Lindy's Five Essential Websites (Non-Major Media) for 2013
[+] Team Summaries

Thursday, November 1, 2012

Kansas State's offense is REALLY good

I listen each week to The Solid Verbal's preview podcast. They do a great job covering college football as very well-informed fans, which is rare. But even when drawing on that wealth of knowledge they have accumulated over the years, sometimes they are very wrong.

I apologize in advance because I have a hard time differentiating between the co-hosts, but this week one of the two, either Ty Hildenbrandt or Dan Rubenstein, picked Oklahoma State to beat Kansas State on the grounds that Kansas State will eventually lose because "despite being very solid on defense, I still have questions about their offense". [Correction: Ty Hildenbrandt has informed me that 1) these are his words and 2) he did not pick OSU to pull off the upset but that he was generally concerned KSU was too dependent on Klein.] He then added that they could be beat if a team figured out a way to take Collin Klein out of the game. Unless you're willing to take a crowbar to his knees, it's about time people recognized that you can't just take Collin Klein out of the offense, and with Klein, Kansas State has the best offense in the country. And it's not that close.

That's a bold claim, especially when Kansas State is only 5th nationally in points per game (ahead of Baylor, Oklahoma State and West Virginia, all teams that are/were supposed to give Kansas State trouble because they could outscore the Wildcats. Absurd). For the sake of simplicity, I'm going to assume that most would argue Oregon has the nation's best offense, and I will focus the analysis on comparing Kansas State to Oregon. (Spoiler: Kansas State wins!)

Let's begin with points per possession. Because teams take turns with the ball, we should really focus more on points per possession and less on points per game. Kansas State leads the country in points per possession, and its not close - 3.86 points per possession for the Wildcats versus 3.47 for 2nd place Oregon. Kansas State has scored touchdowns on 50% of its possessions. At Oregon's game speed, they would have needed to score 50 more points, one complete game, to keep pace with Kansas State.

The follow up is that Oregon pulls starters early in games with big leads. First, if you haven't been paying attention, Kansas State has also been playing from ahead all season (Klein is in on 65% of 2nd half possessions). Second, the stats still add up for Kansas State. The EP3+ is a measure of performance designed, among other things, to adjust for the quality of the competition and the whether or not the game is still competitive. And what team has the top rated offense according to the EP3+? Kansas State. And it's not close. Alabama also finishes ahead of the Ducks. (Oregon scores significantly higher on defense, because Oregon has a better defense than Kansas State.)

Now Oregon does have a solid lead on Kansas State is yards per rush: 6.2 to Kansas State's measly 5.5. But it is no secret that the Kansas State offense has played a tougher schedule than Oregon's. When we adjust those numbers for the strength of opposing defenses, Oregon's advantage shrinks to .35 yards per carry. Kansas State sits in 6th place and Oregon falls to 2nd behind Johnny Football and company.

When we look at passing the ball, the Kansas State advantage is glaring. Collin Klein averages around 9.8 yards per pass - almost a first down per attempt. Oregon? 7.4 yards per pass is good enough for 63rd nationally. Oregon falls to 67th nationally when we adjust for the defense while Kansas State climbs to #2.

And why is the Kansas State passing attack so efficient? Besides being a human wrecking ball, Collin Klein is the most accurate passer in college football. Klein is 4th in raw completion percentage, but when we adjust for yards per catch and the strength of the opponent's pass defense, he comes out #1. He also leads the country in EPA and NEPA. And consider this: Kansas State is drawing on its biggest strength, Collin Klein's arm, on only 33% of plays while Oregon is running the ball 65% of the time. In addition to the EP3+ stats I cited earlier, it is on these grounds that I dismiss the argument that Oregon's offense is better but they haven't really been trying.

But the Kansas State offense is boring and predictable, right? Oregon averages 27.9 explosive plays (>25 yards) per 100 possessions. Kansas State? 32.6. If we look only at the first half (both Kansas State and Oregon starters accumulate stats on more than 94% of 1st half possessions), Oregon climbs to 32.6, but Kansas State inches up as well to 33.3 explosive plays per 100 possessions.

On top of all that, Kansas State does the little things well. Only three teams have fewer penalty yards per game, only 1 has fewer turnovers per possession, and nobody in the country has a better average starting field position. Oregon, on the other hand, is in the bottom half of the country in turnovers per possession and penalty yards per game, and 23rd in average starting field position.

So Kansas State scores more points when they get the ball, they get more yards per play, they get more big plays, and they do fewer things to shoot themselves in the foot; and the do these things late in games, early in games and, I presume, in the middle of games. Q.E.D.

