Saturday, May 18, 2013

The Yankees at the One Quarter Mark: Why Are They Exceeding Expectations? It's Not What You Think (May 18, 2013)


I’ve read many articles and heard many commentators discussing the surprising Yankees, offering theories on why the Yankees are doing so well in the first quarter of the season.  The most popular notion is that the veterans the Yankees picked up off the “scrap heap,” the so-called “2006 All Stars,” are turning the clock back and slugging the team into first.

It is a compelling story.  You will recall in the off-season the Yankees lost Nick Swisher, Russell Martin, Eric Chavez, Raul Ibanez and Andruw Jones, a fairly devastating blow to the offense.  And then, in short order, the Yankees lost Alex Rodriguez, Derek Jeter, Mark Teixeira and Curtis Granderson for lengthy periods before the season started.  And then, once the season started, Francisco Cervelli bit the dust after a hot start.  This essentially left Robinson Cano as the only holdover starter year over year.  (Brett Gardner was injured for virtually all of 2012.)

To the rescue came the Over the Hill Gang (I will refer to them as "OTHG" henceforth):  Ichiru Suzuki (re-signed to replace Swisher), Kevin Youkilis (A Rod), Vernon Wells (Granderson), Travis Hafner (Ibanez) and Lyle Overbay (Teixiera).  In truth, only Ichiro and Wells made the All-Star team in 2006, but all of them did indeed have fine seasons – a superb collective weighted average OPS of .885.  And all of them have faded mightily since, achieving a collective weighted-average OPS of only .726 in 2012, quite a dismal level for players who play power positions.

And yes, the Yankees are off to an amazing start.  And yes, the OTHG is indeed outperforming their 2012 selves and doing more than plugging a few gaping holes.

Great story!  Except for one thing:  it is simply is not true.  The Yankees’ surprising start has little to do with the OTHG.

So what is the real reason?

First. Let’s establish the terms of the discussion.  The Yankees, after 41 games, are in first place in the tough AL East with a 25-16 record, a .610 won-loss percentage that projects to 99 wins.  This is obviously an excellent record, especially with the injuries – but how can we analytically compare it to what was “expected” of them?

I’ll use my Yankee predictive model (see: www.borntorunthenumbers.com/2013/03/yankees-2013-prediction-is-it-1965-i_31.html) as a basis for comparison.  I had them winning 90 games for the full year, based on a projected Team OPS of .757 and Team ERA of 3.86 (figures I derived from trend data, expected at bats and inning pitched, and a regression equation).

Thus, I might have expected them to have won 23 games by now, 41 games into the season.  But even though I knew about most of the injuries at the time of the projection, I did not project the 41 game number; I assumed all the injured players would come back at various points.  So I adjusted my model using the actual plate appearances and innings pitched each player has after 41 games, but still using their projected (not actual) performance; based on those figures, the model says the Yanks would have had 21 wins, 2 less than the model.

So, the analytical question is:  why do the Yankees have 25 wins at this point instead of 21? 

Well, it is not because they are wildly exceeding their expected performance statistically.  While the pitching has been a bit better than expected (3.67 ERA versus a projected 3.86), the hitting has been right on (an OPS of .729 versus a projected .725).  So you can chalk up 1 of the 4 wins to “better pitching,” specifically better relief pitching, with an ERA of 3.31 versus 3.91. 

As for the hitters?  Well, the OTHG has indeed outperformed expectations, but they are hardly “turning back the clock to 2006."  Collectively, their weighted OPS is .788, well above the .723 from 2012, but a far cry from 2006’s .885…in fact, it would be perfectly accurate to say that the OTHG has turned the clock only “halfway back” to 2006.


           On Base plus Slugging (OPS)

2006
2012
2013
2013

Actual
Actual
Projection
Actual
 Youkilis
0.810
0.745
0.750
0.769
 Suzuki
0.786
0.696
0.680
0.608
 Hafner
1.097
0.784
0.775
0.893
 Overbay
0.880
0.727
0.725
0.789
 Wells
0.899
0.682
0.680
0.881
Wtg. Average
0.885
0.723
0.726
0.788


In fact, if the OTHG had actually performed as planned through the first 41 games – that is, at roughly 2012 levels – the Yanks would have won exactly one fewer game.  That excess production has been wonderful, but worth exactly one win to the team.

And this excess contribution has been exactly offset by the rest of the team.  Most of the rest – Gardner, Nix, Nunez, Francisco and Boesch – have underperformed, and that has completely offset the extra win the OTHG has achieved….hence overall the team is hitting exactly as planned.

So how to account for the other three extra wins? 

The real reason the Yankees have overachieved?  Because they are this year’s Baltimore Orioles.  They have been winning close games, those decided by two or less runs, by an astonishing margin, nearly three-quarters of them (.727 percentage) while essentially only splitting games decided by three or more runs.


Won
Lost
1 or 2 run games
16
6
3+ run games
9
10

And that has more to do with superb late relief pitching.  The Yankees have had 20 save situations this year and have saved 18 of them or 90%.  And that of course is all about Mariano Rivera, who is 16 for 16, at 100%.  The average major league team has 15 save situations and saved 11 of them, or 69%.

If the Yankees had saved only 69% of their save opportunities, they would have won four fewer games.  So while their batting is slightly below expectation, and pitching is slightly above, and those have offset, their very clutch pitching is sensational, and this is causing them to win a much higher proportion of close games that an average team.  And do better than their overall stats would imply.   Obviously one of those four wins has already been attributed to the relief staff because of their better-than-anticipated ERA.  Now, we see the other three extra wins can be attributed to pitching very well when it really counted, in those close end-of-game situations.

