Beyond the Box Score: Special Teams

by The Boy on Jul 6, 2008 11:32 AM EDT

So after yesterday's look at Turnovers, it's now time to establish point values for special teams. Leaving PATs out of it for now (it will obviously be easy to add them later), there are three major special teams category (and a fourth minor one): Field Goals, Punts, and Kickoffs (and then Free Kicks). Let's attack them one at a time.

Field Goals

Figuring out what to do about Field Goals was by far the easiest of these categories. I broke FGs into 5-yard increments (18-22 yards, 23-27, 28-32, etc.), looked at the % made in each group, and determined the expected number of points from each kick. Here's what I found:

18-22: kickers made 91.4% of these kicks. 3 points * 0.914 = 2.74 expected points
23-27: 88.1%, 2.64 expected points
28-32: 80.3%, 2.41 expected points
33-37: 69.4%, 2.08 expected points
38-42: 67.1%, 2.01 expected points
43-47: 58.1%, 1.74 expected points
48-52: 45.6%, 1.37 expected points
53-57: 35.0%, 1.05 expected points
58+: 20.0% (1-for-5), 0.60 expected points

So with that, we can treat every FG like an addition or loss of points. For instance, if you miss a 25-yard FG, it's a loss of 2.64 points. If you make it, it's worth 0.36 points. That may not seem like a lot, but you have to remember that the team has been adding (and possibly subtracting) points all the way up the field. To get to the opponent's 8-yard line, they've probably earned at least somewhere in the neighborhood of 2-3 EqPts, so the 0.36 points seems a lot more reasonable in that regard.

So for the season, who benefited the most from their field goal kickers?

Top 10, FG pts/game

1. Arizona State (1.27/g)
2. Indiana (1.12)
3. North Carolina (1.07)
4. Utah (1.07)
5. UCLA (1.07)
6. New Mexico (1.03)
7. UTEP (1.01)
8. Illinois (1.01)
9. Florida State (0.89)
10. Georgia Tech (0.82)

24. Texas (0.48)
31. Nebraska (0.45)
33. Oklahoma (0.42)
37. Missouri (0.34)

2007 All-American kickers included Arizona State's Thomas Weber, Indiana's Austin Starr, Utah's Louie Sakoda, and New Mexico's John Sullivan, so this seems right to me. It also seems to me that UNC's Connor Barth and UCLAs Kai Forbath didn't get enough recognition.

Bottom 5

126. San Jose State (-1.56)
125. FCS Tier 3 (-1.34)
124. Duke (-1.26)
123. Miami-OH (-1.00)
122. Oklahoma State (-0.90)

(As a method of verifying this, I looked at OSU's 2007 stats. Their two kickers--Jason Ricks and Dan Bailey--made 10 FGs, 9 of which were from under 30 yards. Meanwhile, not only did they miss an under-30 kick, but they were also 1-for-8 from beyond 30. Sounds like a Bottom 5 performance to me.

On to the kicking/return game...

Punts

The FG idea above was something of a no-brainer for me, but for Punts, Kickoffs, and Free Kicks, I had to toss around a few different ideas. Here's what I did (and this applies roughly to all three):

Take the receiving team's point value for the line of scrimmage of the punt. for example, you're punting from your 20 yard line. If your opponent had the ball on your 20 yard line, it would be worth 3.898 points.
Take the point value of where the ball ended up. In the above example, let's say the ball was punted 40 yards (to your opponent's 40) and returned 10 (to the 50). The point value of the 50 is 2.095 points.
3.898 - 2.095 = 1.803. That punt was worth 1.803 points.
Got it? So the higher the point total, the better it is for the kicking team. The lower, the better for the receiving team.

Top 10 Teams, Punting

1. Ohio (2.03 pts/punt) (#53 in net punting-yards)
2. Texas A&M (1.89) (#4)
3. Toledo (1.86) (#2)
4. West Virginia (1.83) (#15)
5. Georgia Tech (1.81) (#7)
6. Eastern Michigan (1.81) (#6)
7. Cincinnati (1.81) (#1)
8. Penn State (1.80) (#3)
9. Oklahoma (1.80) (#24)
10. New Mexico (1.79) (#12)

12. Kansas State (1.75)
34. Texas Tech (1.64)
44. Iowa State (1.61)

So other than the crazy Ohio outlier, there's nothing too strange about these numbers. When you think about it, the biggest difference between my numbers and the typical net punting-yards numbers are that...well, a touchback isn't too big a penalty with my numbers. The difference between downing a punt at the 3 and having a touchback may be 17 yards (which is huge), but it's only 0.903 EqPts, a miniscule point margin compared to other areas of the field.

