What does it take to win?

by BoddickerIsClutch on Sep 29, 2009 2:02 PM CDT Comment 18 comments

I got this idea from something said on the radio the other day, and so I wanted to run some numbers and see what I got. The basic premise is, do you really need to run the football effectively to win a game? The old idea is ground control, clock control, less turnovers, manageable third downs and all the other great things that come from a successful run game.

But is that idea out of date? What does it really take to win games now, today (or in this case, last year) in the Big 12?

The one advantage the NFL has over college is the relative parity in the league. Sure, some teams are better than others, but all are populated by extremely talented players, and genuinely capable coaches. The disparity in college ball is far larger.

So for this review, I took Texas, OU, Tech, Ok State, Mizzou, KU, and Nebraska as my representatives. The games I looked at were all the intermatches between these teams. For more data points, I also tossed in any out of conference games against quality opponents, including bowl games.

Here is what I found. The team that wins...

Gets the most first downs 80% of the time.

Has a higher 3rd down conversion rate 88% of the time.

Outgains the other team 88% of the time.

Out passes the other team 56% of the time.

Out runs the other team 72% of the time.

Has less penalties called against them 72% of the time.

Has less penalty yards 72% of the time.

Gets more turnovers 56% of the time.

Wins the time of possession battle 68% of the time.

Obviously there is some correlation here. Those teams that get the most yards will usually have the most first downs and therefore time of possession. Less penalties also goes toward that end as they can kill drives. In what was certainly a pass happy league, the passing yards were just about a wash. This is most likely due to teams that are behind resulting to the pass to try and catch up, skewing the losing teams stats closer to the middle.

Other items of note:

Only one team won without getting either more first downs or a better 3rd down conversion rate, that was Missouri against Northwestern. (Winner had a special teams TD)
In conference play, the only winner (out of 15 games) to not have a better 3rd down conversion rate was Texas against OU. (Winner had a special teams TD)
Of the nine categories I listed, the typical winner wins 6.48 of them.
The two teams out of 25 games that won all 9 were Texas against Mizzou, and OU against KU.
One team won while winning only 3 categories (Mizzou against Northwestern), two teams won while only winning 4 (UF over OU, and Texas over OkSU).
4/25 winners won the TOP battle without also rushing for more yards, 5/25 winners had more rushing yards, but less TOP, and 3/25 didnt win either.

So In general the adage still holds, although it would be more accurate to say, the team that can convert the third down, and thus get more first downs, wins. Though it would appear the running game has a large effect on these two numbers, as expected.

Any thoughts? I'll keep my crude spreadsheet around for a while if anyone has any other combinations they want looked at.

All comments, FanPosts, and FanShots are the views of the reader-authors who create them.

4 recs | Comment 18 comments | Add comment

Story-email Email Printer Print

Comments

Nice post

One metric that might subsume several of those is number of plays

by burnt in ny on Sep 29, 2009 3:29 PM CDT reply actions 0 recs

Also, I've seen correlations like this just take into account first half stats

to balance out the fact that strategies can change after halftime depending on who is ahead or who is behind that may not reflect the overall “team” strategy for success/failure.

by notsofst on Sep 29, 2009 3:52 PM CDT reply actions 0 recs

at least in games that are not close, i.e. decided by two or more scores.

by notsofst on Sep 29, 2009 3:53 PM CDT up reply actions 0 recs

If I had a better way of getting stats

Then opening a ton of ESPN pages, that would be a good idea.

by BoddickerIsClutch on Sep 29, 2009 4:00 PM CDT up reply actions 0 recs

Is cfbstats.com insufficient?

In any case, I like what you’re doing BiC.

proud to swim home

by learned hand on Sep 29, 2009 4:07 PM CDT up reply actions 0 recs

That would work

The way i was recording my data was comparatively. IE, for this game, the winner did better in this category yes or no.

Box score stats are good cause I can quickly match up that info and make the comparison.

I could use cfbstats.com, but then I would have to input the raw data. Not impossible, just not how I was doing it. If I do an update later in the year, Ill look into that. Although for what notsofast was suggesting, cfbstats wouldnt help. That would take real work, something I dont know if Im dedicated enough to provide.

by BoddickerIsClutch on Sep 29, 2009 4:16 PM CDT up reply actions 0 recs

I'm confused

For more data points, I also tossed in any out of conference games against quality opponents

Missouri against Northwestern

But seriously, thanks for the analysis. As far as the old idea, seems to me that it is founded on the, also old, idea of passing. “Three things can happen when you pass and two of ’em are bad.”

Spread offenses seem to have changed that. Passing is now much more likely to be successful. I don’t really think you have to be a running team to be considered a ball-control offense anymore. As long as you can consistently dink and dunk you’ll be alright. Our 2008 offense epitomizes that concept (I’m reserving judgment on 2009 for the moment). I also think that’s why GD doesn’t go vertical very often.

