Who should be Number 2?

[Razor]

Originally Posted by keylime_5

of course it was his "average" but that one 30 yard run skews it a bit. He was doing badly in that game outside of one good run. 4.7 isn't great in college football anyway, it's decent but it's not outstanding, especially considering one run bumped it up a whole lot.

Chris Johnson had a 5.6 ypc last season. But if I were to follow your logic CJ is a terrible player because I have to disregard every big run he makes? Come on...

[iowatreat54]

The lack of basic statistical understanding when it comes to sports and fans is pretty entertaining sometimes. (not just here, in general)

Yes, James averaged 4.7 ypc against OSU, but just because it's called "Average Yards per Carry" doesn't mean it's automatically representative of performance, or even average performance. If you have 14 similar outputs, and one ridiculously high/low output, the one that is not like the others is a statistical outlier. Does it count? Absolutely, why shouldn't it? But it skews perceptions on the data set as a whole. To say James averaged 4.7 ypc would indeed be correct, but if you wanted to represent his true performance based on a statistical average, you would have to go more indepth into statistical formulas that weighs each out put separately and accounts for outliers, rather than saying each of the 15 runs were equal.

I suck at stats, so I have no idea how to do that.

Other than that, James is pretty ridiculous and electrifying, but to say James >>>> OSU because he averaged 4.7 ypc is dumb. I mean, even without all the boring stuff I just wrote about, 4.7 ypc in college for one game is not great, and actually pretty average.

[Razor]

Originally Posted by iowatreat54

The lack of basic statistical understanding when it comes to sports and fans is pretty entertaining sometimes. (not just here, in general)

Yes, James averaged 4.7 ypc against OSU, but just because it's called "Average Yards per Carry" doesn't mean it's automatically representative of performance, or even average performance. If you have 14 similar outputs, and one ridiculously high/low output, the one that is not like the others is a statistical outlier. Does it count? Absolutely, why shouldn't it? But it skews perceptions on the data set as a whole. To say James averaged 4.7 ypc would indeed be correct, but if you wanted to represent his true performance based on a statistical average, you would have to go more indepth into statistical formulas that weighs each out put separately and accounts for outliers, rather than saying each of the 15 runs were equal.

I suck at stats, so I have no idea how to do that.

You can't put weights on statistical outliers, at least not something I've ever done in my work with statistics. Basically, iirc, you can't really account for outliers without creating a bias but perhaps somebody else knows something about this. I agree that a 30 yard run is an outlier, but in football you can't really say that it's an anomaly. To go back to the CJ example, one would have to account for all of his 22 runs of 20+ yards and his 7 runs of 40+ yards last year to determine if he had a great season? No. And 4.7 ypc is pretty good against a big, strong OSU DL. And his 6.9 ypc for the season is everything but average, so I stand by my initial statement. And Oregon should be ranked #2.

Edit:

Originally Posted by iowatreat54

Other than that, James is pretty ridiculous and electrifying, but to say James >>>> OSU because he averaged 4.7 ypc is dumb. I mean, even without all the boring stuff I just wrote about, 4.7 ypc in college for one game is not great, and actually pretty average.

And if you go back and look at my initial post you'll find that I didn't reference the OSU game, another poster did. I meant this year. But if it makes you feel better about your sorry existence to call somebody you don't know dumb then go ahead...

[keylime_5]

Oregon has given up like 1000+ yards and 62 points the last two weeks and they should move up in the rankings? If they play a team with a front seven like Alabama's or Ohio State's they won't put up those gaudy offensive stats. I just don't follow the logic and they won't and shouldn't get in the NC game over OSU if both go undefeated I am certain of that.

[iowatreat54]

Originally Posted by Razor

You can't put weights on statistical outliers, at least not something I've ever done in my work with statistics. Basically, iirc, you can't really account for outliers without creating a bias but perhaps somebody else knows something about this. I agree that a 30 yard run is an outlier, but in football you can't really say that it's an anomaly. To go back to the CJ example, one would have to account for all of his 22 runs of 20+ yards and his 7 runs of 40+ yards last year to determine if he had a great season? No. And 4.7 ypc is pretty good against a big, strong OSU DL. And his 6.9 ypc for the season is everything but average, so I stand by my initial statement. And Oregon should be ranked #2.