Bottom 5 Teams, Punting

126. Missouri (1.14)
125. San Jose State (1.19)
124. Kent State (1.23)
123. Nevada (1.27)
122. Oregon State (1.31)

Wow. Strangely, I'm now more excited for 2008. The single biggest question mark Mizzou has is at punter, where Adam Crossett is gone and Jake Harry and Random Walk-Ons #1-3 are fighting it out, but...well, we were already the worst in the nation in punting last year, and it didn't slow us down much! What's it going to be in '08...worse than last place?

So what about the punt return side of things?

Top 10 Teams, Punt Returns (remember, the lower the better)

1. Kansas State (0.92 points per punt) (#1 overall in punt return avg-yds)
2. Utah State (1.23) (#2)
3. San Jose State (1.23) (#10)
4. Western Kentucky (1.26) (#9)
5. Army (1.27) (#5)
6. Ole Miss (1.27) (#95?)
7. Florida (1.27) (#7)
8. Alabama (1.29) (#11)
9. Idaho (1.33) (#3)
10. Navy (1.34) (#13)

And the worst?

126. Wyoming (2.04) (#101)
125. Middle Tennessee (1.91) (#86)
124. Virginia (1.88) (#56)
123. Southern Miss (1.79) (#93)
122. South Florida (1.78) (#65)

Kickoffs

For simplicity's sake, I set this up exactly the same. You kick off from the 30, so that's the first point value in consideration. The second is, naturally, where the ball ends up. Simple = better in my eyes here. I mean, I played around with the idea of figuring out the average point value of each kick (for kickoffs that was 1.46) and comparing teams' averages to that (so that about half the teams would be positive, half negative). However, that basically leads you to the same order of teams, so in the end it just became an extra, meaningless step.

So here we go. The only difference between kickoffs and punts is that the lower number is better here (this is because the team listed as 'offense' in punts is the punting team, while the 'offense' here is the returning team--simply a technicality).

Top 10, Kickoffs

1. Virginia Tech (0.98)
2. UTEP (1.07)
3. Virginia (1.13)
4. Western Kentucky (1.19)
5. San Diego State (1.19)
6. Rice (1.19)
7. Iowa (1.21)
8. USC (1.23)
9. Ohio State (1.23)
10. Georgia Tech (1.24)

Beamer Ball strikes again.

Bottom 5, Kickoffs

(FCS Tier 5 and FCS Tier 6 were the bottom two, so we'll skip those)

126. New Mexico State (1.74)
125. Iowa State (1.73)
124. Temple (1.69)
123. Wyoming (1.68)
122. Washington State (1.66)

On the other end of things...

Top 10, Kickoff Returns

1. Akron (1.74)
2. UCLA (1.71)
3. Fresno State (1.69)
4. UConn (1.69)
5. Kansas (1.68)
6. Wake Forest (1.67)
7. Hawaii (1.67)
8. East Carolina (1.65)
9. Oklahoma (1.64)
10. Purdue (1.64)

Bottom 5, Kickoff Returns

126. Oregon State (1.08)
125. Duke (1.11)
124. New Mexico State (1.13)
123. Northern Illinois (1.14)
122. Virginia (1.14)

Free Kicks

I mentioned that this was a minor category--that's simply because, of 141,000+ plays in 2007, there were 54 free kicks. They make a difference...but not really. Very few teams were involved in more than one free kick in 2007, so we won't waste the time listing bests and worsts here. Just know that they're set up exactly the same as kickoffs, only they're from the 20 instead of the 30.

Special Teams Average

I've got a bit of Phil Steele in me, obviously. Part of the reason I've done all these 'points' measures is for predictive purposes, by all means, but...like Steele, I just love ranking stuff. (How many "power rankings" does he discuss at given times? Like, 1,403?) So with that in mind, you knew I'd attempt to come up with an "overall" Special Teams ranking as well. I did this by adding together the 'higher is better' numbers, subtracting the 'lower is better' numbers. So we get something like this:

Kickoff Return Avg + Punt Avg + (FG Avg * 2) - Kickoff Avg - Punt Return Avg = Special Teams Avg

(I multiplied FG by 2 so that FGs would carry the same weight as kickoffs and punts.)

