Saturday, December 31, 2011

My Ideal Offense

A while back over at MGoBlog, there was some discussion about what the ideal offense is. This started as a mailbag question for Brian, then migrated to the board where everyone described their ideal offense. I described my ideal offense as:

I would have an offense based off of a lot of option plays. From shotgun and pistol, I'd mix up Read Option, Veers, Inverted Veers, etc. on different linemen so they couldn't cheat  on it. I'd also throw in some speed option. I'd use Pistol, I Form, and Wishbone to do Triple options.

Passing would look like a passing spread with some WCO concepts mixed in. Particularly, pre-snap reads to determine WR routes. It might even be that a play call is actually 2-3 plays and the defense determines which to run.

The RB would be a Brandon Minor type where if he would consistently use the lower numbers in the box to break a tackle and run for another 30 yards. My QB doesn't need to be a burner, but needs to be mobile.  It's more important that he can throw accurately, can make good reads, and does really good fake handoffs for the options and PA. I'd have some utility-type players that can play multiple positions, e.g. TEs that can play receiver, HB, and occasionally FB as well.

There would be no-huddle with lots of pre-snap motion, a lot of times moving utility players around to completely change formations. Maybe we start out in Wishbone and then motion to 3WR shotgun or 3TE I Form.

I play a lot of football on my Xbox so this question has stuck around in my head for quite some time, especially since this is very similar to the offense that I run in the game (Option, Power, Quick passes). After a while, I realized that, using motion along with changing the dive and pitch backs, you can run almost any type of option play out of a single formation.

I sat down and did a lot of reading (Thanks, Smart Football) on lots of different styles of offense. I found out that there is a running offense very similar to what I had come up with in my mind: the Ski-gun. If you’ve never heard of it, that’s fine, because it’s only been around for a few years and is still only utilized at the high school level, as far as I know. Basically, Muskegon (get it? Mus-ski-gun…) High School used to be a standard triple option team out of the Flexbone (more on that later). It’s been a while since I did my research on it, but I believe that Muskegon was having a hard time passing out of the Flexbone, so they decided they needed to spread the field a bit and put the QB in shotgun/pistol to make it easier for the QB to see the field and make his reads. They, of course, needed to do this without getting rid of their offense in the Flexbone.

Enter Ski-gun. The Ski-gun is basically the Flexbone offense, but out of the pistol. For those unfamiliar with the Flexbone, it is a running formation based on the triple option used by Georgia Tech, Navy, Army, and a few other teams. It involves a dive back, about 4 yards behind the LOS, two slot backs (basically, speed backs) as wings on either side of the line, and two WRs covering the wings. It looks like this:

A good amount of the time, one of the SBs will go in motion behind the dive back. The ball is snapped shortly after the motion starts. The dive back heads toward the back of the guard in the direction of the motion. The QB will read the DE on that side of the line. If he goes for the dive back, the QB keeps and runs with the SB for a pitch option. If the DE stays to guard the pitch, the QB gives the ball to the dive back for what should be an easy 3+ yards.

The Ski-gun however looks more like a pistol spread. The QB is now 2 yards behind the center, with the dive back two yards behind him (he has not moved). The slot backs are now actually in the slot, 2 yards outside the tackles and aligned with the QB. That’s the only difference in formation. When the slot goes in motion, he runs straight toward the QB until the ball is snapped to prevent the defense from keying in on his motion. After the snap, he can continue his motion in front of the QB to be the effective dive back in a zone read triple option, or he can go the usual route for the veer triple option. Essentially, this is what you end up with:

BaseFormation

This is the base of my ideal offense. Pistol spread with a slot going in motion on a majority of snaps, allowing for a variety of different types of option and power plays. There are a few things that I like about this formation to begin with. First, it’s completely balanced. Aside from the motion, there is no indication which side is the strong side. Second, Pistol keeps the QB’s eyes on the field during passes and allows the dive back to begin running straight at the line at the beginning of the play. Third, the motion or lack of motion allows for a small set of formations to be run out of one formation. Without motion, you have the obvious formation of Pistol Spread. Snapping the ball as the slot gets to the left tackle is like running out of the Pistol Weak Slot. Snapping the ball a bit later, allows you to run out of the Pistol Strong Slot. And waiting for the motion man to completely cross the formation moves you into a trips formation.