Edit:


And if you go back and look at my initial post you'll find that I didn't reference the OSU game, another poster did. I meant this year. But if it makes you feel better about your sorry existence to call somebody you don't know dumb then go ahead...

First and foremost, your original post referencing James compared to OSU was:



That's it. There is no explaination or note that you indeed meant for this season only. Furthermore, if you want to compare James to OSU's defense, the most logical comparison would be from their only meeting. If you want to say this year only, then that obviously doesn't work, but then your argument is entirely baseless and opinionated, and thus not a good argument at all.

Additionally, I never called you dumb, so get don't get your panties in a bunch. I said that saying James >>>> OSU is dumb, because it is. It's a dumb statement and argument based on little to no substance, and what little substance there is would actually show James is slightly above average at best compared to OSU's D.

Second, I understand that you can't overanalyze a players every single play in order to show the type of player he is. It would be tedious, and not the best indicator of him overall. However, for a one game basis, it is entirely acceptable to analyze performance based on each play, especially when it's something like carries which is usually around 20 per game. That's not a lot. Further, I didn't say it was the best way to do it, only that it would be more accurate than using simple averages to determine a single perceived performance. Using simple averages is good when you have a large sample size because it tends to follow a norm, but for a single game you can't assume it is the norm.

Finally, I said 4.7 ypc is pretty average for a normal game at RB. It obviously becomes more impressive depending on the strength of the opponent (OSU for example), however, 4.7 ypc even vs. OSU is still good, not great, as was contested in this thread. I believe someone said James faired pretty well against OSU with a 4.7 ypc, which is true, but it's not like he had a great game.

[JoeJoeBrown]

Originally Posted by iowatreat54

First and foremost, your original post referencing James compared to OSU was:



That's it. There is no explaination or note that you indeed meant for this season only. Furthermore, if you want to compare James to OSU's defense, the most logical comparison would be from their only meeting. If you want to say this year only, then that obviously doesn't work, but then your argument is entirely baseless and opinionated, and thus not a good argument at all.

Additionally, I never called you dumb, so get don't get your panties in a bunch. I said that saying James >>>> OSU is dumb, because it is. It's a dumb statement and argument based on little to no substance, and what little substance there is would actually show James is slightly above average at best compared to OSU's D.

Second, I understand that you can't overanalyze a players every single play in order to show the type of player he is. It would be tedious, and not the best indicator of him overall. However, for a one game basis, it is entirely acceptable to analyze performance based on each play, especially when it's something like carries which is usually around 20 per game. That's not a lot. Further, I didn't say it was the best way to do it, only that it would be more accurate than using simple averages to determine a single perceived performance. Using simple averages is good when you have a large sample size because it tends to follow a norm, but for a single game you can't assume it is the norm.

Finally, I said 4.7 ypc is pretty average for a normal game at RB. It obviously becomes more impressive depending on the strength of the opponent (OSU for example), however, 4.7 ypc even vs. OSU is still good, not great, as was contested in this thread. I believe someone said James faired pretty well against OSU with a 4.7 ypc, which is true, but it's not like he had a great game.

All well put. There are many ways to measure impact. Statistics can be manipulated many different ways. For example, one could try and measure how many first downs he picked up in short yardage situations. How often did he get stuffed? Etc...

The fact is, he is a great running back, and he is exciting to watch. I just took umbrage with Razor's post that he is uber-greater than OSU. The only recent measuring point is a game that occurred 6 games ago, with teams that haven't had a lot of turnover.

A game which OSU's D dominated. A game in which OSU's offense stayed on the field for 40+ minutes. Boring ball control offense and stiffling defense. The kind of offense where Tressel settles for 3 instead of going balls out to get 7 (I don't like it, as it's lost more games than it should have by keeping games close and putting way too much pressure on the D).


Look at the drive chart. That's ownage.

[iowatreat54]

Originally Posted by JoeJoeBrown

All well put. There are many ways to measure impact. Statistics can be manipulated many different ways. For example, one could try and measure how many first downs he picked up in short yardage situations. How often did he get stuffed? Etc...

The fact is, he is a great running back, and he is exciting to watch. I just took umbrage with Razor's post that he is uber-greater than OSU. The only recent measuring point is a game that occurred 6 games ago, with teams that haven't had a lot of turnover.