For the hell of it, here are all 120 teams ranked by Special Teams Avg (I removed the 6 FCS 'tier 1-6' teams from this list):

San Diego State (1.69)
Arizona State (1.55)
Georgia (1.50)
UTEP (1.43)
Indiana (1.38)
UCLA (1.36)
Nebraska (1.35)
Illinois (1.27)
Toledo (1.26)
Utah (1.21)
Oklahoma (1.18)
Virginia Tech (1.12)
NC State (1.09)
Wake Forest (1.07)
Kansas State (1.02)
North Carolina (1.01)
Ohio (0.98)
Georgia Tech (0.95)
UConn (0.93)
USC (0.91)
Troy (0.87)
Idaho (0.86)
South Carolina (0.84)
UAB (0.83)
New Mexico (0.81)
Purdue (0.77)
Boise State (0.76)
Akron (0.72)
Central Florida (0.67)
Texas (0.66)
LSU (0.63)
Florida State (0.63)
Maryland (0.61)
UNLV (0.57)
Arizona (0.57)
Wisconsin (0.55)
Hawaii (0.52)
Ohio State (0.44)
West Virginia (0.38)
Bowling Green (0.35)
Syracuse (0.34)
Auburn (0.33)
Buffalo (0.30)
Penn State (0.30)
Alabama (0.30)
Tennessee (0.30)
Missouri (0.24)
Air Force (0.21)
TCU (0.21)
Mississippi State (0.20)
Western Kentucky (0.17)
Oregon (0.16)
Colorado State (0.16)
Fresno State (0.13)
Texas Tech (0.11)
Navy (0.11)
Louisville (0.05)
Marshall (0.05)
Nevada (0.05)
Pittsburgh (0.03)
Virginia (0.02)
Arkansas (-0.01)
Kansas (-0.02)
Washington (-0.02)
Tulane (-0.10)
Florida (-0.11)
Rutgers (-0.14)
Southern Miss (-0.15)
SMU (-0.16)
Utah State (-0.16)
Kent State (-0.18)
South Florida (-0.18)
Clemson (-0.19)
Western Michigan (-0.24)
Florida Atlantic (-0.26)
BYU (-0.26)
Michigan (-0.26)
Michigan State (-0.29)
UL-Lafayette (-0.36)
Minnesota (-0.39)
Cincinnati (-0.41)
Texas A&M (-0.43)
Stanford (-0.44)
California (-0.45)
Wyoming (-0.45)
Colorado (-0.47)
Arkansas State (-0.51)
Kentucky (-0.51)
Florida International (-0.52)
Washington State (-0.54)
Northwestern (-0.58)
Eastern Michigan (-0.58)
Oregon State (-0.59)
Boston College (-0.62)
Iowa (-0.69)
Miami-FL (-0.74)
Houston (-0.74)
Louisiana Tech (-0.75)
Northern Illinois (-0.75)
Miami-OH (-0.76)
Central Michigan (-0.78)
Ball State (-0.78)
Ole Miss (-0.79)
Rice (-0.79)
Army (-0.85)
UL-Monroe (-0.96)
Memphis (-1.07)
Tulsa (-1.11)
East Carolina (-1.11)
Iowa State (-1.12)
Vanderbilt (-1.14)
Oklahoma State (-1.35)
New Mexico State (-1.41)
Temple (-1.46)
North Texas (-1.52)
San Jose State (-1.66)
Notre Dame (-1.67)
Baylor (-1.73)
Middle Tenenssee (-1.99)
Duke (-2.95)

I've bolded the teams who finished in the AP Top 10, just to see how much of an impact Special Teams made on their success. As you see, some finished high on the list, some in the middle, but 9 of 10 were on the plus side of the ledger. (Of course, Boston College finished tied for #10 with Texas, and they were pretty bad on this list.) So I guess this shows that you don't necessarily have to be great in the special teams department, but you can't be too bad unless you're playing a weak schedule (Kansas) or in a weak conference (Boston College).

My original intent here was to rank teams based on a "special teams points per game" type of measure. However, there's a problem with that--the teams that score a lot are penalized a lot because, well, they also kickoff a lot. So while the per-game numbers are important as far as adding up the components of a final score goes, it's not very good for ranking teams on their special teams units. The above way works better, I'd say.

