Writers of Pro Football Prospectus 2008

Most Recent FO Features

SmithSte01.jpg

» The Week In Quotes: August 29, 2014

This week: Josh Shaw lies, Steve Smith intimidates, Le'Veon Bell relaxes, Matt Simms dances, and Clint Trickett kisses and tells.

07 Sep 2012

VN: Embrace The Randomness

by Bill Connelly

In my second ever Varsity Numbers column in 2008, I discussed how I tend to view turnovers using the Equivalent Points measure:

In just about any football game, you'll see a reference to turnover margin, or maybe points off of turnovers. But it doesn't take in-depth thinking to realize that not all turnovers are created equal. If a running back fumbles on his opponent's 1-yard line, that's a huge turnover because his team had a high level of expected points, and he threw them away. And if he fumbles on his own 1, that's also huge because it hands his opponents a high level of expected points. And if he fumbles on his opponent's 1, and it's returned for a touchdown, that's doubly huge -- it cost his team quite a few points and handed his opponents a touchdown. But in turnover margin, all three of those fumbles count the same as if some backup quarterback fumbled at midfield on the last play of a 49-7 game.

It seems clear that, as FO has covered in the past, counting the significance of two values -- the team's field position when the turnover happened, and the opponent's resulting field position -- gives you a much better view of a turnover's true costliness. And that's what we're going to try to do.

Let's look at two turnovers:

* Turnover 1: Team A is about to score. Their running back carries the ball inside the 1, then fumbles. The ball bounces into the arms of a safety, who carries it 100 yards for a touchdown.

* Turnover 2: Team A's quarterback fumbles at his 45-yard line, and Team B recovers. Team B then immediately goes three-and-out and punts the ball back to Team A.

Using my numbers, Turnover 1 was worth 12.62 points (5.62 points for being at the opponent's 1, 7.00 points for being returned for a touchdown). Turnover 2 was worth 4.26 points (1.92 points lost/prevented, plus 2.34 points given/taken). Is that not a much more accurate read of which turnover truly impacted the result of the game and which did not?

I have used the Turnover Points concept frequently in my work at SB Nation, and it really does fill in quite a few gaps in terms of the difference between points scored, yards allowed, and huge special teams plays made. With Brian Fremeau revisiting the turnovers concept earlier this week, I thought it would be interesting to look at turnovers myself and see what we can glean from Week 1's action.

In all, Week 1 saw FBS teams play 76 games, 32 against FCS teams and 44 among each other. On average, the games saw 3.5 turnovers worth approximately 5.2 Turnover Points each. When you think about it, that is pretty staggering. For all the analysis we do about talent, matchups, etc., about 18 points per game (3.5 x 5.2) are determined simply by your ability to either force terrible mistakes, avoid them, or get lucky and fall on one when it hits the turf. No wonder football coaches often go gray faster than U.S. presidents.

In all, the results of 11 Week 1 games were decided almost entirely by turnovers. Below you will see those games broken out with the following columns: week, date, teams, score, turnovers committed, equivalent point value of those turnovers, scoring margin, turnover points margin, and "Non-TO Margin," which signifies the difference between the scoring margin and turnover points margin. For the 11 games below, the winning team had a positive turnover points margin and a negative "Non-TO Margin."

Week Date Teams Score TO TO
Points
Point
Margin
TO Pts.
Margin
Non-TO
Margin
Week 1 8/30/12 Kent State 41 0 0.0 +20 +27.1 -7.1
Towson 21 6 27.1
Week 1 9/1/12 Illinois 24 1 5.5 +17 +17.5 -0.5
Western Michigan 7 4 23.0
Week 1 9/1/12 Ohio 24 0 0.0 +10 +13.9 -3.9
Penn State 14 3 13.9
Week 1 9/1/12 Washington 21 1 5.5 +9 +14.1 -5.1
San Diego State 12 3 19.6
Week 1 9/1/12 Miami 41 1 4.1 +9 +16.2 -7.2
Boston College 32 3 20.3
Week 1 9/1/12 Indiana 24 0 0.0 +7 +8.9 -1.9
Indiana State 17 2 8.9
Week 1 8/30/12 McNeese State 27 0 0.0 +6 +9.9 -3.9
Middle Tennessee 21 3 9.9
Week 1 9/1/12 Wake Forest 20 1 4.3 +3 +9.2 -6.2
Liberty 17 2 13.5
Week 1 8/31/12 Stanford 20 0 0.0 +3 +9.9 -6.9
San Jose State 17 2 9.9
Week 1 9/1/12 UTSA 33 1 6.1 +2 +13.9 -11.9
South Alabama 31 4 20.0
Week 1 9/1/12 Northwestern 42 1 4.3 +1 +15.7 -14.7
Syracuse 41 3 20.0
Week 1 9/1/12 Iowa 18 0 0.0 +1 +5.2 -4.2
Northern Illinois 17 1 5.2
Week 1 9/3/12 Virginia Tech 20 0 0.0 +3 +4.4 -1.4
Georgia Tech 17 1 4.4

For a couple of these games, you could make the case that turnovers defined the game but didn't decide them. If Kent State and Illinois had not benefited so significantly from turnovers, they may not have taken the foot off of the accelerator, so to speak, and they may have done enough damage to win the game regardless. (Then again, Illinois quarterback Nathan Scheelhaase got hurt, so...)