So to break it all down:


Games Won
Expected Wins
21
OTHG
1
Other batters
-1
Starting pitchers
0
Relief pitchers
1
Save opportunities
3
Actual wins
25


The Yankees are not overachieving because of the Over the Hill Gang.  No, it is all about that 43-year old guy who has yet to see the other side of that same hill…yes, Mariano Rivera. 

Thursday, May 9, 2013

The Econometer Turns Blue, Good for the Democrats, Amidst a Crazy World (May 9, 2013)

It was a momentous month in the world, a month of disasters, both real and threatened.  President Obama was able to demonstrate strength as “First Healer” in the aftermath of the Boston Marathon bombing, as well as patience and restraint with respect to North Korea nuclear threats and Syrian use of chemical weapons.  It was also a month for good economic data, all of which might have translated into significantly higher Obama approval numbers.  They were indeed very modestly higher, but the failure of the Senate to pass any gun control measures hurt him and his Democratic congressional colleagues as well, and thus the outcomes were a bit muted.  (See my April Month in Review for my update on all the major political events of the month at:  http://www.borntorunthenumbers.com/2013/05/april-month-in-review-startling.html)

The “Econometer” captures what was generally a good set of economic data for the past month.  The unemployment report not only was fairly robust in the private sector in April, but there were substantial upward job growth revisions to the February and March reports as well.  The unemployment rate thus fell to 7.5%.  Consumer confidence rose, and the GDP for the first quarter came in at 2.5%, well above the anemic 0.5% of the fourth quarter.  The Dow has soared to record highs, and even the price of gas has dropped to below Election Day levels.  In short, each measure movied in a positive direction this month.  None of this points to a dramatic recovery, but it also refutes the notion that we were having a third “spring slump” in succession.

Remember, the Econometer measures how the economy has moved since Election Day on the theory that positive news should benefit the Democrats come 2014.

Econometer
7-Nov

7-Feb
8-Mar
8-Apr
8-May
Econometer
0.0

-11.9
0.6
-1.6
13.5
  Unemployment Rate
7.9

7.9
7.7
7.6
7.5
  Consumer Confidence
73.1

57.3
68.0
59.7
68.1
  Price of Gas
3.71

3.45
3.79
3.75
3.62
  Dow Jones
  13,330

  13,728
  14,054
  14,520
  14,747
  GDP
3.1

0.4
0.4
0.4
2.5

Obama’s approval rating nudged over 50% again, and is now a net +3.0 points, a bit better than Election Day.  This represents his highest level since February, although in general Obama’s approval rating has settled into the 50% range after the election.

Obama Approval Rating
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Approve
49.6

51.7
50.1
49.0
50.8
  Disapprove
47.4

44.2
45.0
46.7
47.8
  Net
2.2

7.5
5.0
2.3
3.0

The Democrats’ margin over the Republicans has narrowed to +2.4 points in the “generic ballot” for Congress.  You can see that the real change was that the Democrats lost three points to “don’t know”; the Republicans remained in the 38% range.  I think this could reflect some frustration with the gun law, which was widely supported by 60%+ of the electorate.  Perhaps they expected widespread Republican opposition but were surprised that several Democrats broke ranks.  It is interesting to note that the “don’t know” portion of the electorate is now over 20%, not exactly a vote of confidence for either party.

Generic Congressional Ballot
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Democrat
46.3

44.5
42.8
43.8
40.7
  Republican
46.0

39.7
38.5
38.0
38.3
  Net
0.3

4.8
4.3
5.8
2.4


There have been no new looks at 2016 so I will repeat last month’s data set and await new polls this coming month.  The headline:  it’s Hillary’s to lose at this point, and the Republican field is deep but split, with Marco Rubio holding a modest lead.

Presidential Preference - Dem.
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Clinton


58
58
64
64
  Biden


19
19
18
18
  Warren


8
8
5
5
  Cuomo


3
3
3
3







Presidential Preference - Repub.
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Rubio


22
22
20
20
  Paul, Rand


10
10
16
16
  Christie


13
13
15
15
  Ryan


15
15
15
15
  Bush, Jeb


13
13
11
11
  Huckabee


11
11
n/a
n/a







Head-to-Head
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Clinton


49
49
49
49
  Rubio


41
41
42
42







Net Favorability - Dem. (among Dems)
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Clinton


69
69
79
79
  Biden


67
67
62
62
  Warren


40
40
25
25
  Cuomo


20
20
22
22







Net Favorability - Repub. (Among Repubs)
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Ryan


69
69
62
62
  Rubio


47
47
43
43
  Paul, Rand


48
48
39
39
  Bush, Jeb


47
47
33
33
  Christie


15
15
9
9
  Huckabee


59
59
n/a
n/a







Senate Projections
7-Nov

7-Feb
8-Mar
8-Apr
8-May
  Total Forecast Democrat


50
50
50
50
    Democrat Not Up


34
34
34
34
    Demomcrat Solid


11
11
11
11
    Democrat Leaning


5
5
5
5







  Total Forecast Republican


50
50
50
50
    Republican Not Up


31
31
31
31
    Pepublican Solid


14
14
14
14
    Republican Leaning


5
5
5
5







* Italics mean no new info from prior month