Stats First PPDvPwr5 Ranking

[AE 87]

Okay, just for review, I calculate points/drive against pwr5 competition.
Points= 7pts for each rushing and passing TD and 3pts for every made FG
Drives = sum of all rushing TDs, passing TDs, FG atts, Punts, Failed 4th down, and lost Turnovers.

So I don't count scores by special teams nor defenses. I don't count 2pt conversions or missed p.a.t. of either kind. Also I don't include safeties in points or in drives. It's not perfect; for example, a lost fumble in a special teams run-back would count as an extra drive, but it works pretty well.

Ranking is based on subtracting the ppd allowed by D from the ppd scored by O for a Differential Points Per Drive. Typically, I only rank teams that have played at least 3 pwr 5 opp's, but I reduce that to at least 2 for this post-week-6 ranking. I've provided the AP top 25 as comparison.

Obviously, my stat like WashSt's last 3 weeks of big wins vs Idaho, Oregon, and Stanford, doesn't see the losses to Eastern Washington and Boise St (which were pretty close). It really likes Ohio State and Michigan, but both of those teams capitalized big on Rutgers being a Big10 school now.

GT is currently #36, #39 Offense and #35 Defense.

 

[bke1984]

I think you should put some calculation in to account for blowouts at some point. i.e. maybe don't count all drives in a game where a team keeps scoring after being up 28 points...so if it's 28-0 and then ends up 63-0, you only count the first four drives...

Probably not the easiest adjustment to make if you're using Excel

 

[AE 87]

I think you should put some calculation in to account for blowouts at some point. i.e. maybe don't count all drives in a game where a team keeps scoring after being up 28 points...so if it's 28-0 and then ends up 63-0, you only count the first four drives...

Probably not the easiest adjustment to make if you're using Excel

Yeah, I'm just using excel and cfbstats free data tables. It's a raw stat, as described.

 

[CuseJacket]

So, if the defense stops a drive with an INT but then turns that same play into a defensive safety, no points/no drive for the offense? Or no points/1 drive? Or something else?

I ask half-tic but was actually wondering at the time how that scenario factors into ppd stats.

 

[AE 87]

So, if the defense stops a drive with an INT but then turns that same play into a defensive safety, no points/no drive for the offense? Or no points/1 drive? Or something else?

I ask half-tic but was actually wondering at the time how that scenario factors into ppd stats.

Team that threw interception had a drive ending in interception. No points 1 drive (for initial offense).

If interceptor fumbles, and it's recovered and run back for TD, then each team gets a drive but no points are credited for stat.

 

[4shotB]

GT is currently #36, #39 Offense and #35 Defense.

thanks for taking the time. Couple of takeaways....1) overall we aren't as bad as most of us would have guessed (imo). 2) defense gets crucified way out of proportion to the O but your data (thank you again!!) doesn't support that. disclaimer: not that our D doesn't have issues and 3) given this data and a win against Pitt we might have a different perspective on this team and much less wailing and gnashing of teeth. Go Jackets.

 

[33jacket]

Early in the season still. So alot can still move. It would be nice to put the records of the teams in a column to see if this is truly a good metric that helps GENERALLY correlate to wins and losses and overall team qualty/performance

 

[AE 87]

Early in the season still. So alot can still move. It would be nice to put the records of the teams in a column to see if this is truly a good metric that helps GENERALLY correlate to wins and losses and overall team qualty/performance

That's not available from the tables I use. Its also a different way of ranking.

For example, WashSt lost to EWash and BoiseSt by 3 each. Then they beat Idaho by 50, Oregon by 26 and Stanford by 22.

What tells more about them as a team, their 3-2 record or their dominance of 2 Pac12 teams? It seems to me that it's a basis for conversations not one def more true than other.

For general comparison, I think the AP ranking is pretty good.

That being said, it is early, and differences in schedule influence early unit rankings more.

 

[33jacket]

That's not available from the tables I use. Its also a different way of ranking.

For example, WashSt lost to EWash and BoiseSt by 3 each. Then they beat Idaho by 50, Oregon by 26 and Stanford by 22.

What tells more about them as a team, their 3-2 record or their dominance of 2 Pac12 teams? It seems to me that it's a basis for conversations not one def more true than other.

For general comparison, I think the AP ranking is pretty good.

That being said, it is early, and differences in schedule influence early unit rankings more.

