With last night's 3-0 shutout in Game #32, the Jays have hit the 20% mark of an already up-and-down 2014 season. The dominant storyline of the season thus far has been the bullpen, and they've certainly deserved a lot of the scorn heaped upon them for blowups at inopportune times. At the same time, there's been a number of leads squandered or largely squandered before they got the ball. To dig a little deeper, I thought it would be useful to look at the season on an inning by inning basis. Have the Jays done a good job or getting and building leads? If so, is it driven offensively or defensively? Are they consistent?
For every inning, I pulled three data points: runs for, runs against, and Win Expectancy from Fangraphs. The first two are outcomes with no game context, which tells us about general productivity; the latter is purely contextual so we can get an idea for timing. This allowed me to calculate by inning the Jays' runs scored per 9 innings (RS/9), their runs against per 9 innings (RA/9), their overall run differential, their average Win Expectancy (WE) and their average Win Probability Added (WPA) for the inning.
Before diving into the data, one important point. This is merely looking backwards at what has happened. Though there's some projecting/extrapolating of wins below, its merely illustrative and not intended to be predictive.
Let's start by breaking down these numbers. On average, in the first inning the Jays have scored runs at a well-above rate of almost 6 runs per 9 innings while allowing runs at just 3.4 runs per 9. The result is an excellent differential of 2.5 runs per game. A full season of the First Inning Blue Jays would mean scoring almost 950 runs in a full season while only giving up about 550.
At the end of the first inning, the Jays have on average had a 52.5% Win Expectancy, which represents a 2.5% average increase over the beginning 50%. The highwater mark of first innings occurred in the second game of the season at Tampa, when the Jays jumped out to a 3-0 lead and would expect to win about 80% of the time (they also jumped out to a 3-run lead at the Rogers Centre on April 26th, but the higher run environment means a 3-run lead is slightly less safe).
Simply put, the Jays are dominating the first inning. Based on their runs scored and against extrapolated across 9 innings, over the entire season they'd be expected to win 119 games (Pythagorean expectation). Likewise, extrapolating the 2.5% increase in Win Expectancy to all inning would suggest a 117 win season. Basically, the 2014 First Inning Blue Jays are the '27 Yankees.
The second inning Blue Jays are slightly less spectacular - they've scored a little less and allowed a fair bit more runs - but still pretty solid. On average, they add 1.2% to their chance of winning, with their best 2nd inning adding 50% and the worst subtracting 35%.
Nonetheless, extrapolating things out, by either system they'd still win almost 100 games if they could maintain the level of performance achieved in the second inning. Across the first two innings, the Jays have been a juggernaut. So why are they sitting at 15-17? Well...
I believe the technical term for the Third Inning Blue Jays is "unmitigated catastrophe". The pitching is again sequentially worse and now below average, but they're not the real culprits. The bats have gone completely MIA, scoring just two runs. They didn't even score in the 3rd inning until the 25th game of the season. It's particularly baffling given for the most part, it should be the top and middle of the order up to bat in the third inning.
The average third inning has wiped out the gains of the first two innings, with an average 49.7% win expectancy, coughing up 4% per 3rd inning. Now, Pythagorean expectation breaks down at the margins, so the 3 expected wins can't be taken seriously. But even extrapolating by WPA, the Third Inning Blue Jays make the 2003 Tigers look like worldbeaters. 24 wins actually puts them in the fine company of the 1899 Cleveland Spiders, whose 20-134 record sets the standard for baseball futility.
Forget the bullpen in the late innings: this is where things have gone off the rails in 2014.
The 4th inning, contrary to its predecessor, has been very kind to the Jays. It's been their best inning, so good that if they accrued win probability for the rest at the game at the same rate as the 4th, they'd win 105% of the time! The 4th inning should generally see the lower part of the order, so the pitching improvement is somewhat attributable to this; the bats less so given that the bottom of the order frequently resembled a black hole for most of April.
The 5th Inning Jays are eccentrics: they haven't scored much and they've given up a lot of runs, but when the dust has settled and the World's Fastest Grounds Crew comes rushing out, the Jays have on average positioned themselves to have a better chance of winning. This is unsustainable, though their run differential will regress in the opposite direction. And as will shortly be seen, luck is a two-way street.
6th and 7th Inning
These two innings are almost perfect mirror opposites - what has been gained in the 6th has been in turn disgorged during the 7th. In fact, from the 3rd to the 7th the Jays have been pretty good at alternating been good and bad. The end of the 6th inning is generally the highwater mark for the Jays: if they could merely maintain their 57% win expectancy for the last three innings, they'd be 18-14, on pace for 92 wins.
The 8th innings has easily been the most offensive, both literally and figuratively. The Jays have put plenty of runs on the board; unfortunately, their opponents have done so even more. The run different is poor but not dismal, but the sequencing has absolutely killed the Jays. The 8th Inning Jays have scored multiple runs 7 times, gaining 108% in WPA (average of 15% per multi-run inning). They've given up multiple runs 6 times, losing 209% in WPA (average 35%).
It's a similar picture for the 9th Inning Jays. They've actually outscored their counterparts, while bleeding WPA. Across the last 2 innings, the Jays have outscored opponents 35-34, while falling from a win expectation of 54% entering the 8th to a below .500 record when the dust settles. There's really nothing else to be said, other than $%#%&#&*! Farrell $@(&^#*.
So what's the takeaway here? We knew the bullpen has been bad. But I think something that has been overlooked is the Jays' inability to land the knockout blow in the middle of games after knocking pitchers around early, frequently allowing them to settle in and go 6 or 7 innings and give their bats a chance. In the first two innings, the Jays have averaged 5.5 runs. From innings 3-5, just 3.5 runs. And yet typically, pitchers tend to allow more runs during this frame as batters see them more times and they tire. This helps to explain why three times they've had a 3-run or greater lead after 2 innings, and yet lost. I have confidence the bullpen will eventually come around, but I'll be paying a lot of attention to the middle innings.