Once you look past the formation itself, the number of variations of option plays that can be run out of this are incredible. Have the motion man go in front of the QB and the RB go the opposite direction for a zone read triple option. Have the RB go in the same direction as the motion man to utilize the RB as a lead blocker. Do the same thing, but switch to a power blocking scheme instead of a zone scheme to run an inverted veer with a pitch option from the motion man to the RB (I thought up this play, then saw Nebraska run it against Michigan). This is intended to be the intro to an in-depth series into my ideal offense, so I will be going over a lot of these plays in the future.

Here are some videos of how Muskegon High School utilizes their new formation.

The Passing Game. Admittedly, the passing game is an afterthought in my offense. However, when I do pass, I don’t just want it to be 5-10 yards at a time. My thought is that there are 2 ways to spread a defense: horizontally and vertically. Essentially, there are three parts of the field.

The first is the box, which includes the entire offensive line, everything behind them, and everything within the first 4-5 yards in front of them, i.e. defensive line and line backers. Sometimes, you will hear announcers talk about 7, 8, or 9 defenders in the box. This is what happens when you don’t stretch the field. Forcing them to move outside of the box by throwing deep, or throwing/running to the sidelines pulls men out of the box and into one of the other two parts of the field

The second is the flats, which is the area between the box and the sidelines, including WRs, CBs, and any defenders covering the slots. Short passes or runs to this part of the field stretch the field horizontally. If you are able to pull a LB out of the box to cover a slot receiver, you then gain an advantage in the running game because they now only have 6 men in the box. Unless, of course, they put a safety in the box to replace him, which brings us to the third part of the field.

The third part of the field over or behind the defense. This is everything beyond the box. Passing deep is the only way to attack this and opens up the field vertically. If you are able to complete passes more than 10 yards down the field, the defense will not be able to creep the safeties into the box to help stop the run. They also won’t be able to put them directly over the slot receivers to allow a LB to stay in the box.

It’s important to stretch the field in both directions because you want to spread the defense thin and prevent them from cheating to stop the run. This is like one of those triangles that you see where you can have two of the options in the triangle, but you can’t have the third. If you don’t stretch vertically, the defense can play close to the line to stop the run and the passes to the flat. If you don’t stretch horizontally, the defense can play the safeties back, keep the linebackers in the box, and have the linebackers drop back to cover the slots when they determine it’s not a run. If you can’t run, they will take everyone out of the box and shut down the entire passing game.

Anyway, when I began thinking about what was important in the passing game, I decided that the most important thing is to keep the playbook simple. My offense is based around the run and practice would need to be focused on that. Keeping the passing plays to a minimum and simple will allow for more time to be spent perfecting the running game. You are also running with your QB. You don’t want the whole offense to shutdown if he gets injured. Keeping it simple allows for a backup to come in and move the ball effectively.

The problem with keeping the playbook small is that there isn’t a whole lot of variation when you pass. I decided to see how other offenses deal with this problem. Specifically, I wanted to see if there were any concepts in those offenses that are essentially option through the air, where the plays change as the defense reacts. This would help keep the play count and complexity for the players down, but increase the difficulty to defend the plays. There are two offenses that I know of that specifically spread the field to pass. Those are the Air Raid from teams such as Texas Tech (while Mike Leach was there), Oklahoma, and WVU and the Run ‘n’ Shoot or Run ‘n’ Gun from teams such as Oklahoma State and Houston.

I first turned did some research on the Air Raid because I was more familiar with it. There were some things that I liked about the Air Raid offense, like Mesh, which involves the slots crossing to scrape the defenders, then sitting between zones, or continuing straight if it’s man coverage. However, the Air Raid offense seems more like a lot of different plays, which are slight variations on each other like X Follow, Y Follow, and Z Follow. Those involve calling the correct play before the ball is snapped. I would prefer more of a way to call Follow, then the alignment or the defensive play determines whether it’s X, Y, or Z.

I did some more reading on the Run ‘n’ Shoot and this appeared to be exactly what I was looking for. Basically, there are a few main concepts in the routes and plays are built around those concepts. Each of these concepts is one of the receivers reading the defense and running toward the weak point of that defense to exploit it. By adding, say, 4-6 passing plays to the playbook, you can effectively stretch the field vertically. The best part is that the Run ‘n’ Shoot almost always runs out of the spread look or trips and it started out by almost always motioning from one to the other. This means that many of the concepts like the motion man running to the flat for a hot route directly apply to my offense.