I hear you. But its why i used the term general. Wsu is odd. But probably shows u how bad oregon is and flawed stanford is too. At the length of a season, not after 5 games, i bet things correlate. Which is why insaid its early. But if u do this ranking and put the records next to it it would be interesting. Maybe this should be all games not just vs pwr 5? Yes it scews early data, but over a season and everyone has cupcakes it normalizes alot

 

[AE 87]

I hear you. But its why i used the term general. Wsu is odd. But probably shows u how bad oregon is and flawed stanford is too. At the length of a season, not after 5 games, i bet things correlate. Which is why insaid its early. But if u do this ranking and put the records next to it it would be interesting. Maybe this should be all games not just vs pwr 5? Yes it scews early data, but over a season and everyone has cupcakes it normalizes alot

OK, it seems you have a fundamental misunderstanding of sports, stats, and conversations about them.

1) None are perfect. Get this, sometimes a worse team beats a better team based on their records against shared opponents.

2) As a result, there's no way to prove wrong a person who thinks they are always right. They can always say that maybe Oregon and Stanford are really bad or whatever.

3) Obviously, since any ranking will have its flaws, they will all be less accurate with less info. Would Washington have the same stats as now if they played FSU's schedule? Would Mich or Ohio State? Probably not. Would they have the same record? That's the debate/conversation.

4) With the addition of more games, differences in pwr5 scheds balances out typically, but other games skew the stats, esp when you have teams that play more fcs or G5 opponents.

 

[33jacket]

OK, it seems you have a fundamental misunderstanding of sports, stats, and conversations about them.

1) None are perfect. Get this, sometimes a worse team beats a better team based on their records against shared opponents.

2) As a result, there's no way to prove wrong a person who thinks they are always right. They can always say that maybe Oregon and Stanford are really bad or whatever.

3) Obviously, since any ranking will have its flaws, they will all be less accurate with less info. Would Washington have the same stats as now if they played FSU's schedule? Would Mich or Ohio State? Probably not. Would they have the same record? That's the debate/conversation.

4) With the addition of more games, differences in pwr5 scheds balances out typically, but other games skew the stats, esp when you have teams that play more fcs or G5 opponents.

thanks for insulting me. But I am pretty smart guy. For this big of an in-depth analysis on simply adding a column next to the teams with their win loss record I am wondering where the fundamental misunderstanding is? The original response to the original question could have been this simple:

the response could be no. or ok. it really is that easy. I don't need any explanation on why you don't include the win loss record next to the teams. Zero. I don't give a **** why you don't include it. I was just asking if you could since it would be interesting to see the records next to all these rankings...I think it would be a good view one can choose to correlate or not.

if you choose not to. Just say nah, I don't want to. Or nah I don't have the space. Ok...cool..thx for considering.

this reminds me of the explanations I get from the engineers that work for me always so interweaved and complicated...when all I was asking was just to consider, which is like ok or nah

this has gotten off topic. My lord. So is it yes or no? You won't add the w/l record next to it? thats fine...just was asking. For the record, good stats when considering multiple factors (o, d etc) will nicely correlate to win/loss. That is my guess....so it would be good to see over a year.

 

[AE 87]

thanks for insulting me. But I am pretty smart guy. For this big of an in-depth analysis on simply adding a column next to the teams with their win loss record I am wondering where the fundamental misunderstanding is? The original response to the original question could have been this simple:

the response could be no. or ok. it really is that easy. I don't need any explanation on why you don't include the win loss record next to the teams. Zero. I don't give a **** why you don't include it. I was just asking if you could since it would be interesting to see the records next to all these rankings...I think it would be a good view one can choose to correlate or not.

if you choose not to. Just say nah, I don't want to. Or nah I don't have the space. Ok...cool..thx for considering.

this reminds me of the explanations I get from the engineers that work for me always so interweaved and complicated...when all I was asking was just to consider, which is like ok or nah

this has gotten off topic. My lord. So is it yes or no? You won't add the w/l record next to it? thats fine...just was asking. For the record, good stats when considering multiple factors (o, d etc) will nicely correlate to win/loss. That is my guess....so it would be good to see over a year.

In my first reply, I started with this:

That's not available from the tables I use.