A game which OSU's D dominated. A game in which OSU's offense stayed on the field for 40+ minutes. Boring ball control offense and stiffling defense. The kind of offense where Tressel settles for 3 instead of going balls out to get 7 (I don't like it, as it's lost more games than it should have by keeping games close and putting way too much pressure on the D).


Look at the drive chart. That's ownage.

This was mainly my point. I didn't mean to offend Razor or anyone, I just meant to point out that it was a dumb argument to say James was so much better than OSU, whether it was based on overall, one game, or this year. It's just a dumb argument.

Also that using ypc isn't the end all be all of evaluating performance, or even that indicative when looking at a small sample size such as one game or 15 carries. A simple average such as ypc can be easily construed if there is a low number of examples because outliers have such a high impact on the outcome. That is why you need to account for them in small sample sizes, and not lump them in as equal to all others.

And again, I think James is a great college RB and is extremely exciting, but he had a decent game at best vs. OSU (in terms of running games vs. OSU's D), and was actually pretty much shut down save for one fluke run. And yes, it was a fluke in relation to his overall game that day.

[MichaelJordanEberle (sabf)]

Looking at Lamichael James' game against OSU and Chris Johnson's 2000 yard season ARE NOT THE SAME THING.

When you're looking at one game where he had one semi-long run that is a major outlier from everything else, you get the best indication of his actual effectiveness.
When you look at a whole season, where CJ proved time and time again he could break long runs, it makes no sense to discount all of his long runs. No one has ever suggested that. It's a case of MULTIPLE long runs vs. ONE SINGLE CARRY of 30 yards.

[iowatreat54]

Thank you Christopher for basically saying what I wanted to say without sounding as dumb or confusing as I did. You are an inspiration.

[YAYareaRB]

NO Boise option? I'm gone

[Giantsfan1080]

Yeah but last year James was known to break a huge run in every single game he played in. His long runs per game were 17, 27, 45, 36, 26, 49, 56, 33, 60, 51, 22, and 52. Obviously like Johnson breaking a long run during the course of a game is part of James' track record.

[bearsfan_51]

Originally Posted by YAYareaRB

NO Boise option? I'm gone

You people are out of your minds if you think Boise State would get in before an undefeated Ohio State, Oregon, Oklahoma, Nebraska, Arizona, or ANY undefeated team from the SEC.

[JoeJoeBrown]

Originally Posted by Giantsfan1080

Yeah but last year James was known to break a huge run in every single game he played in. His long runs per game were 17, 27, 45, 36, 26, 49, 56, 33, 60, 51, 22, and 52. Obviously like Johnson breaking a long run during the course of a game is part of James' track record.

Again, 70 yards against the Buckeyes. 42% of which were on one single run.

That's a pretty mediocre game. He did not come close to owning the Buckeyes as Razor asserted.

That's all that everyone (and it looks like people who don't like the Bucks are included in this) is trying to explain.

Personally, I would love to have him start for OSU, but the statement was ridiculous.

[superman]

pretty close voting

gonna give it to oregon though

[soybean]

Originally Posted by superman

pretty close voting

gonna give it to oregon though

awesome sig.

[Giantsfan1080]

I was commenting towards Chris saying that there's a previous pattern for James to have a long run in a game that increases his rushing average which is similar to Johnson. Yes it was only one game but James was known to do that all of last year so it does matter when looking at his game against Ohio State. I don't he dominated but he did have a very good game you can't just take out the long run.

[Sniper]

Originally Posted by JoeJoeBrown

Again, 70 yards against the Buckeyes. 42% of which were on one single run.

47 percent of Terrelle Pryor's passing yards against Miami came on two plays. One was a shovel pass to Dan Herron and the other a bomb to DeVier Posey, yet all the Ohio State fans were using this to boost Pryor's Heisman resume. See, we can do this both ways.

[P-L]

Originally Posted by Giantsfan1080

I don't he dominated but he did have a very good game you can't just take out the long run.

You can't ignore it, but you have to look at it logically. The guy carried the ball 15 times and was a total non-factor on 14 of those carries. He did not have a good game because his running hurt Oregon more than he helped them. You can't just look at his average and pretend that one successful run makes up for 14 [mostly] unsuccessful ones.

