Monday, December 28, 2015

Armed Forces Bowl Preview

I wanted two things this holiday season: I wanted a Cal Football bowl game. Check. And I wanted a Red Ryder BB Gun. No, just kidding. I wanted said bowl game to take place on the weekend, or in the evening, or on one of the numerous days I have off from work. So naturally, we end up in a bowl that kicks off at 11:00am on a normal workday Tuesday. I honestly feel like we got jobbed a little bit, but whatever. No use crying over spilt milk right? We can only play the opponent they put in front of us.

Like the other service academy schools, Air Force’s offense is heavily centered on the run game. With highly strict physical requirements for student athletes, Air Force is generally unable to recruit the kind of massive offensive and defensive line talent other schools covet. This means they will almost always find themselves at a disadvantage in the trenches. Their option offense (commonly referred to as “The Veer”) is designed to minimize that disadvantage, which is both totally obvious and totally brilliant.

This next bit might not seem obvious or brilliant, but just stick with me. The Air Force Veer offense often starts by leaving two defenders completely unblocked. These defenders are called “keys.” The quarterback is responsible for reading both “keys” and reacting appropriately. From what I can tell, the Veer almost always has two backs in the backfield, and an option to send one of those backs straight up the middle. If the first unblocked key is not in position to stop this running back, the QB will give it to him in the hopes he can gain a chunk of yardage through the traffic in the trenches. But if the first unblocked key moves to stop this running back, the quarterback keeps the ball himself and rolls to the outside, while the other back rolls with him. From there, you have the QB and the second RB against one defender (if the offensive line has blocked the play correctly). Think of it like a 2-on-1 fast break in basketball. The QB can keep the ball and run himself or pitch to his RB, depending on which one the defender attacks (and by all means he should attack SOMEONE, there is nothing worse in this situation than indecision).

I know this is a lot to try to explain without video evidence, so take a look at a couple different plays from the Mountain West Championship game that matched Air Force against San Diego State:

https://www.youtube.com/watch?v=GUkTg9QkYco

Start at 22:15. The first key on this play is SDSU #99, he jumps inside to fill the hole where Air Force’s first RB is headed. The QB sees this, and pulls the ball to keep it himself. He starts to roll to his right, but the second key (#27) immediately breaks down towards the QB. The pitch goes to RB2 who races in for the 8-yard score.

Next is 38:30. SDSU #40 actually does a pretty good job of trying to contain both the QB and RB1, but he ends up just a half step too close to the RB1 option. The qb wisely pulls the ball and scampers around the first key. #42 decides he’s cutting off the RB2 option (look how far outside he ends up!), so the QB simply heads up-field and gains solid yardage.

Jump to 48:05. Good old #99 is back again here, and he gets through the line quickly. Unfortunately, he ends up too far away from RB1’s running lane. The rest of the offensive line has all blocked down towards the bottom of your screen, so the QB hands off to RB1 who hits the line with a head of steam and breaks through for a huge gain.

Don’t get too intimidated here, these were some of the “best case scenario” examples I could find. But I also came across some not so good ones.

At 39:11, SDSU brings the heat and 3 defenders are in the QB’s face before he can even make his reads. RB1 would have been stuffed on the play, and the QB doesn’t even have a chance to get outside with RB2. Pressure and poor blocking blew up this play before it ever started.

Lastly, at 46:47, we see both a great defensive play and a poor read by the QB. #40 for SDSU gets into the backfield and appears to seal the edge (where the QB would attempt to roll). But if you watch closely, he is already hedging towards RB1 before the QB even hands him the ball. Thinking that his roll out lane is cut off, the QB hands off up the middle, but he has guessed wrong. #40 has already jumped into THAT lane and stuffs the play for no gain. Not only can we tell that the incorrect read was made because the play was stuffed up the middle, but also because there is SO MUCH ROOM on the nearside of the formation. QB/RB2 would have had tons of space for a big gain if the QB had kept the ball on his initial read.

If it hasn’t been made abundantly clear at this point, Air Force is going to run the ball. They’ll run on first down, they’ll run on third down. They don’t really care about convention. But that doesn’t mean they won’t pass at all. Every once in awhile, they’ll pop off a play action pass. And when they do, it usually doesn’t go well for the defense. According to Football Outsiders’ S&P+ ratings (super convoluted ratings system, don’t sweat the details), Air Force has the single most explosive passing offense in FBS. Of course it’s a small sample size, their starting quarterback only attempted 134 passes all season (For reference, Jared Goff attempted 492). With Cal’s suspect secondary though, the deep threat cannot be ignored, even if it only comes up a few times throughout the game.