You can contact Scott at scott@cfbtn.com if you want to let him know how smart and clever you think he is. If you want to tell him he's doing a terrible job, send your thoughts to genechizik@auburn.umd.edu 

43 comments:

  1. How do you get the voters to pay attention to an article like this? Not sure I agree with the assessment of the Defenses (K-State has held better offenses to WAY under their average compared to Oregon), but everything else is spot on.

    ReplyDelete
    Replies
    1. I would like nothing more than for voters to pay more attention to this. I'd say tell your friends and tell them to tell their friends. Post it on facebook and twitter. Harass sports media outlets with it. Attach it to the leg of a carrier pigeon. Smoke signals, maybe.

      Delete
  2. In the time it took to read this, Oregon had two scoring drives on USC.

    ReplyDelete
    Replies
    1. Only two? It's 900 words. I guess USC had to score twice while you were reading, too, right?

      Delete
    2. Actually it was 3, they also gave up 2 from USC.

      Delete
  3. K-State hasnt played Baylor yet...

    ReplyDelete
  4. I believe Kstate has also played 6 consecutive top 25 teams and have blown most of them out.

    ReplyDelete
    Replies
    1. That is a lie. Name them.

      Delete
    2. They have played 4 in a row and 5 out of the last 6 have been ranked. Oklahoma, (KU), Iowa State(ranked in coaches poll only), West Virginia, Texas Tech, and Oklahoma State. He said I believe Kstate played..... ie. he wasn't sure. Calling him a liar is way over the top.

      Delete
    3. 12 oklahoma & 22 texas tech. that = 2

      Delete
    4. I believe he meant at the time that they played them they were ranked.

      Still, out of currently ranked teams, K-State holds a 2-1 advantage over Oregon.

      Delete
  5. What a rediculous homer piece. Spin and manipulation of stats is not the truth.

    ReplyDelete
    Replies
    1. Oregon is being exposed. No way the get to the BCS champ game if KSU and Alabama remain undefeated. Deal with it.

      Delete
    2. An ironic strategy considering I cite evidence and you don't. I would just laugh this off, seeing as how I have no association with Kansas State and will remember Dec. 5, 1998 on my death bed . . . and smile, but I find it morally corrupt and contrary to the pursuit of truth to blindly accuse someone of spin and manipulation just because they present evidence that you don't like.

      Delete
    3. On no, you're an Aggie!!! Lol...at least an unbiased one.

      I hope you guys take down Bama.

      Delete
    4. The OP has no idea what "manipulation of the stats" means. Presenting them is not manipulating them...did the author run a transformation and pick and choose what datums to use and not use? No...you're being ignorant, and show that your whole "argument" is motivated by passion and not logic.

      Delete
    5. I don't see how these stats Prove K-State has a better offense. These are stats based off who each team has played. Both teams are in separate conferences, and the only head to head in the two conferences was (W)Arizona 5-4(2-4) vs (L)Oklahoma State 5-3 (3-2) I'm not claiming the ducks are a better O, but you saying your stats are evidence is false. Games are still played on the field and not on paper last time i checked

      Delete
    6. Eric,
      Show me a game where the offenses play each other on the field and then we will talk.

      Also, as Collin not Klein said, all the stats COME from games that were played on the field. I'm pretty sure that Scott didn't fabricate them or get them from a dream he had one night.

      Delete
    7. Remember, remember the fifth of December
      BCS, Aggies and plot.
      I see no reason why BCS, Aggies
      should ever be forgot...

      Remember, A&M. I am looking forward to things coming full circle this weekend.

      Delete
  6. Collin not kleinNovember 06, 2012

    Oh no! The old paper vs the field argument! So you probably watch games on mute and never watch or read anything espn puts out. We are all well aware football is played on a field, which amazingly is where the performances happen in which we render analysis and statistical measures. And truth is oregons chances at the ncg lie with notre dame losing, and specifically losing to USC. Then oregons comp rankings gain ground on more than one facet.

    ReplyDelete
  7. LOL..the big 12 stinks...Zona, the Pacs 7th best team scored 59 and crushed on of your 4 best. Big 12 would get hammered by the SEC just like usual. If Alabama, Oregon, and K State are undefeated Ksate has about a 10% chance of making the NC

    ReplyDelete
  8. I don't really see the big "argument". One team averages 33 explosion plays per 100 and the other 28? One team averages 6.2 ypc while the other 5.5? One QB ranks #1 in the nation in PE while the other is #7? One team is scoring 3.86655432 points per possession and the other is 3.47665549?

