was a prime example of that in their loss to South Carolina. Dominated the entire game but had something like 3 red zone turnovers that made them look worse than what they are. Teams (like tripple option teams or air raid teams) that rely on consistent short yardage plays always look better than they should in these type of models due to limiting possesions and higher chances of explosive plays. Its why most advanced stats (FEI, S&P+, Massey, FPI) all use a composite of things like PPD or PPP (points per play) to measure a combination of everything as well as field position and Special teams effects, 3rd down % to measure consistency, and yards per play to measure explosiveness, and TOM as it is far and out the highest correlation for winning. Food for thought.

 

[Whiskey_Clear]

Only a Tech fan would put this information in a chart. It is interesting but I am not sure that it means that much. Gailey may have done better in one score games because that defense was so hard to move on in late game situations. They just kept coming and eventually they would get to the quarterback and it was over. One game they played Miami and that guy was sacked like 7 times in the game. It's a wonder he didn't end up in the hospital. Kai Michael Hall hit him so hard one time I thought he would't get up. The answer for CPJ is obvious: we need better athletes on defense especially on the perimeter. Team speed is lacking at certain positions which is exploited by some teams. Better defense and more depth will result in more one score games going our way.

It would also lead to us winning more blowouts...more wins of all kinds actually. Now we just have to figure out how to recruit a 5 star 300lb DT every year.

 

[FightWinDrink]

It would also lead to us winning more blowouts...more wins of all kinds actually. Now we just have to figure out how to recruit a 5 star 300lb DT every year.

Bioengineering department needs to get their s*** together

 

[AE 87]

I like PPD, but it never really goes into the whole story. Its good at retroactively stating how good teams are but doesn't really indicate how well teams will do in the future. UNC in 2015 was a prime example of that in their loss to South Carolina. Dominated the entire game but had something like 3 red zone turnovers that made them look worse than what they are. Teams (like tripple option teams or air raid teams) that rely on consistent short yardage plays always look better than they should in these type of models due to limiting possesions and higher chances of explosive plays. Its why most advanced stats (FEI, S&P+, Massey, FPI) all use a composite of things like PPD or PPP (points per play) to measure a combination of everything as well as field position and Special teams effects, 3rd down % to measure consistency, and yards per play to measure explosiveness, and TOM as it is far and out the highest correlation for winning. Food for thought.

No statistical measure would predict that UNC SC game. The same goes for our Game vs UT or NC State's game versus SC.

You are right that the advanced statistics are better than raw statistics, and derived raw statistics like my PPDvsPWR5. So, just take it for what it's worth.

 

[Sideways]

It would also lead to us winning more blowouts...more wins of all kinds actually. Now we just have to figure out how to recruit a 5 star 300lb DT every year.

Do you know any?

 

[bke1984]

I understand that not all one score games are created equal, but a one score differential is by far the simplest measure to determine if a game was close or not. Plus, we all know that one crazy play can happen in a football game that results in a score...this could be the difference between a win and a loss in a single score decision (see 2015 FSU).

I already have game data for Tech, so let's take a look at it by coach. My thoughts on the matter is that over time a coach's winning % in single score decisions should level out at 50%...doesn't exactly seem to be the cast though...

So 35% over the past 8 years...does it mean he's a bad coach? No...it's just interesting. Look at the guys who had a percentage like that in the 30's for their careers. I know Alexander won a NC, but go look at his record during the 30's. Bobby Dodd once said he any other coach would have been fired for losing as many games as Alexander did during that time.

So anyhow...let's let the stats go for a minute....

Right now it seems like we either win in blowout fashion, lose a close hard fought game, or win by some freakin' miracle (see 2016 BC, 2016 uGA, 2015 FSU, 2014 uGA, 2014 VT....2014 Georgia Southern...both a blowout and a miracle at the same time )

Every game is different, but we seem to have trouble in tight games right now without having something crazy happen. Our coach also tends to make some really crazy decisions at times in these contests...the Qua TD to win in Athens was amazing, but c'mon...THAT'S what you had to resort to with first and goal within the 5...with Dedrick Mills on your squad? An AB toss pass? We could have run the ball up the middle three straight times and won that game. For another example that didn't turn out so well, see the 2009 uGA game...the decision to throw deep multiple times in a row at the end is still mind boggling.

