Schrödinger's cat: Is the passing game dead or alive?
Numbers say Jake should play well against LSU
Richard Linde, 11 May 2009

Although people are skeptical of a quarterback's passing efficiency rating, Jake Locker's per-game ratings seem meaningful under the light of the criticism being leveled at his throwing ability. His ratings when compared with some defensive numbers suggest he is a better passer than people think, his high throws and wild passes notwithstanding.

Over his 16-game career at Washington, Locker has completed 48.7% of his passes, having thrown 421 times and completed 205 passes for 2,574 yards,15 touchdowns and 15 picks. Those numbers translate to a pass efficiency rating of 104.68.  Last season, Mark Hartline of Kentucky posted a PE of 104.7, which ranked him 100th in the country.

Many fans think Locker isn't a good enough passer to be playing quarterback in the Pac-10 and, that because of his exceptional athleticism, he should be playing at another position. Some suggest that Locker will be drafted by the Pros as a safety.

New head coach Steve Sarkisian is moving Locker under center and using him in a pro-style offense, in spite of the fact Locker has extraordinary running skills. Out of the spread formation, Locker rushed for 987 yards in the 2007 season. He suffered a season-ending injury in his fourth game last season.

Under center in the spring game, Locker posted a whopping pass-efficiency rating of 218.89, completing 16 out of 18 passes.

However, Chip Kelly, Oregon's new coach, apparently wasn't impressed by his performance. 

"We’re not going to line up our best against the rest like the Huskies did in their spring game," he said. "They ran their number one offense against the number two and three defenses so Jake Locker could go 16 of 18 with two drops. And they had their number one defense up against the second and third offense so they could shut someone out. We’re going to compete. The number one offense will go against the number one defense. That’s the way it’s been all spring. Walter Thurmond III (starting corner) will be lined up against Jamere Holland (starting receiver), head to head and we’ll see who comes out ahead.” [Ken Woody's blog in the Eugene Register-Guard].

Incidentally, Kelly's comments made the "The Border War," supplying the hot-running feud with more ammunition.

However, according to the numbers shown in the second graph below, Jake should have a good game throwing against LSU, finishing with a PE of 120 or higher, Kelly's comments notwithstanding. Excluding injuries from consideration, there is a pattern to Locker's passing performance.

Locker's passing efficiency rating over his last 8 games, which includes four scrimmages, is 130.29. He has completed 61.1% of his passes in these last 8 outings. See Table 1 below.

With Chip Kelly's comments in mind, I guess you should probably view the numbers in Table 1 skeptically.

However, Locker's numbers in the Oklahoma game (passing efficiency 120.57) jumped out at me. After all, Oklahoma played Florida in the BCS title game last season. So I looked at the Sooners' pass defense. The Sooners had a pass efficiency defense of 118.09, ranking them 46th in the country, their pass defense being 99th best in the country. Locker's rating (120.57) seemed to be a function of the pass defense (118.09) he faced in that game.

So, I looked back at Locker's PE ratings over the 2007 season and compared them with the quality of the defenses he faced.

In Table 2 below, I try to make sense of Locker's numbers from the 2007 season, which are labeled "Schrödinger's cat" over at bearsnecessity.com. See (Give me bad numbers).

Locker's stats below mostly correlate (the 3 rows highlighted in purple don't) with the quality of the passing efficiency defenses he faced in the 2007 season. I mean there is an explanation for everything.

• Ohio State (4th in FBS), USC (6th), and ASU (15th), each having a strong PED, held him in check, for example. They had pass defense efficiency ratings that were under 109 on the season, and Locker posted ratings of 79.25, 66.29, and 82.96 against them respectively.

• He had five games in which he bettered the PED he was facing. In two of those games, he matched up against non-BCS opponents, Boise State and Hawaii. In the other games, the three teams' overall pass defenses were rated 70th, 71st and 86th in the country.

• Purple row: He was injured and lost to the team in the Oregon State game, which explains his mediocre number of 52.46. Finishing up the game for Locker, Carl Bonnell posted an exceptional PE of 144.69.

• Purple row: The game against WSU marked Locker's return to action after that injury, and he had a poor day passing. Call it rust and anxiety.

• Purple row: The glaring anomaly was in the Stanford game (83.39 versus a 134.4).  He completed half of his passes in that game and threw one pick against no touchdown passes. He had 16 rushes for 97 yards, as part of 388 yards rushing the Huskies posted on the game. Perhaps, his passing wasn't needed in that game, in a game the Huskies won.

• He underachieved in the Syracuse game, posting a PE of 136.46 against a defensive rating of 147.04. However, that game marked Locker's baptism of fire and he did post a nice number.