For the other nine games, however, turnovers both defined and potentially decided the winner. Bounces ruled in the Ohio-Penn State game; not only did Ohio benefit significantly from turnovers, but they also benefited from a non-turnover. On their first touchdown, Ohio quarterback Tyler Tettleton saw a horrendous pass bounce off of the hands of a Penn State defender (said defender was actually stepping in front of another Penn State defender who might have picked the pass off) and into the hands of his intended receiver, who caught the pass and ran for an unlikely touchdown. Make no mistake: Ohio made the plays it needed to win the game late. But without turnovers, they don't win.

Turnovers made the difference in quite a few upset bids as well. Indiana and Wake Forest each avoided embarrassing losses to FCS teams that had otherwise potentially outplayed them (Liberty actually outgained Wake, 363 to 293), while Stanford narrowly avoided a loss in their first game post-Andrew Luck thanks to 10 turnover points. Washington escaped a bit of a stinker and held off San Diego State thanks to a 14-point turnover margin. UTSA and Northwestern each scored last-second road wins in part because turnovers kept them close enough through the game's first 59 minutes; meanwhile, Virginia Tech and Georgia Tech were in complete gridlock heading into overtime, but Tevin Washington's incredibly misguided overtime interception made the difference. And, finally, turnovers sealed an upset bid (if you can call it an upset) for McNeese State over poor Middle Tennessee, which has fallen off the map over the last couple of years.

So much of college football is determined not only by who can avoid throwing stupid passes, but who is lucky enough to intercept a given stupid pass (instead of simply breaking it up, or, in Penn State's case, batting it to an opposing receiver). It's determined not only by forcing or avoiding fumbles, but picking up the loose ball when it begins to bounce around in an abstract fashion. Turnovers obviously aren't based entirely on luck, but they are just random enough to drive you crazy. As much as or more than any other sport, college football allows fans of a given losing school to defiantly say "We gave them that game," or "They didn't beat us, we beat ourselves." It is oddly comforting to say such things, and as Week 1 proved, it isn't entirely inaccurate, either.

New F/+ Rankings

Week 2 F/+ rankings are now updated here. Obviously these are still based mostly on preseason projections, with a little bit of raw Week 1 data thrown into the mix. With no true opponent adjustments of which to speak, obviously the data is going to overreact to certain games (Alabama over Michigan, Duke over Florida International, or the S&P+-specific Random FBS Team who dominated Random FCS Team), but such is life. These kinks will work themselves out as the opponents adjustments take hold.

(Note: you can find updated S&P+ rankings at the same page. I was hoping to set up this year's S&P+ page this week but ran out of time. You can find the numbers there regardless.)

F/+ Rk Team Record F/+ Last Wk. Change S&P+ S&P+ Rk FEI FEI Rk
1 Alabama 1-0 +34.3% 1 0 278.3 1 0.311 1
2 LSU 1-0 +27.9% 2 0 260.9 4 0.268 3
3 Oklahoma State 1-0 +26.6% 7 +4 268.6 2 0.204 13
4 Oklahoma 1-0 +26.5% 3 -1 253.0 6 0.281 2
5 Oregon 1-0 +25.1% 4 -1 263.2 3 0.201 14
6 USC 1-0 +25.1% 5 -1 257.6 5 0.229 5
7 Florida State 1-0 +22.6% 6 -1 250.2 7 0.215 10
8 Georgia 1-0 +20.9% 8 0 243.3 8 0.215 10
9 West Virginia 1-0 +20.5% 20 +11 240.2 10 0.224 7
10 Ohio State 1-0 +20.5% 13 +3 239.3 11 0.228 6
11 Notre Dame 1-0 +20.2% 14 +3 239.2 12 0.222 9
12 Florida 1-0 +19.8% 10 -2 236.2 14 0.230 4
13 Texas 1-0 +19.4% 9 -4 240.7 9 0.199 15
14 Arkansas 1-0 +17.9% 15 +1 238.9 13 0.177 19
15 Virginia Tech 1-0 +17.6% 18 +3 228.7 22 0.223 8
F/+ Rk Team Record F/+ Last Wk. Change S&P+ S&P+ Rk FEI FEI Rk
16 Stanford 1-0 +16.9% 11 -5 234.8 15 0.178 18
17 South Carolina 1-0 +16.4% 17 0 230.6 20 0.188 16
18 Nebraska 1-0 +16.3% 23 +5 234.0 16 0.170 20
19 Clemson 1-0 +15.8% 21 +2 225.1 25 0.205 12
20 TCU 0-0 +15.5% 12 -8 231.2 19 0.168 22
21 Wisconsin 1-0 +14.0% 19 -2 229.1 21 0.148 26
22 Michigan State 1-0 +13.9% 24 +2 226.2 24 0.160 24
23 Tennessee 1-0 +13.6% 27 +4 224.0 28 0.165 23
24 Missouri 1-0 +13.4% 28 +4 232.0 18 0.122 32
25 North Carolina 1-0 +13.1% 32 +7 233.9 17 0.105 39

This Week At SB Nation

Once again, I will try to share with you some relevant pieces I wrote this week at SB Nation and elsewhere.

Posted by: Bill Connelly on 07 Sep 2012

1 comment, Last at 07 Sep 2012, 5:03pm by Aaron Brooks Good Twin

Comments

1
by Aaron Brooks Go... :: Fri, 09/07/2012 - 5:03pm

"meanwhile, Virginia Tech and Georgia Tech were in complete gridlock heading into overtime, but Tevin Washington's incredibly misguided overtime interception made the difference."

That game never gets to OT without a blown handle on a bad long snap to the punter in the 2nd quarter. Which while not technically a turnover (the vagary of the punter being on the bottom of the pile instead of a defender, on a 4th-down play), it was still a 50-yard loss of field position. Which, if I'm correct, would be worth 2-3 points.