    All seems pretty comparable when comparing different teams in different conferences across 9 games with different opponents. Even some of the points like penalties seem a bit pointless. 8 of the 20 most penalized teams in D1 right now are in the Pac-12 and 10 of 12 teams in the Pac-12 are rated 65th or worse in penalties. Pac-12 refs like their flags.

    ReplyDelete
    Replies
    1. Stupid! Interpretation of stats in this manner is just wrong. The Ducks have only played their first team on both defense and offence an average of a 1/2 per game for every game except USC......There were times that 3rd, 4th, 5th and yes even 6th string players were playing total quarters of a game. Of course the numbers will be skewed if you look at total possessions and not account for this FACT! Stupid! Stupid! Stupid! If you really want to compare apples to apples take this into consideration regarding your idiotic analysis.

      Delete
    2. You didn't read the whole post, did you? For example, this part: "The follow up is that Oregon pulls starters early in games with big leads." Expressing an opinion on something that you haven't actually read is, how did you put it, "Stupid! Stupid! Stupid!"

      Delete
    3. To the OP, you choose to dismiss K-State's advantage offensively because it is small. I disagree, but it's a reasonable position. In many ways the two teams are similar. To say that there is no argument, though, seems like a stretch. A number of people have reacted like I'm committing heresy by even suggesting that the Oregon offense is fallible. And the difference is not negligible as you suggest. Assuming 15 possessions per game, a .4 point per possession advantage comes out to 6 points. That advantage for K-State is slightly larger when you adjust for the competition. If you were to spot one of two evenly matched teams a 6 point advantage they would win 70% of the time. So, yeah, that's kind of a big deal.

      Delete
    4. Scott I did read it the entire argument. I just didn't think this argument was worth a longer response than I originally provided. I am saying to do this analysis correctly, you have to go back to each game that both teams played and determine the point of time that first teams were pull and re-model the data accordingly...that time point is when it is obvious that a victory is in hand. Any statistical comparison that whomever is trying to make here is just not correct or valid without doing that. Now, once that is done then we can discuss strength of schedule, the style of play (passing vs. running) etc. and how to include that in the model. Just throwing statistics out for statistics sake and justifying it with vague statements that are truly biased is simply stupid!

      Delete
    5. That's what the EP3+ does. It uses point differential and personnel (based on stat accumulation) on each drive to weight drives as more or less competitive. If a drive is deemed "non-competitive" it is dropped from the analysis. Then it evaluates each drive based on points scored and the change in field position and finds the best fit for each competitive drive in every game. And Kansas State has a bigger lead in the EP3+O (offense) than in does in raw points per possession.

      Delete
  9. Oregon has been shutting down the offense early in games because
    Chip Kelly doesn't want the bad karma of scoring 100 points on a team.

    ReplyDelete
  10. Kansas State uni's are old school UGLY...therefore the Ducks automatically outclass them on the eyeball test. They do not have the athletes or depth at position to hang with the Ducks for four quarters. If you think they do then you live in OZ...yo Toto, this isn't Kansas anymore!

    ReplyDelete
    Replies
    1. Wow Doug, thanks for confirming what we all know.... It's all about the uniforms! This is real football, something you granolas from the coast don't and won't understand. I'm glad kids don't come to my school "because of the unifroms" I'm glad they are here to play football and represent themselves and the University.

      Delete
    2. Oregon's uniforms are better than KState's???? Only a mother could love that crap that Oregon rolls out each game. This is football, not a car show. Oregon's uniforms scream "thug" more than "class". Worst argument ever.

      Delete
  11. I got to thinking about the "analysis" and came up with a different angle to get out who might be "best". How about points per minute of offensive possession?

    Oregon is averaging 54.33 points per game on 27:48 minutes of possession.

    KSU is averaging 44.33 points per game on 32:30 minutes of possession.

    So, Oregon is averaging about 1.97 points per minute on offense while KSU is averaging about 1.37 points per minute on offense.

    1.97 would appear to pretty significant over 1.37.

    Give both offenses the ball for 5 minutes and Oregon scores 9.85 and KSU scores 6.80.

    See what can be done with "stats"?