There isn’t a perfect solution to stopping Air Force’s Veer. The obvious big picture goal is to force Air Force into 3rd-and-long situations. Despite their potential explosiveness through the air, their pass game simply isn’t consistent enough to drive down the field in chunks. But getting them into obvious passing situations is easier said than done. How do you stop the run, even when you know the run is coming? It seems to me like the best path here is NOT to stack the box and attempt to snuff out the triple option before it gets started (like SDSU did in the “no chance” example). When it works, it’s great. But if you guess wrong or miss a tackle, you could be looking at a monster gain for the offense. It seems like it may be more prudent to try and limit the offense to short gains. Plug the middle and push everything towards the boundaries. Make them go east and west before they can head north and south. Keep the options in front of you. Stick to your assignment. Make them beat us 3 yards at a time if they can. But don’t let them break the big play. Our defense has been playing this way against the pass all year. Why change now?

The best thing that can be said for Cal is that they have had over a month to prepare for this very specific offensive scheme. Cal can realistically simulate what the triple option may look like, and run through the ways to defend it. On the other hand, Air Force can prepare all they want for Cal’s offense, but I’ll bet they don’t have a way to realistically simulate playing against Jared Goff.

And really, there isn’t all that much to discuss in terms of Air Force’s defense. Their run defense is solid and will limit big plays, but you can get 3-4 yards a pop if you need it, just because your offensive line is probably noticeably larger than their defensive front. Their pass defense is definitely susceptible, Goff should be able to get the looks he wants. While execution is never a given with Cal, I’m not too worried about seeing our offense shut down. If Cal loses to Air Force, I doubt it will be because we couldn’t score.

The danger is evident. Think about the Texas game. Think about Texas’ QB galloping through our defense like a majestic wildebeest. Our run defense has not exactly been stellar throughout the season, and that’s without them having to worry about all the different looks Air Force will throw at them. I won’t be surprised if Air Force runs all over us in the early part of the game. It’s going to take our guys some time to settle in against a scheme they’ve probably never actually played against before. But if we can respond on offense, and make some defensive adjustments throughout the game, I believe our edge in pure athleticism will be enough to get us over the hump. 

Cal pulls away late on the arm of Jared Goff to win their first bowl game since 2008.

Saturday, December 19, 2015

Art of Aggression

The “Will-He, Won’t-He?”, “He Said, She Said”, “She Sells Sea Shells by the Sea Shore” extravaganza that was “Sonny Dykes Extension Watch 2015” didn’t turn out to be all that extravagant. First of all, it lasted just 12 days. The final football game of Cal’s regular season was November 28. Dykes’ extension was announced on December 10. I’m pretty sure longer and more tense negotiations have taken place. And secondly, I highly doubt whether there was ever real discord between Dykes and the Cal administration anyways. Cal offered, Dykes countered/used leverage gained by showing interest in other possible coaching jobs. Welcome to every business transaction ever. Personally, I never worried that the two parties wouldn’t reach an amicable resolution.


The deal itself seems like a fair one for both sides. Sonny is extended through the 2019 season, and will average $2.825 million over the next four years. This takes Dykes from being the lowest paid Pac-12 head football coach to middle of the pack, which parallels perfectly the team’s progress from worst in the Pac-12 to middle of the pack over the last 3 seasons. The administration also agreed to shell out a little more for assistant salaries. This may not sound terribly important, but in the context of Cal Football, it may be the most integral detail. Not only was it reported to be an important sticking point for Dykes, it also shows a willingness to compete monetarily with the bigger names in the athletic realm. Cal still won’t match the amounts UCLA, Oregon, and USC pay their coaching staffs (USC is based on pure assumption, as a private institution they are not required to publicize salaries - so of course, they do not), but with these increases, they have closed the gap significantly.


I am surprised, however, by this one-year rollover clause that was thrown into the deal. It states that Dykes will automatically be extended “an additional year for each year he wins seven regular-season games and the team achieves an annual Academic Progress Rate (APR) of 980 or higher”.