SUMMARY

So what have we accomplished today, class? Well, we've ranked teams. That's always fun. And beyond that, we've now come up with ways to measure the remaining component of the typical football game that impacts the number of points put on the scoreboard. You can see where this is going, I'm assuming--next up, we'll rank the teams! I'm a-twitter just thinking about it, and I know you are too!!

All comments on 'Sunday Morning Quarterback' are the views of the individual commenter and do not necessarily reflect the genius of SMQ, Sports Blog Nation, etc.

0 recs | Comment 12 comments

Read Related

Story-email Email | Print |

Comments

The other KO method

If you had used the KO method with the baseline as the avg return, then the fact that good teams KO more wouldnt have mattered for doing a per game ranking. If they KO a lot and are good at it, it helps them even more, which it does. If the KO a lot and suck at it, then they are giving back some of the points from scoring.

Something similar would probably have to be done with punts too, but maybe not, sent a punt is, almost always, a negative play. How did you figure muffed punts recovered by the kicking team? Was that counted as a return to the point of the fumbe + a turnover (which was covered in your other diary)?

by gtne91 on Jul 6, 2008 10:34 PM EDT 0 recs

Onside kicks

How were these done? Obviously, ones recovered by the receiving team count as a long return. But what about ones recovered by the kicking team? That should be a big positive to the kicking team, right?

by gtne91 on Jul 6, 2008 10:35 PM EDT up 0 recs

good questions...

1. I should look into the ‘baseline’ idea to see if the teams who kicked off often were indeed penalized less.

2. Botched and muffed punts were counted both in the turnover and special teams departments. Basically a bad snap resulting in a 30 yard loss and change of possession would be both a turnover and a -30 yard punt. A muffed punt would count as a return to the point of the fumble. In that case, the punt would have an extremely high point value. Here’s an example from the Rutgers/Army game (it’s the first Punt/Fumble I found).

Rutgers punts from its 34 yard line. The starting point value for the punt is 1.566 (that’s the point value of Army having the ball on Rutgers’ 34 yard line).

Army muffs the punt, and Rutgers recovers at the Army 8. The ending point value is -4.600. Why? Because Rutgers having the ball at the Army 8 is worth 4.600 points. The negative value is because these values are for the receiving team.

1.566 – (-4.600) = 6.166 points on the punt. The higher the better for the punting team, and it doesn’t get much higher than that.

3. I tossed around a lot of different ways to gauge on-side kicks, and here’s what I ended up with:

If the receiving team recovers it at the kicking team’s 40 yard line, it’s treated as a normal kickoff returned to the 40. The kicking team knows the risk it’s running on an onside kick—it’s giving away some points if it doesn’t recover.

If the kicking team recovers at its 40 yard line, it’s treated as a return to the 40 and a fumble. It’s not a turnover (though you could actually make a decent case for it being treated as one), but as far as point values go, that’s how it’s treated.

Let me know if you have more questions…I love answering them…

http://www.rockmnation.com
Thrust nunchuk upward!

by The Boy on Jul 7, 2008 8:54 AM EDT up 0 recs

Baselines

First, sounds like you handle fumbled punts/onside kicks the way I would expect.

As far as the baseline goes, the avg kick may be the best (making an avg kick worth zero points), but there are some other baseline options that could be used. Your current idea of baselining at the kicking teams 30 just seems wrong, for some reason, to me. Here are some other baselines that I think would work better:

1. A return to the 35 (or is it 40 now?). The baseline is a kick out of bounds. With the exception of those cases where the team penalizes 5 yards instead, the kicking team can “play it safe” and kick oob. Making that a zero point return, the kicking team can choose to gamble and kick it in bounds and go from there. My problem with this is that I think a return to the 35/40 is a negative result for the kicking team, so should be treated as such.

2. A touchback as baseline. If you tackle the other team inside the 20, the kicking team gains a small value of points. A return outside the 20 gains pts for the return team. Like the current system, the avg will be positive for the return team, meaning a team that kicks off a lot (and doesnt have a guy who regularly puts it out of the end zone) may be disadvantaged. But, much less than the current method.

My only problem with the avg return idea is that it changes from year to year. I like the idea of a consistent baseline to measure against.