If it’s easy to do, Mack and I would also be interested in seeing how “explosives” correlate.

by gwh65 on Sep 29, 2009 4:28 PM CDT reply actions 0 recs

That was actually a very competitve game.

I didnt watch it, so I have no idea how Mizzou won, but I guess the difference was Maclin. Satistically speaking the thing that makes the least sense from their box score is the final score – http://sports.espn.go.com/ncf/boxscore?gameId=283640077

by BoddickerIsClutch on Sep 29, 2009 4:34 PM CDT up reply actions 0 recs

Also more evidence for my thesis of last year

That Mizzou actually sucked, and just fooled everyone all year somehow into thinking they were a good team.

See also games against all of the big 12 south teams (Baylor almost got em)
See also game against KU
See also Bowl Game

by BoddickerIsClutch on Sep 29, 2009 4:36 PM CDT up reply actions 0 recs

Hey now

The wildcats have a decent season to good every decade or so. Last year was it. This year, well, I’m hoping they beat Indiana.

proud to swim home

by learned hand on Sep 29, 2009 4:36 PM CDT up reply actions 0 recs

unimpressive proofreading

proud to swim home

by learned hand on Sep 29, 2009 4:38 PM CDT up reply actions 1 recs

I did something very similar in 2007.

Here it is: http://www.burntorangenation.com/2007/9/4/75624/82027

by billyzane on Sep 29, 2009 4:38 PM CDT reply actions 0 recs

Great stuff here

I have no idea how to “get comfortable” with statistics, but if I’m trying to coach a winning football team here’s what I emphasize:

Ball possession.

This is not necessarily time of possession. The concept is this: If I have the ball, you don’t. This is totally critical in an era, and especially in a conference, where offenses so dictate the outcome of games. If I have the ball I can: (1) score points; (2) better my field position, which worsens yours; (3) rest my defense; (4) achieve, or maintain, the statistically unmeasurable concept we call “momentum.”

To win the game, I want to run more plays than you do. Also, to go back to Mack’s favorite two-step, I want to at least break even in explosives and turnovers.

by edsp on Sep 29, 2009 4:40 PM CDT reply actions 0 recs

Don't tell your philosophy to the Miami Dolphins

I know, I know that was a fluke.

by aaronlybrand on Sep 30, 2009 12:25 AM CDT up reply actions 0 recs

Turnovers = Skewed Data Point

I think turnovers are a bigger factor than your data would suggest. The fact that a much larger % of competitive games would have teams with equal number of turnovers as opposed to the other metrics. (ie – Time of possession almost never 30-30). If you stated – if you have more turnovers – you would win x% of the time – this factor would be much greater.

by realmccoy on Sep 30, 2009 3:45 AM CDT reply actions 0 recs

I think I get what your saying

But I did say if you have more turn overs you win 52% of the time. Except I didnt word it that way to avoid suggesting causality.

Turnovers will have more ties than any other factor, but if its a tie isnt it a wash? Is a 0-0 turnover game generally have its outcome effected by turnovers? Obviously not. What about a 2-2 turnover game? Thats tougher, to get a real answer youd have to look at it situationally. I hoped to take enough data points so that it would even itself out, meaning in one game a 2-2 TO game could have been huge for the victor, while in another a 2-2 benefited the loser more.

At any rate, its the discrepancy in turnovers that I wanted to illustrate. And I was surprised by the result. You would think the extra opportunities combined with what should be either a favorable field position, or a great reversal of fortune would have a larger impact on the outcome. Perhaps it would be useful as well the losing percentage of teams with less turnovers.

To that end, I just went through all the games where the winning team did not have more turnovers. First, I had one game wrong, so the turnover percentage is actually 56% (It was the Texas/TTU game. Honest mistake I swear, not a Freudian slip).

5 of the 25 games had the team with less turnovers actually winning, while 6/25 had a tie at the turnover mark. 2 of the 5 were at a -2 in turnovers even.

by BoddickerIsClutch on Sep 30, 2009 8:41 AM CDT up reply actions 0 recs

causality

The problem here is separating the causes from the side-effects. I would think that teams wearing colors do better than ones wearing whites. But it is not the uniform that is the causal factor. It’s that home teams most often wear their colors and home teams have an advantage. So when we look at factors like time of possession, we need to ask whether running clock is itself a path to victory. Teams running hurry up offenses may think not.

Before you criticize someone, walk a mile in their shoes. If they get mad, you're a mile away AND you have their shoes.

by Caradoc on Sep 30, 2009 9:03 AM CDT reply actions 0 recs