As for the poll question, I don't know yet. I have been more impressed with Oregon this year than I have Ohio State. With that said, I think Ohio State would beat them head-to-head.

[Giantsfan1080]

Yes you're right PL but that's part of James running style. In previous games he has a track record of having a similar statistical pattern of having 1 very long run per game. If this was a one time thing I'd say your right but it wasn't.

[JoeJoeBrown]

Originally Posted by Sniper

47 percent of Terrelle Pryor's passing yards against Miami came on two plays. One was a shovel pass to Dan Herron and the other a bomb to DeVier Posey, yet all the Ohio State fans were using this to boost Pryor's Heisman resume. See, we can do this both ways.

Which is fine, I don't think Pryor has any business being in the Heisman chase. Also, you are ignoring total performance in the game (70 yds = subpar if you are going to say you are owning someone in a loss vs 233 yds = average in a win.)

You must be fun at parties.

[JoeJoeBrown]

Originally Posted by P-L

You can't ignore it, but you have to look at it logically. The guy carried the ball 15 times and was a total non-factor on 14 of those carries. He did not have a good game because his running hurt Oregon more than he helped them. You can't just look at his average and pretend that one successful run makes up for 14 [mostly] unsuccessful ones.

As for the poll question, I don't know yet. I have been more impressed with Oregon this year than I have Ohio State. With that said, I think Ohio State would beat them head-to-head.

Well reasoned and therein lies the problem and joy of CFB: Too many teams, nowhere near enough direct competition, and the transitive property just doesn't work in common opponent comparisons. Too many variables.

[P-L]

Originally Posted by Giantsfan1080

Yes you're right PL but that's part of James running style. In previous games he has a track record of having a similar statistical pattern of having 1 very long run per game. If this was a one time thing I'd say your right but it wasn't.

If you take LaMichael James' longest run from each regular season game last season and separate them from the rest he still averaged a very respectable 4.9 yards per carry while his long runs averaged out to be 39.5 yards per carry. There is a huge difference between averaging about 5 yards a carry then breaking a 40 yarder and averaging less than 3 yards a carry then breaking a 30 yarder. His track record says that he is a very consistent running back who can make a huge play once or twice per game. Against Ohio State, he looked like a poor running back who ended up making one big play.

[keylime_5]

Originally Posted by Sniper

47 percent of Terrelle Pryor's passing yards against Miami came on two plays. One was a shovel pass to Dan Herron and the other a bomb to DeVier Posey, yet all the Ohio State fans were using this to boost Pryor's Heisman resume. See, we can do this both ways.

yes but outside of those 2 plays Pryor still had a great day against Miami. Without those 2 big plays he still had 100+ rushing yards and 2 touchdowns and no turnovers and he did pass for like 230 yards in that game or something like that, so taking away 47% of 230 still leaves over 100 yards passing, on a rainy day to boot.
Also, no one going around posting Terrelle Pryor>>>>>>>Miami.

Not quite the same as having 40 yards on 14 carries plus a 30 yard run. Even if you don't take away the 30 yard run like everyone is saying, 70 yards on 15 carries is just decent, not great.

[iowatreat54]

Originally Posted by JoeJoeBrown

Which is fine, I don't think Pryor has any business being in the Heisman chase. Also, you are ignoring total performance in the game (70 yds = subpar if you are going to say you are owning someone in a loss vs 233 yds = average in a win.)

You must be fun at parties.

Sniper is a lot of fun at parties....pants parties. At my pants parties. Sex.

Since I've yet to contribute as to the original topic, here it goes.

Right now, I think OSU really hasn't done enough to deserve to be replaced at #2. Sure, they struggled vs. an underwhelming Illinois team, but imo the game was never in doubt. That said, Oregon has looked better overall, and their win over Stanford is better than any win OSU has, imo.

So, you have the one team who is pretty boring but built on a great D that allows them to have off days on offense and still feel comfortable. Or, you have the seemingly unstoppable offense that will score on anyone, anytime, but also let any team score on them anytime.

So my choice: The ESU Timberwolves.

[lowlife]

Since Oregon has Casey Matthews they automatically win out in every tiebreaker.

You can't discount the Matthews factor.
Sackmaster.