Thoughts?

by gtne91 on Jul 7, 2008 10:24 AM EDT up 0 recs

well, there are two things at work here...

...#1 is coming up with a way to actually evaluate the kicking teams. #2 is how to apply an accurate point value to the kicks themselves. For #1, there’s a very interesting debate to be had between using my overall point value averages, or the 35, or touchbacks, as you mentioned. For #2, though, my hands are kind of tied to what the actual point value of the kick was. Either way, I guess, if I’m using averages of any kind, you can see who was best and worst at kickoffs…but I’m open to debate on the technicalities…

http://www.rockmnation.com
Thrust nunchuk upward!

by The Boy on Jul 7, 2008 10:39 AM EDT up 0 recs

Just had a thought

What if a KO is treated as if it were a turnover, but with no loss to the kicking team. On turnovers, you calculate 2 parts, the expected value of field position of the offense that is given up and the expected value of starting position by the recovering team. So, the value to the return team would be the value of the starting position but with a zero for the loss of field position by the kicking team. That has some problems in that it still leads to a kicking team always having a negative value for the kick (unless they recover an onside kick), but I dont think that is a problem.

If you treat a TD as a combo of a TD + kickoff, basically what you have is 7 pts to the offense/kicking team plus a turnover. Which is what scoring is.

Take the following example. Team A drives down the field and score a TD. They kick off, team B returns it for a TD. B kicks off, touchback. This scores the same as if A had never scored in the first place and only drove to the 20 so far. Its the dividing it up that becomes interesting.

A would have 7 to the O for the initial TD. -7 credited to special teams. Then +whatever a return to the 20 is worth for the special teams.

Now, that may not be all that useful for measuring the per game value of a kickoff team. It works fine for per kick avg. But any baseline works for that.

Maybe Im oversimplifying your system though.

by gtne91 on Jul 7, 2008 11:21 AM EDT 0 recs

Great work, Bill

The only question I have, related to the Army example, is whether counting those points as ‘plus’ for the punting team is really accurate—it’s certainly a negative for Army, having fumbled the ball inside its own ten, but all Rutgers had to do is fall on it. The huge number resulting from that doesn’t really reflect RU’s ability as a punt coverage unit. I’d say it distorts it, actually.

I don’t know if you can do anything about that, though.

Otherwise, the idea of ‘expected points’ continues to work very well.

by SMQ on Jul 7, 2008 12:08 PM EDT 0 recs

On second thought

You could treat the recovery (from Rutgers’ perspective, anyway) the same way you would had Army just made a fair catch at the 8 and took over there on offense. It just doesn’t seem to me that, if the the goal is to quantify performance, that a punt team deserves points that it didn’t create, i.e. falling on a fumble. The points should count against the fumbling team in full, but not for the coverage team.

by SMQ on Jul 7, 2008 1:22 PM EDT up 0 recs

There is a skill to falling on a fumble

check out last year’s GT-UGA game. If Tech can fall on any of the fumbles in the end zone, including IIRC, at least one on a punt play, that game could have turned out different.

by gtne91 on Jul 7, 2008 1:30 PM EDT up 0 recs

this speaks to the same...

...”evaluation” versus “points” argument I mentioned earlier. That play more or less “created” about 4 points for Rutgers, and it’s documented in that way. However, maybe Rutgers did absolutely nothing right on the play…maybe the ball bounced directly into an RU player’s hands and stuck. Rutgers should maybe not get credit for it, but we’re entering a major gray area here…say a running back trips and falls in the backfield on 4th-and-1 without a finger laid on him. It still counts as a “stop” for the defense, right? For stats, it has to stay pretty black and white…with the idea that luck will balance out in the end. Maybe the exact same thing happened to Rutgers later in the season? Rutgers punted over 50 times throughout the course of the season, so their overall average is still going to basically represent how they covered punts…the one muffed punt aside…

http://www.rockmnation.com
Thrust nunchuk upward!

by The Boy on Jul 7, 2008 1:52 PM EDT up 0 recs

They’re relatively rare, but I’d love to see the statistics on blocked kicks factored in. Given the massive momentum shift those tend to bring, they have an influence that tends to be all out of proportion to their frequency. Even at Virginia Tech, Frank Beamer’s teams only accumulated 102 blocked kicks in his first 200 games as a head coach at Tech.

by Alaska Hokie on Jul 7, 2008 11:39 PM EDT 0 recs