FYI, I love our coach. He's got balls and he keeps us competitive against the toughest teams on our schedule. A lot of what I've noticed could easily be bad luck...but it's pretty clear to me that it's been a trend. Hopefully this stuff turns around...

 

[Deleted member 2897]

I understand that not all one score games are created equal, but a one score differential is by far the simplest measure to determine if a game was close or not. Plus, we all know that one crazy play can happen in a football game that results in a score...this could be the difference between a win and a loss in a single score decision (see 2015 FSU).

I already have game data for Tech, so let's take a look at it by coach. My thoughts on the matter is that over time a coach's winning % in single score decisions should level out at 50%...doesn't exactly seem to be the cast though...

So 35% over the past 8 years...does it mean he's a bad coach? No...it's just interesting. Look at the guys who had a percentage like that in the 30's for their careers. I know Alexander won a NC, but go look at his record during the 30's. Bobby Dodd once said he any other coach would have been fired for losing as many games as Alexander did during that time.

View attachment 2992


So anyhow...let's let the stats go for a minute....

Right now it seems like we either win in blowout fashion, lose a close hard fought game, or win by some freakin' miracle (see 2016 BC, 2016 uGA, 2015 FSU, 2014 uGA, 2014 VT....2014 Georgia Southern...both a blowout and a miracle at the same time )

Every game is different, but we seem to have trouble in tight games right now without having something crazy happen. Our coach also tends to make some really crazy decisions at times in these contests...the Qua TD to win in Athens was amazing, but c'mon...THAT'S what you had to resort to with first and goal within the 5...with Dedrick Mills on your squad? An AB toss pass? We could have run the ball up the middle three straight times and won that game. For another example that didn't turn out so well, see the 2009 uGA game...the decision to throw deep multiple times in a row at the end is still mind boggling.

FYI, I love our coach. He's got balls and he keeps us competitive against the toughest teams on our schedule. A lot of what I've noticed could easily be bad luck...but it's pretty clear to me that it's been a trend. Hopefully this stuff turns around...

Re: "Voldemort" - I see you, well played.

 

[AE 87]

I understand that not all one score games are created equal, but a one score differential is by far the simplest measure to determine if a game was close or not. Plus, we all know that one crazy play can happen in a football game that results in a score...this could be the difference between a win and a loss in a single score decision (see 2015 FSU).

I already have game data for Tech, so let's take a look at it by coach. My thoughts on the matter is that over time a coach's winning % in single score decisions should level out at 50%...doesn't exactly seem to be the cast though...

So 35% over the past 8 years...does it mean he's a bad coach? No...it's just interesting. Look at the guys who had a percentage like that in the 30's for their careers. I know Alexander won a NC, but go look at his record during the 30's. Bobby Dodd once said he any other coach would have been fired for losing as many games as Alexander did during that time.

View attachment 2992


So anyhow...let's let the stats go for a minute....

Right now it seems like we either win in blowout fashion, lose a close hard fought game, or win by some freakin' miracle (see 2016 BC, 2016 uGA, 2015 FSU, 2014 uGA, 2014 VT....2014 Georgia Southern...both a blowout and a miracle at the same time )

Every game is different, but we seem to have trouble in tight games right now without having something crazy happen. Our coach also tends to make some really crazy decisions at times in these contests...the Qua TD to win in Athens was amazing, but c'mon...THAT'S what you had to resort to with first and goal within the 5...with Dedrick Mills on your squad? An AB toss pass? We could have run the ball up the middle three straight times and won that game. For another example that didn't turn out so well, see the 2009 uGA game...the decision to throw deep multiple times in a row at the end is still mind boggling.