What these numbers suggest (Tables 1 and 2) is that when Locker's offensive line and receivers are the equal or the better of the pass defense he is facing, he will be as good a passer as any other quarterback in the Pac-10 next season, if not better. The tutoring he's receiving from quarterbacks coach Doug Nussmeier and Sarkisian, a former quarterback, will only make Locker a better passer.

Now for graphs, tabular data and that damned cat.

In the chart below, a graphical image of the data in Table 2, Locker's passing efficiency is represented by the green line and the opposing team's pass defense efficiency by the orange line. Six points are below 100 and six are above. Of the six points below 100, three of them can be explained by the quality of the pass defense Locker faced, and two of the other three by his injury in the Oregon State game. His performance in the Stanford game remains a mystery, as does the health of Schrödinger's cat until you open the box ;-)

A graphical image of all 16 games that Locker has played in, as ordered chronologically. Locker's numbers are in blue. PED is in orange. Locker has finished either close to or higher than the PED in 9 of his games. USC and ASU had his number. Injuries account for his performance against OSU and WSU and in the second game against the Tree. The first game against the Tree remains a mystery. If LSU's pass defense efficiency is close to what it was in 2008 (116.62), Jake's passing statistics against the Tigers should be close to those points near the 120 line, e.g., BSU, UCLA, Hawaii, BYU, and Oklahoma below. Perhaps, even higher since LSU's overall pass defense ranked 73rd in the nation.

Table 1. Chip's dilemma. Locker's passing efficiency over his last 8 games. The cumulative ratings are shown on the right side of the table.

Game	C	A	Y	T	I	PE	C	A	Y	T	I	PE
Spring '08	13	17	159	1	1	161.69	13	17	159	1	1	161.69
Oregon	12	28	103	0	0	73.76	25	45	262	1	1	107.35
BYU	17	32	204	1	0	116.19	42	77	466	2	1	111.36
Oklahoma	16	24	154	0	0	120.57	58	101	620	2	1	113.54
Stanford	5	9	51	0	0	103.16	63	110	671	2	1	112.69
04/11/09	10	18	142	1	0	140.16	73	128	813	3	1	116.56
04/18/09	10	16	133	1	1	140.45	83	144	946	4	2	119.21
04/25/09	16	18	200	2	0	218.89	99	162	1146	6	2	130.29

Table 2. Schrödinger's cat. Jake Locker's throwing (middle column) versus pass defenses in the 2007 season. So, are his numbers really a conundrum? His numbers all make sense except for the Stanford game.

Team	Pass Def Eff	Pass Eff	PD
Ohio State	98.73	79.25	1
USC	102.0	66.29	15
ASU	109.3	82.96	61
Hawaii	111.29	111.34	37
Boise State	113.42	122.05	26
UO	114.2	144.48	86
UCLA	115.0	123.18	70
Arizona	115.5	159.41	71
Oregon State	116.0	52.46	72
California	121.4		38
WSU	133.4	86.05	93
Stanford	134.4	83.39	107
UW	141.5		104
Syracuse	147.04	136.46	102

Table 3. This guy stinks. Jake Locker's career passing efficiency. Black = games played in 2007. Purple = games played in 2008. The numbers in bold highlight his +100 games.

Team	C	A	Y	T	I	PE	C	A	Y	T	I	PE
Syracuse	14	19	142	0	0	136.46	14	19	142	0	0	136.46
BSU	13	25	193	1	1	122.05	27	44	335	1	1	128.27
tOSU	16	33	153	1	3	79.25	43	77	488	2	4	107.26
UCLA	17	36	216	4	2	123.18	60	113	704	6	6	112.33
USC	13	28	90	0	1	66.29	73	141	794	6	7	103.19
ASU	10	28	142	1	1	82.96	83	169	936	7	8	99.84
UO	12	31	257	4	1	144.48	95	200	1193	11	9	106.76
UA	17	30	336	2	2	159.41	112	230	1529	13	11	113.62
Tree	16	32	151	0	1	83.39	128	262	1680	13	12	109.93
OSU	6	14	16	0	0	52.46	134	276	1696	13	12	107.02
WSU	12	35	224	1	2	86.05	146	311	1920	14	14	104.06
Hawaii	9	17	142	0	1	111.34	155	328	2062	14	15	105
UO	12	28	103	0	0	73.76	167	356	2165	14	15	102.54
BYU	17	32	204	1	0	116.19	184	388	2369	15	15	103.74
OU	16	24	154	0	0	120.57	200	412	2523	15	15	104.72
Tree	5	9	51	0	0	103.16	205	421	2574	15	15	104.68