I wasn't trying to say that you are not a smart guy, but I did assume that you had understood, "That's not available from the tables I use" as explaining that I cannot include it. I then explained why I don't think it served your suggested purpose of seeing ...

if this is truly a good metric that helps GENERALLY correlate to wins and losses and overall team qualty/performance

In other words, I was saying I can't simply add win-loss because it's not in my source data, but I was also rejecting the premise of why you wanted me to do it. That's how conversations take place between thoughtful people. I also recognized the "GENERALLY" aspect of your post and pointed to the AP top 25 as also providing such a comparison.

So, when your reply seemed to simply ignore that I was actually being responsive to your post, and picked on the particulars of my example, I took it as a rejection both of the value of my ranking without it and of my argument against W-L as serving a better comparison than AP. Indeed, when you repeated your suggestion for including W-L, when I already told you that data's not available to me, I thought you were continuing the suggestion that it is a more accurate, rather than different, way of ranking teams. So, I responded to your post as actually being a reply to my post. You said, "I hear you," so, I interpreted what follows as a response to the substance of my post. I responded to you accordingly (I also remember you saying that you like to troll with most of your posts in this forum).

Now, that being said, with this most recent post, it's clear that you ignored or misunderstood that I answered your question directly with the first sentence of my reply, "That's not available from the tables I use." It's also clear that rather than responding to the substance of my reply and everything I said about different ways of ranking and that I used AP ranking as a general comparison, your post was just saying/defending why you still think W-L would be good to see. It may have been more helpful if you began, "I don't care what you say," rather than, "I hear you," if that's what you mean.

FWIW, in previous years, I've included all games, and I've included teams that play only 1 or 2 PWR5/BCS AQ games. In my opinion, the differences between Pwr5 schedules balances out over the course of the season, but not the other FBS and FCS. A good G5 team that dominates G5 competition is not necessarily comparable to a good Pwr5 team that dominates its competition. Now, I don't think that the advantages of a good Pwr5 team over an extra 1-3 G5 or FCS do average out over the course of the season as much as you think; however, the bigger problem arises from the stats of the G5 teams who play only, or almost only, G5 or lesser opponents.

Again, it's not a perfect stat, but none are.

 

[33jacket]

In my first reply, I started with this:


I wasn't trying to say that you are not a smart guy, but I did assume that you had understood, "That's not available from the tables I use" as explaining that I cannot include it. I then explained why I don't think it served your suggested purpose of seeing ...


In other words, I was saying I can't simply add win-loss because it's not in my source data, but I was also rejecting the premise of why you wanted me to do it. That's how conversations take place between thoughtful people. I also recognized the "GENERALLY" aspect of your post and pointed to the AP top 25 as also providing such a comparison.

So, when your reply seemed to simply ignore that I was actually being responsive to your post, and picked on the particulars of my example, I took it as a rejection both of the value of my ranking without it and of my argument against W-L as serving a better comparison than AP. Indeed, when you repeated your suggestion for including W-L, when I already told you that data's not available to me, I thought you were continuing the suggestion that it is a more accurate, rather than different, way of ranking teams. So, I responded to your post as actually being a reply to my post. You said, "I hear you," so, I interpreted what follows as a response to the substance of my post. I responded to you accordingly (I also remember you saying that you like to troll with most of your posts in this forum).

Now, that being said, with this most recent post, it's clear that you ignored or misunderstood that I answered your question directly with the first sentence of my reply, "That's not available from the tables I use." It's also clear that rather than responding to the substance of my reply and everything I said about different ways of ranking and that I used AP ranking as a general comparison, your post was just saying/defending why you still think W-L would be good to see. It may have been more helpful if you began, "I don't care what you say," rather than, "I hear you," if that's what you mean.

FWIW, in previous years, I've included all games, and I've included teams that play only 1 or 2 PWR5/BCS AQ games. In my opinion, the differences between Pwr5 schedules balances out over the course of the season, but not the other FBS and FCS. A good G5 team that dominates G5 competition is not necessarily comparable to a good Pwr5 team that dominates its competition. Now, I don't think that the advantages of a good Pwr5 team over an extra 1-3 G5 or FCS do average out over the course of the season as much as you think; however, the bigger problem arises from the stats of the G5 teams who play only, or almost only, G5 or lesser opponents.

Again, it's not a perfect stat, but none are.

I dont know dude. I have had a rough morning. So excuse my posts here....i think your look is one of the better ones so why i was trying to add to it