My Ideal Offense. So, my ideal offense is basically a hybrid offense that is meant to incorporate the best parts of the Flexbone Triple Option offense and the Run ‘n’ Shoot offense. The lone formation is the Pistol Spread with the QB 2 yards behind center, the dive back 2 yards behind him, the slots 2 yards off of the tackles and aligned with the QB, and 2 WRs covering the slots.

The QB is ideally an accurate, smart QB that is just fast enough to make a defense respect him if he keeps the ball to run. Excessive arm strength is not a necessity. Neither is raw speed, although he can’t be a statue. A few examples would be Drew Brees, Tate Forcier, Chase Daniel, Drew Henson, etc.

The dive back (D in my diagram above) would be somewhere between a Power Back and a Full Back. He needs to be able to break a tackle up the middle. This offense is going to give some one-on-one opportunities for the dive back and he needs to be able to punish that defender and get extra yards. A few examples would be pretty much any Wisconsin starting RB, Mark Ingram, BJ Askew, and Chris Perry (except bigger).

The slots are speed backs. They need to get to the edge, be able to catch, and handle one-on-one situations with DBs. Not only are they motioning, though, but they will be running a deep route a good amount of the time. There is also a fair amount of responsibility in reading the defense on passing routes and any potential trick plays, so a certain level of football smarts is required. This is where you put Fitzgerald Toussaint, Steve Breaston, and Percy Harvin.

The X and Z receivers are typical X and Z receivers. They need some speed, height, and jump ball ability. They are the main components in stretching the field vertically. This system is going to get them a lot of one-on-one matchups against CBs or safeties, so either being able to get behind them or being able to out jump them for the ball will be a necessity. Pick a star receiver and put him here.

There are about 3-4 read plays with motion, each with a few variations to get Triple Option, QB lead, or RB lead. There are a couple of running plays without motion like RB Draw, QB Lead Draw, and Veer. We’ll say there are 4-5 passing plays from the Run ‘n’ Shoot, Mesh from Air Raid, and, of course, the bubble screen. Throw in a few PA plays, which will look like an existing running play, but the routes will be one of the passing plays. All-in-all you are probably looking at 5-8 running plays, 6-8 passing plays, and 4 or so PA plays to give you 15-20 plays, all of which would be practiced to perfection.

As a spread team, they would always be in no-huddle. Everyone reads the plays from the side line. If you want to hurry up, you hurry up. If you want to kill some clock, walk to the line and take your time getting set.

The Plan. As I briefly mentioned, my goal is to make this into a series. I plan on going a little deeper into the concepts of my ideal offense, maybe a little history and evolution of some of the concepts, then diagramming a new play concept in each post. I’m thinking it will end up being about 10 posts long. It really depends on how much I plan to cram into each one.

Sunday, December 4, 2011

Top 25: 2011 Post-Season (Bonus: Playoff Proposal)

It’s been a while since I have updated my rankings. That’s just due to other hobbies taking priority for the past few weeks.

To the ranking. I’m in a bit of a predicament with my ranking system. There are two components to the rankings. One is margin of victory combined with the opponent and the other is win-loss percentage. Last year I had an issue with Auburn being ranked 4th because they didn’t score enough points, so I put less emphasis on margin of victory. This is the formula that I have been using this season, and it looks mostly correct. The problem that I have now is that Oklahoma and Houston are both ranked too high, in my opinion.

The problem is that a loss does not discredit you in this system, but you only have wins to make your case. So, if a 9-win Oklahoma team can “impress” enough in 9 games to be ranked ahead of a team with 10 or 11 wins, then so be it. However, that sword cuts both ways. Houston should not be ranked 6th. I tried removing emphasis on the overall record, but that sky rockets Oklahoma to #2, which is obviously not correct. I tried placing more emphasis on record, but it seemed to be too much emphasis on record. Oklahoma dropped to 13th, but Southern Miss moved into the top 10 (already 12, which is a product of beating down the #6 team). I tried some other tweaks like allowing losses to affect a team’s points, but that didn’t do much of anything at the top. In the end, I decided it was best to leave the formula in tact to keep continuity between the weeks and accept that it’s not perfect (I won’t say it’s never going to be perfect because I will find a way).