Let’s knock out this APR bit first. If you are unfamiliar, APR basically measures an athletic program’s ability to keep student athletes in school and on track to graduate. I know 1,000 is a perfect score. I’m not sure exactly how low a score can be, but in 2013 Cal Football registered a shockingly low 923 APR score, that was widely referred to as “an embarrassment”. This year, Cal Football’s APR is expected (official scores are released next Spring) to be 997. I know it has been said many times, but I cannot stress enough how much credit Sonny deserves for the academic turnaround he has engineered, and I have no issue with the inclusion of the “980+ APR” corollary.


The more curious part of the equation is the regular season wins. Doesn’t seven wins feel like setting the bar a little too low? We won seven games this year, and we weren’t even that good! We didn’t resolve any of our “beating California rivals” issues, and we were never really in the race for the North division. As long as I’m understanding this correctly, Sonny Dykes could coach the team to this exact same (somewhat disappointing) Cal Football season every year, and he would remain our coach forever. I won’t say that this clause is a completely awful idea, but I also won’t deny that it raises my eyebrows just a smidge.


But, I digress from the gory contract and legal details. You don’t care about all this. This is not Bob Loblaw’s Law Blog. With the uncertainty of Dykes’ contract situation safely behind us, let’s get into the actual football stuff, shall we? Let’s get right to the criticism!


I have publicly criticized (I say that like anyone cares) Dykes’ decision making and level of aggression throughout the year. I recall watching a bunch of strange short distance play calls, lots of punts on reasonable-to-go-for-it 4th downs, and a few ill-advised fakes sprinkled in for good measure. At certain points, it truly felt like Sonny chose to do the opposite of whatever I screamed at him to do through my television screen. I understand that what I want in the heat of the moment is not always the smart play, but hopefully you can understand my frustration.


Measuring whether being aggressive in certain situations is a good or bad thing is a tricky task. It’s all too easy to deem a successful choice as a good one, and a failed choice as a bad one. But that’s not how football works. It would be impossible to go through every scenario and decide definitively whether Dykes made the right call or not. It’s simply too hard to project whether or not an imaginary outcome would change the reality of the final score. In fact, I have heard several people offer the opinion that every choice made in a win is a good one, because they all led, and contributed in some way, to a victory. This is total BS my friends, but everyone is entitled to their opinion… I suppose.


What we can do is look at raw, overall statistics, without imposing our own subjective viewpoints on outcomes. I have compiled data for all Pac-12 teams from the 2015 regular season, tracking any 4th down situation in which they had 5 or fewer yards to gain, that took place between their own 40 yard line and their opponent’s end zone. Let’s refer to these as “opportunities”. I figured that these opportunities will, most of the time, present reasonable options to go for it or not, but I do have to admit that this part IS subjective. You could certainly argue that the criteria for yards to gain should be shorter or longer, or that it is only reasonable to go for a 4th and 5 if you are on your opponent’s side of the field. These are just the criteria I landed on. My defense is as follows:


It’s my blog. Deal with it.


So let’s take a look at how each coach approached these 4th down opportunities. The table below shows how many opportunities each coach faced, and how many times he chose to go for it. The made up stat “Aggression %” is simply the resulting percentage. Of course, specific context is left out here - take with just a few grains of salt. We don’t know how many of these were obvious go-for-it scenarios and how many were obvious punt situations, but it presents an interesting overall picture of aggressiveness.


Coach
Team
Opportunities
Went For It
“Aggression %”
Leach
Washington State
38
23
60.5%
Rodriguez
Arizona
36
20
55.6%
Helfrich
Oregon
33
17
51.5%
Shaw
Stanford
28
14
50.0%
Sarkisian/Helton
USC
30
15
50.0%
Petersen
Washington
30
15
50.0%
MacIntyre
Colorado
38
19
50.0%
Dykes
Cal
31
15
48.4%
Whittingham
Utah
35
13
37.1%
Andersen
Oregon State
20
7
35.0%
Graham
Arizona State
42
12
28.6%
Mora
UCLA
26
7
26.9%
AVERAGE
45.7%


Well, color me surprised! Turns out Coach Dykes isn’t as ultra-conservative as I would have you believe. Yes, he’s towards the bottom of the table, but he is very close to the 5 coaches ahead of him. In fact, if he had gone for it one more time, he would have ranked third (16/31 = 51.6%). Remove Mike Leach from the equation (because seriously guys, he’s a crazy person), and Dykes is within shouting distance of all his coaching counterparts.  At the very least, he’s not Jim Mora. Holy guacamole, what are you doing Jim? For someone with such a naturally aggressive personality, you would think he might roll the dice more than a quarter of the time. Must be he’s got a lousy run game and a great punter (there’s my patented sarcasm again).