FYI, I love our coach. He's got balls and he keeps us competitive against the toughest teams on our schedule. A lot of what I've noticed could easily be bad luck...but it's pretty clear to me that it's been a trend. Hopefully this stuff turns around...

I still think that your use of this stat is nonsense, but I'm willing to be corrected. I earlier raised some of my concerns in a reply to Cuse, and I invite you to address them if you don't mind. If you don't want to address all of them, then perhaps you can focus on just the last two weeks. Two weeks ago, FSU had a 4 pt lead with about a minute and a half left, and Miami drove the field and scored a TD. This past week, we had a 2 pt lead with about 2 1/2 minutes left, and Miami drove the field and scored a FG. I want you to explain how our game functions in assessing our coach's ability in close games. Please address questions like whether we recruit comparably to FSU and Miami.

In other words, please explain why this same data doesn't show that our coach gets us into 1 score games where another coach would have us losing by multiple scores. Thanks.

Also, when you put "stats aside," it also seems that your characterization is one sided. Why for example, do you lump GT's close wins as "by some freakin miracle," but losses as losing hard fought when UNC's QB scooping and scoring on a fumble and Miami's WR catching a double-rebounded ball or Orwin giving Miami a Safety on a KR somehow fall in as evidence of CPJ not being able to win the hard-fought ones. So, if you could discuss that also, it would help me better understand your position. Thanks

 

[bke1984]

Valid points for sure...the stats aside argument is addressing everyone that replied that they rely on gut feelings. My entire post originated from a gut feeling I had, and I found data that shows we're losing a lot of close games. Is it all Johnson's fault? Hell no...

Honestly, I think Johsnon has done a great job of getting us into closer games...see below...especially with the higher end opponents on our schedule. So yeah, ultimately the players are the ones that have to make plays...and the stuff you pointed out happened on the field. Whether it's the coach's fault or not, we're still losing more close games than we're winning right now. Maybe we need better players, maybe we need to learn how to defend a bubble screen, maybe we need to focus a bit more on fundamentals...or maybe we will just have the ball bounce our way a few more times. Right now it just seems like 2014 aside, the ball has bounced the other teams way a whole lot more times than it has bounced ours

Time to work. Let's just beat Wake (by more than one score, preferably) and move on with the season...

 

[sidewalkGTfan]

Right now it seems like we either win in blowout fashion, lose a close hard fought game, or win by some freakin' miracle (see 2016 BC, 2016 uGA, 2015 FSU, 2014 uGA, 2014 VT....2014 Georgia Southern...both a blowout and a miracle at the same time )

Every game is different, but we seem to have trouble in tight games right now without having something crazy happen. Our coach also tends to make some really crazy decisions at times in these contests...the Qua TD to win in Athens was amazing, but c'mon...THAT'S what you had to resort to with first and goal within the 5...with Dedrick Mills on your squad? An AB toss pass? We could have run the ball up the middle three straight times and won that game. For another example that didn't turn out so well, see the 2009 uGA game...the decision to throw deep multiple times in a row at the end is still mind boggling.

FYI, I love our coach. He's got balls and he keeps us competitive against the toughest teams on our schedule. A lot of what I've noticed could easily be bad luck...but it's pretty clear to me that it's been a trend. Hopefully this stuff turns around...

Ok, some of these games you have as "miracles" that we won, are games that were close throughout most of the game, could've went either way, and our guys made plays to win them. How come you don't list the "miracles" that we lost?

2015 Pitt - they kick a 56? yard FG to win
2016 Pitt - they get a game tying TD on a deflected pass
2017 UT - they recover a late fumble after a long run on a drive that would've iced the game away
2017 Miami - on 4th and 10, they have a receiver catch a deflected pass, while falling to the ground

Why do our own fans want to give the other team all of the credit when they beat us, but whenever we win a tight game, it's random luck?

 

[sidewalkGTfan]

Everyone in this thread should read this link. It's a year old but you'll be surprised where the might Nick Saban is on the list.

https://www.cbssports.com/college-f...ootball-coaches-in-close-games-entering-2016/