That said, here is the poll. Rematch validated, unfortunately. For what it’s worth, before the OSU-OU game, I calculated that OSU would need at least 36 points to pass Bama in my ranking. They scored 34. Had they not let OU score in the end, they’d be number 2 here, and possibly in the BCS. I’ll get to my feelings about this in the bonus section. In none of my tinkering did Oklahoma State ever move up to #2. There were cases where Oklahoma or Boise jumped Bama, but those were obviously not good formulas. I’ve added records to the output. These are records against BCS teams, so some teams are missing a win.    

1. LSU(12-0) 1.6414908946071
2. Alabama(10-1) 2.0346791039512
3. Oklahoma St.(11-1) 2.04930723097138
4. Boise St.(11-1) 2.06727988816191
5. Stanford(11-1) 2.18027952897559
6. Houston(11-1) 2.3320805451465
7. Oklahoma(9-3) 2.33229548184857
8. Southern California(10-2) 2.35678754689992
9. Wisconsin(10-2) 2.42456116254947
10. Michigan(10-2) 2.46666250026333
11. Arkansas(9-2) 2.53513831475187
12. Southern Miss.(10-2) 2.53915931843441
13. Kansas St.(9-2) 2.54534227302646
14. Oregon(10-2) 2.59692398706121
15. South Carolina(9-2) 2.64366999396126
16. Virginia Tech(10-2) 2.71186798147264
17. TCU(9-2) 2.8086999231112
18. Georgia(9-3) 2.81198896593114
19. Michigan St.(9-3) 2.92015263969457
20. Baylor(8-3) 2.93054569371464
21. Nebraska(8-3) 2.94937708942283
22. Notre Dame(8-4) 2.96894103414904
23. Clemson(9-3) 3.05512967713607
24. West Virginia(8-3) 3.072811643377
25. Cincinnati(8-3) 3.13173424969166

 

BONUS: Playoff Proposal

So, I’m not happy about the LSU-Bama rematch. Pick a reason and I agree. 2006 UM-OSU. The regular season is no longer a playoff. Bama didn’t win their division, let alone conference. Both teams from one conference just means the conference championship format is wrong.

This is the time of year when everyone is posting their playoff formats. Even Craig James and Kirk Herbstreit are throwing out their opinions on the BCS selection show. So here is mine.

- 8 teams starting with the 4 BCS bowl games. I like the bowl games and think there is room to combine the current system with the new. Rounds 2 and 3 are home games for the higher ranked opponents. This puts extra emphasis on winning during the regular season. Plus, who wouldn’t love to see LSU or Bama play Michigan or Ohio State in a snow storm?

- There are no auto-qualifiers. This currently takes an emphasis off of the season. Seriously, how can you let two teams that don’t even qualify as at-large teams into the Orange Bowl?

- You cannot have more than two teams from any conference. Period.

- Any conference champions in the top 16 are automatically in. If there are more than 8, then the top 8 ranked conference champions are selected. This makes winning your conference essential if you want into the playoff.

- If there is a non-champion that beat the champion of another conference, has more wins or fewer losses, is ranked higher than the other team, and is one of the top 2 ranked teams in their conference, they replace the other team. The best example of this is if Georgia had beat LSU, Boise State would have taken Georgia’s place in the playoff because they beat them in head-to-head, had fewer losses, and were ranked higher than them at the end of the year. If two teams from the same conference would replace conference champions, the team replacing the lower ranked champion will be selected.

- If there were not enough conference champions in the top 16 to fill the 8-team playoff, add the highest-ranked, eligible (can’t be 3rd from a conference) team that has not already been selected until you are full.

Using my rankings, the teams would be:

LSU (SEC Champ)
Oklahoma State (Big 12 Champ)
Wisconsin (Big Ten Champ)
Southern Mississippi (C-USA Champ)
Oregon (Pac 12 Champ)
TCU (MWC Champ)
Alabama (#2)
Boise State(#4
)

Using the BCS, the teams would be:

LSU (SEC Champ)
Oklahoma State (Big 12 Champ)
Oregon (Pac 12 Champ)
Wisconsin (Big Ten Champ)
Clemson (ACC Champ) (this would go to South Carolina, but they are the 4th ranked team in the SEC)

Alabama (#2)
Stanford(#4)

Boise State(#7
)

 

The rules for the matchups are as follows:

- If there is a current conference-bowl affiliation and a team from that conference is selected, that team will be placed in that bowl. If there are two teams from the same conference, the conference champion will be placed in the bowl. If neither of the two teams are the conference champion, the higher ranked team will be placed in the bowl.

