~ ~

Saturday, November 16, 2013

Countdown To Roses - MSU at Nebraska Post Game

Predictions were made based on statistical analysis in Countdown To Roses - MSU at Nebraska.

Numbers do not win games, no matter how cleverly they are presented and/or manipulated. Statistics can only give probabilities based on quality data. In Countdown, it was acknowledged that there were insufficient data for a good analysis. However, I also wrote that I had my spoon ready and I expect the crow to fly.

Prediction, Michigan State yards rushing: 182 (+57/-63) yds. Actual: 168
Prediction, Michigan State yards passing: 175 (+50/-53) yds. Actual: 193
Prediction, Michigan State scoring: 24 (+8/-7) pts. Actual: 41

Prediction, Nebraska yards rushing: 71 (+15/-28) yards. Actual: 182
Prediction, Nebraska yards passing: 144 (+44/-43) yds. Actual: 210
Prediction, Nebraska scoring: 14 (+5/-5) pts. Actual: 28

Only 2 of 6 predictions were accurate.
Crow is pretty good barbecued with salsa and beer.

------------- Remember to check Spartan Headline links, updated real time, at the bottom of the page (Web version only). -------------- Please click on COMMENTS below the post to enter and view reader comments. ***If you are receiving this post via automated email, ***you need to go to the site to view headline links and embedded videos in this post: (Web version only)


  1. SM76 there is no need to be so hard on yourself. To begin with, you were 100% accurate on the most important of all the predictions and statistics... MSU would win that game. At the end of the day it doesn't matter how many yards you pick up or how many they get, as long as you put that W up everything else is just details. That prediction is worth double points in my book.

    Also, the 64 yards passing that Nebraska picked up on that final drive was hardly earned. It didn't look like Narduzzi was dialing up anything more than the don't give up more than 15 yards on any given play defense. Taking that away, puts the actual numbers in your wheelhouse.

    Finally, your prediction of our points scored was a little under where we ended up. I am guessing your thought process didn't take into account Nebraska fumbling the ball 6 times (4 recovered by us) and throwing a pick too. 5 TO's is going to skew point predictions every time.

    And hey... its stats. There are always going to be outliers. So sometimes the numbers are going to lean high and sometimes they will lean low (hello Michigan. How is Gardner feeling these days?)

  2. Hey 76,

    Glad to be in agreement with ATown (as usual). There is no need for you to order a plate of Crow. Desmond Howard is still working on his order, and he may need some more time to shove it all down his oversize gullet.

    Your statistical predictions were cool, and you should keep working on them. I may not be the stats-guy you are, I guess I'm more of an aggregated-data guy, but even you said it takes 100 samples to get a good picture of something. So you created one sample. Keep 'em coming.

    ATown makes two excellent points about unpredictable variables in turnovers and the garbage yards at the end. The last drive resulted in 64 extra passing yards and seven extra points, so you were much closer in two categories than you are allowing yourself credit for. The TOs were claimed to result in something like 17 extra points, without which you hit that other nail on the head, too. For what it's worth, I always assumed your predictions were based on an EVEN turnover ratio.

    I suppose you might have been a bit off on the Nebraska Rushing totals, but much of that was "created" by the effort of a single player who got more from several plays than was actually there. Take away a few of his big runs, and even those numbers come much closer to the bulls-eye you predicted.

    In the Macro-End, you predicted us to win by 10, and we won by 13. WE WON AND YOU WON. And you would have even won in Vegas with that.


    1. SM82, You shouldn't agree with me. I have been agreeing with me for years and its never a good thing.

      The Vegas point is also more validation to the predictions. Vegas had the over under at 41-41.5... MSU covered that alone. Vegas also had the line at 6. So your line was way more accurate. Basically, Vegas gets paid for its predictions and your numbers were either just as skewed as theirs or you were more accurate.

      I am probably with SM82 and I like looking at everything to tell the story but I love the numbers. They are fun to use as predictors and to use as validation for what you saw actually happen. I like the predictions and going out on a limb to be specific on them. Keep that stuff coming and I will defend them for you.

  3. Guys,

    Was I really being hard on myself? No. I was merely laughing at myself an activity grossly lacking it today's society, present company excluded.

    Anyone who understands statistics and probability understands that "likely" and "probably" are not guarantees. The simplistic odds for each of those predictions were not better than 2:1. As SM82 noted, the lack of the minimum data sample size was noted in the analysis and lacking a minimum sample size negates any real predictions.

    If you have never actually had crow (the black bird), it really does go well with salsa and we all know that everything goes better with a cold one.

    If I am "guilty" of anything with this post-game blog, it is being pre-emptive, of assuming not a few chUMps would show up stanching up the place with their dUMb talk.


Please sign in using the method most convenient for you. We do not receive your login information. This function is provided by Blogger.