    ReplyDelete
    Replies
    1. That's real neat how you keep trying, Oregon fan. You're college football's Tienanmen Square man, facing down the statistical tank with your rubber chicken - Hey look it, our team snaps the ball quickly and can't maintain a drive - flashy like our uniforms. You know who doesn't have flashy uniforms, doesn't snap the ball quickly, does maintain drives, and wins national championships? The same team that will roll Oregon like a bowling ball if you're stupid enough to test the SEC again.
      PS - A little tip: when you put a lot of words in quotes it reads like a 14 year old boy, rolling his eyes at his parents because he's deeply ashamed of his acne and his tiny manhood.

      Delete
  12. KSU would handle Oregon no problem. Their offense controls the ball and no way they (Oregon) get into any type of rhythme. By the time their offense starts to roll, cats are up by 24 and cruising...

    ReplyDelete
  13. "PS - A little tip: when you put a lot of words in quotes it reads like a 14 year old boy, rolling his eyes at his parents because he's deeply ashamed of his acne and his tiny manhood."

    What a strange reference. Sounds like some trauma left over from youth. There are some people you can talk to who can help.

    ReplyDelete
    Replies
    1. Forgot to add, Oregon can't "maintain" drives? ESPN reported after the USC game in their ESPN box research notes, Oregon had eight TD drives of 75 yards or more, which is the most by ANY D1 team in ANY game in the last 9 years. That includes every game played by every SEC team in 9 years.

      Oregon also has 489 points so far this year, and is averaging 54.33 points per game. If it averages 46 points per game in its final 5 games, it will break Oklahoma's NCAA record of 716 points in a season. It is on pace to score 761. That includes three games where Oregon essentially packed it in on offense after half time (up about 50), and only scored a total of 14 points in those 3 second halves.

      If Oregon breaks 30+ against Cal, it will become the first team to score 30+ in 23 straight games (the record of 22 shared with Oklahoma State in 2011-12 and Hawaii 2006-07). If Oregon scores 40+ against Cal it will become the first team in the BCS era to score 40+ in 13 straight games (Texas 2004-05).

      There is some more "rubber chicken" for you.

      Delete
    2. Be careful or you'll put an eye out swinging your rubber chicken around like that. I'm not going to say Oregon's performance against USC wasn't good - it was a real nice show from the #2 offense in the country. But the rest of the stats are a product of tempo. If you play fast enough you'll eventually score a lot of points. Too many idiot football fans confuse playing fast with having a good offense and a bad defense, which is why you give up 51 to USC and the Oregon defense is always underrated. That's caveman analysis, some real quality rubber chicken. It is a real, nonnegotiable fact that Kansas State scores more points per possession, so if Oregon scores more points per game it is only because they (and their opponents) have more possessions. I am convinced from the stats I've seen that the Kansas State starters do more per possession (move the ball and score) than Oregon. If you disagree, then I accept your right to be wrong, but don't come at me with points per minute or points per game. That's rubber chicken.

      Delete
  14. Well, you got me interested so I took a look. For possessions, I did not include any punt or kickoff returns for TDs (as the offense never takes the field); or, any defensive scores; and, tried for each team to not count running out the clock at the end of halves/games. I used ESPN boxes. I'll list what I found by drives for TD then drives for any score against total possessions:

    KSU:

    Missouri State: 5-12, 8-12
    Miami: 7-11, 8-11
    N. Texas: 4-9, 4-9
    Oklahoma: 2-9, 3-9
    Kansas: 8-12, 8-12
    Iowa State: 3-10, 5-10
    WV: 7-10, 9-10
    TT: 6-11, 8-11
    Okie State: 3-12, 6-12

    TOTALS: TD per possession: 45 out of 96 (.468)
    Scores per possession: 59 out of 96 (.615)

    Oregon:

    Arkansas St: 8-14, 8-14
    Fresno St: 6-14, 6-14
    TT: 9-16, 9-16
    Ariz: 4-14, 6-14
    WSU: 6-14, 7-14
    UW: 6-11, 7-11
    ASU: 6-16, 6-16
    Color: 9-13, 9-13
    USC: 9-12, 9-12

    TOTALS: TDs per possession: 63 out of 124 (.508)
    Scores per possession: 67 out of 124 (.540)

    So, despite just "running more snaps", Oregon is now the one with a higher percentage of TDs per possession. However, KSU has kicked a lot of FG, which count and overall, still scoring some points at a higher % of drives.

    TOTAL POINTS PER POSSESSION: once again this will discount defensive scores and punt and kick returns and will include TDs, XP, 2P, and FG.

    We a VERY close margin: KSU (3.72), Oregon (3.67). So, every 10 possessions, KSU is scoring (0.4) more, less than half a point.