- Two teams from the same conference cannot play each other.

- Rematches must be avoided if possible. If total avoidance is impossible, then the rematches must be limited.

- The highest ranked team will play the lowest ranked team that does not violate the previous two rules. The pairings go on like this until all teams have a matchup.

- The second round works the same way, except the games are played at the home of the higher ranked team.

- The championship game is played at the home of the higher ranked team.

 

So, here would be the select order using the BCS.

Sugar gets LSU due to conference affiliation.
Fiesta gets Oklahoma State due to conference affiliation.
Rose gets Oregon and Wisconsin due to conference affiliation.
Orange gets Clemson due to conference affiliation.
Sugar gets Boise State as the lowest team available.
Oranage gets Bama because Bama gets the next lowest team, which is Clemson
Fiesta gets Stanford as the last remaining team.

Sugar – #1 LSU vs. #7 Boise State
Fiesta – #3 Oklahoma State vs. #4 Stanford
Rose – #5 Oregon vs. #10 Wisconsin
Orange – #2 Alabama vs. #15 Clemson

Assuming that the higher ranked teams win, we’d have LSU, Oklahoma State, Oregon, and Alabama. LSU and Alabama cannot play because they are in the same conference and have already played. LSU and Oregon have already played, so that means LSU and Oklahoma State need to play. So, the second round matchups would be

#3 Oklahoma State @ #1 LSU
#5 Oregon @ #2 Alabama

If LSU and Alabama both win, then the rematch is warranted. Alabama @ LSU

This might look like LSU is getting the shaft, but it’s just luck of the draw. If Wisconsin or Clemson win, then LSU faces them in the second round, which is fair.

 

The reason I like this system is that it puts a huge emphasis on the season and winning your conference. You better win your conference or be ranked in the top few spots or you aren’t getting in. It gives everyone a fair shot. There is no more worry about whether a 12-0 Houston deserves a shot because of their weak schedule. They automatically get in and can prove their worth over 3 of the top ranked teams in the country. If they had won their game, they are playing 3 of the top 10 (likely top 5) teams in the country.

It keeps the bowls kind of in tact. I love watching the Pac 12 and Big Ten champions face off. It would pain me to lose that.

It prevents rematches and conference matchups if possible. I don’t want to see LSU play Bama or Oregon again. I want to see if there are any other teams that might be better than LSU. Allowing Boise State and Oklahoma State both to get shots at LSU is the whole point of this system.

It encourages out of conference matchups to be good. If LSU had lost to Bama or Georgia, then they take Oregon’s spot. If PSU had won the Big Ten, Bama would have taken their spot, opening another spot for a top-ranked team. Playing tough matchups in the out-of-conference schedule does two things. It gives you a chance at some insurance in case you don’t win your conference and it increases your strength of schedule, leaving you higher ranked.

Anyone who has an argument to play in the championship game gets their chance. If you won your conference and are ranked in the top 16, you deserve a shot at the national championship. The replacement step is meant to upgrade teams if one is, without argument, better than another.

All that said, the BCS will probably just change their rules or format again to keep the money lining their pockets for a few more years. They’ll keep doing that as long as it’s profitable. /rant

Alternate Picking Proposal

Pick the top rated conference champion and then pick the all of the at-large teams ranked above them in order of ranking, skipping teams that would be the 3rd from that conference. Continue doing this until you have 8 teams. This would be LSU, Oklahoma State, Bama, Oregon, Stanford, Wisconsin, (Skip Arkansas) Boise State, Kansas State. 8 of the top 10 teams with the 2 teams being left out being #3 and #4 in the SEC. I don't think anyone outside of the top 10 has an argument to be in the tournament. Bowls would be LSU vs. KSU (Sugar), Bama vs. BSU (Orange), OSU vs. Stanford (Fiesta), Oregon vs. Wisconsin (Rose). OSU @ LSU and Oregon @ Bama in the second round. Bama @ LSU in the championship.