Furthermore, Sonny comes in above the conference average. He is actually slightly MORE aggressive than the average Pac-12 coach, given any particular “4th down opportunity.” I’m not excusing him from the field goal parade in Palo Alto, or the abhorrent fake punt in Pasadena. But on the whole, he fits right in with normal convention. Is it possible that Sonny Dykes is more levelheaded than I am?


However, as the saying goes - what goes up, must come crashing down to earth in a fiery wreckage causing widespread devastation and despair. Here are the corresponding conversion rates for each team (how many times did the team gain a first down or touchdown, out of the total times they went for it on a 4th down opportunity).


Coach
Team
Went For It
Converted
Conversion %
Shaw
Stanford
14
12
85.7%
Mora
UCLA
7
6
85.7%
Whittingham
Utah
13
11
84.6%
Sarkisian/Helton
USC
15
12
80.0%
Helfrich
Oregon
17
13
76.5%
Graham
Arizona State
12
9
75.0%
Leach
Washington State
23
16
69.6%
MacIntyre
Colorado
19
11
57.9%
Dykes
Cal
15
8
53.3%
Rodriguez
Arizona
20
10
50.0%
Petersen
Washington
15
7
46.7%
Andersen
Oregon State
7
2
28.6%
AVERAGE
66.1%


Ok, I was overreacting, it’s not that bad. We weren’t “Oregon State” bad at converting 4th down opportunities. But we still weren’t great. Unlike the previous table, there is a wide gap between the top and bottom. While Sonny’s low position in the first table did him a disservice, this one might actually give him too much credit. You’ll notice he is much closer to the 11th ranked coach than he is to the ones at the top of the group. He is also significantly below the conference average. If Cal had been able to just convert at that 66% average, they would have extended 2 additional drives (and potentially added 2 more viable scoring chances). Considering Cal lost two games by 6 points (and played Stanford close for 3 quarters), who knows what those two extra chances could have yielded?


But look at all those teams that are WELL above the average. Stanford, UCLA (Earth to Mora: You're converting at 86%, GO FOR IT MORE), Utah, USC, Oregon: All 8-game winners. Weird how all these successful programs appear at the top of our conversion table, huh? While all the not-so-successful ones are at the bottom. It almost makes… too much sense.


What if we were to group the teams into the top 6 teams in the conference, and the bottom 6, in terms of 2015 record? The top is comprised of Stanford, Oregon, Washington State, USC, Utah, and UCLA. The bottom is comprised of Cal, Washington, Oregon State, Arizona State, Arizona, and Colorado.


Here’s what we find:




Opportunities
Aggression %
Conversion %
Composite Record
Top 6 Teams
190
46.8%
78.7%
53-21
Bottom 6 Teams
197
44.7%
53.4%
31-42


Hmmm, similar number of opportunities. Similar levels of aggressiveness. But my goodness, look at the difference in conversion. That 25% gap represents a whopping 23 extra conversions on the season, 4 extra scoring chances per team. Again, during a short 12-game season, that is meaningful. The teams that won more games held a huge advantage in extending drives and creating more scoring opportunities. Imagine that.


And gee, that bottom 6 conversion average looks familiar. Oh, that’s because it’s nearly identical to Cal’s own conversion rate. These are subjective statistics people. You have zero arguments against the fact that Cal converted these 4th down opportunities like an average bottom 6 team in the Pac-12. And thus, Cal finishes as a bottom 6 team in the Pac-12.


This study represents a very small piece of the pie. By no means will 4th down conversion rate (under specific criteria, to boot) always be an indicator for overall success. But it’s definitely something to chew on. During a season in which we felt so close, yet so far away on more than one occasion, it’s hard not to imagine what a better 4th down conversion rate may have turned into. Is it possible that our poor short yardage run offense cost us a win or two this season? Absolutely. Alas, that is a fruitless game to play. The world will never know.


Bowl preview is coming soon.

Go Bears.