    ReplyDelete
  15. I'm not sure how the other PPP stat was calculated. It might just be total points/total possessions, which obviously would be less accurate. Defensive or special teams scores should clearly not be a part of the analysis, nor should "kneel down" type possessions matter.

    So, KSU still leads. However, (always a however), as mentioned Oregon has played in a lot of blowouts this year. I think 6 of 9 have been over by or before half and 8 out of 9 over in the third.

    I see KSU with 4 of 9 similar games (Miami, WV, TT, and Kansas). The other games I have entering the 4th quarter at 16-9, 21-13, 10-16, 24-14, and 38-20. Kansas State has had more games where it has had to keep being efficient offensively in order to secure the win.

    Oregon has had at least 7 games where it almost looked like it stopped trying to score at some point: Arkansas State (50-3 in the second quarter), Fresno State (35-3 in the second), TT (35-0 at half), ASU (43-0 at half), UW (35-7 at half), Colorado (56-0 at half). What you find is Oregon's "points per possession" is great early in those games and falls off as it gets way ahead. For example:

    Oregon went 7-7 with 7 TDs against Arkansas State. Rest of the game? 1-7. TT? 7-10 with 7 TDs. Rest of the game? 2-6. ASU? 6-7 with 6 TDs. Rest of the game? 0-9. Colorado? 7-7 with 7 TDs. Rest of the game? 2-6. I think one gets the point. "Calling off the dogs", which arguably Oregon has done most of the year for long stretches, has effected its "points per possession" quite a bit with a lot of non-scoring drives. Even if a team is running its 3rd string and walk-ons between the tackles to run out the clock, that drive still shows up in the "points per possession" stat.

    KSU has some of that as well. Not as much considering a lot more of its games have gone into the 4th still live. Finally, I noticed some of KSU's most efficient "points per possession games" have been against the weaker teams. It was 2-9 scoring TDs against Oklahoma, 6-11 against TT, 3-12 against Okie State, and 3-10 against Iowa State. Great against West Virginia (7-10) and Kansas (8-12).

    Oregon despite smashing ASU 43-0 at half, only ends up 6-16 for TDs on the game. It's worse outing on paper is probably UA, which is something like 105th in total defense (where it ended up winning 49-0). It was solid versus UW (6-11) and excellent versus USC (9-12).

    I attempted to do my best, but there might be an error here or there, that is why I listed it in detail for anyone who wants to check. The final point, if you are convinced "points per possession" is key, if Oregon passes KSU in that stat, will it mean Oregon is now better than KSU, and it isn't just about "running more plays"? Because, depending on how the games go Saturday, it just might happen.

    ReplyDelete
  16. Rechecking my numbers as it was a big undertaking, caught one error. Oklahoma for KSU should read:

    Oklahoma: 3/10, 4/10.

    It actually was in how it was listed in the ESPN box. It made it look like Tuggle scored on the fumble recovery, when there actually was a 1 yard 1 play drive for a TD for KSU.

    That changes the numbers to:

    KSU:

    46 out of 97 (.474)
    60 out of 97 (.619)

    Points per possession:

    KSU: 3.75
    Oregon: 3.67

    Difference in a 50 play game: 0.8 points. I'd just edit it but I don't see an edit function.

    ReplyDelete
  17. And one final thought. So far this year KSU has averaged 10.8 possessions per game and Oregon 13.7. Roughly 11 and 14.

    To discount for Duck fans complaints about "garbage time", let's take the first 10 possessions for each team in the first 9 games and figure "points per possession".

    That should account for most of KSU's possessions, and perhaps drop off a bit of "garbage time". 10 possessions in most games includes pretty much 3 quarters of all its games, into the 4th. For Oregon, 10 possessions gets it into the third; and, generally will "drop off" the ends of these games, of which 8 of 9 have been pretty much locked in.

    Looking at just the first 10 we find:

    KSU: 43 TDs in 89 attempts (one game KSU only had 9 possessions total).
    KSU: 56 total scores in 89 attempts (43 TDs, 13 FGs)

    Once again not counting return or defensive scores. That is 340 points in 89 possessions for a "points per possession" of 3.82.

    ORE: 55 TDs in 90 attempts
    ORE: 59 total scores in 90 attempts (55 TDs, 4 FGs, 2 TP conversions)

    That is 399 points in 90 possessions for a "points per possession" of 4.43.

    Obviously, 4.43 > 3.82. That is just the first 10 in each game, so it "discounts" the idea the numbers just come from "running more plays".

    Give each team 10 drives, Oregon scores 44.3 and KSU 38.2.

    ReplyDelete