Monday, October 17, 2011

Top 25: 2011 Week 8

Oklahoma leaps Bama into the #2 spot. Michigan falls from their loss to MSU, but is still ranked ahead of them. Michigan should fall further due to a bye week this week. Illinois only falls 4 from their loss at home to OSU. Kansas St. replaces them and everyone else in the top 10 slides up to fill the voids.

The Big Ten has the best showing thus far with 6 teams in the top 25, although their weak schedules aside from MSU vs. Wisconsin will probably hurt that next week. Look for the winner of that game to vault into the top 5, possibly top 3, next week.

Homefield advantage drops significantly, as expected. I’m guessing it will head toward 4 points after all the games from this coming weekend, then continue toward somewhere between 3 and 3.8 like what we saw last season.

HVA: 4.50 points

Rank Prev Team Score
1 1 LSU 2.45890648786532
2 3 Oklahoma 2.5637554347037
3 2 Alabama 2.64802102648971
4 4 Boise St. 2.66292055388939
5 6 Oklahoma St. 2.68590523866395
6 7 Wisconsin 2.82016077606777
7 8 Clemson 2.93895071624342
8 11 Kansas St. 3.02556342568788
9 10 Stanford 3.1573487153316
10 5 Michigan 3.45398832508829
11 12 Houston 3.50108486847672
12   Michigan St. 3.59638932054817
13 9 Illinois 3.59676262910921
14 19 Virginia Tech 3.61901166608766
15 14 West Virginia 3.65429029484404
16 20 Southern California 3.68487175944827
17 22 Rutgers 3.7316386209211
18 18 South Carolina 3.76081537693958
19   Oregon 3.77544505695606
20 17 Arkansas 3.84350195566227
21 23 Nebraska 3.88896572844642
22   Penn St. 3.89971002077327
23 25 Washington 3.921810258696
24   SMU 4.0645567166626
25 13 Georgia Tech 4.07024293676979

Dropped out: (15)Texas, (16)Arizona St., (20)South Fla., (24) North Carolina

Conference Breakdown:

ACC 3
Big East 2
Big XII 3
Big Ten 6
C-USA 2
Pac 12 4
SEC 4
WAC 1

Monday, October 10, 2011

Top 25: 2011 Week 7

 

Here is my first ranking of the season (and this blog for that matter). I don’t have too much commentary on it this week as I’m just trying to get this out there before I start moving on with the code to produce more content. Obviously, at this point your top 3 are LSU, Bama, and Oklahoma. I think the biggest surprise is Michigan at number 5. Without looking too much into the details, my guess is that the formula likes the big wins over a relatively diverse group of teams. As the conference season continues on, I’d expect them to fall even without a loss as other teams start picking up meaningful wins. Stanford is a good example of this. I think you can see why the BCS waits until week 8 to release their first rankings, but, like I said, I wanted to get these out and I figured that they are close enough for now.

The average home field advantage (HVA) is quite high so far this year. I expect that to drop as the season wears on and we have much closer games. Last year, for comparison, ended with a 3.8-point HVA, which is a bit closer to what you usually hear. Some of the other content that I am working on is more statistical analysis, including expected margin of victory (MOV) and probability of a win based on the calculated scores. I have some basic formulas down, but I’m working to on some alternatives to see if I can decrease the uncertainty a bit more. If you have any suggestions for features or statistics, let me know and I’ll see if I can work them in in the future.

HVA: 5.36 points

1 LSU 2.80088407221375
2 Alabama 2.82525523265761
3 Oklahoma 2.90103732778685
4 Boise St. 3.01757379664688
5 Michigan 3.09446518600681
6 Oklahoma St. 3.21048272126539
7 Wisconsin 3.23106635038202
8 Clemson 3.27141358831144
9 Illinois 3.37987900100243
10 Stanford 3.45772677823538
11 Kansas St. 3.70587375482202
12 Houston 3.75508016447462
13 Georgia Tech 3.78180409782285
14 West Virginia 3.99170721067731
15 Texas 4.03840795266698
16 Arizona St. 4.18733618929933
17 Arkansas 4.27256324632263
18 South Carolina 4.30393480993576
19 Virginia Tech 4.3286948727631
20 Southern California 4.35235935579354
21 South Fla. 4.42527369946208
22 Rutgers 4.49889209952053
23 Nebraska 4.53640877223486
24 North Carolina 4.73368571924313
25 Washington 4.83997853709854