A number of sites have collected information on Nobel prize winners over the years.

Nobel site: http://www.nobel.se
http:///www.suttontrust.com/text/nobel.doc
list of institutions at time of discovery:
http://www.nobelprize.org/nobel_prizes/lists/universities.html

You can download my summary analysis: nobel-analysis-v1.
There is also an interesting book discussing many of the prizes, “A Century of Nobel Prize Recipients: Chemistry, Physics, and Medicine” by Francis Leroy.

Physics analysis: from 1952 to 2011 (60 years)
1952-1981: S=  27    C=  10    E=  22    T=  15; S=  73%  C=  27%  E=  59%  T=  41%
1982-2011: S=  17    C=  18    E=  26    T=  8;   S=  49%  C=  51%  E=  74%  T=  23%
[S=simplifiers, C=constructors, E=expt, T=theory]
[Assignments depend on detailed consideration of each discovery]

– there is a rise in constructors (~doubles in 30 years), and more of them are experimental discoveries. We might say that science becomes more surprise  led. There has been a rise in the ability to make controlled and novel physical systems.

– would we say that liquid helium is ‘natural’ or ‘constructed’? Cooling helium gas to a few degrees above absolute zero is very unnatural, but we can say that it exists naturally (maybe somewhere else in the universe?). What might we say then about ultracold atoms: these have to be optically pumped in special rigs (definitely constructed) to get them a few nanoKelvin above absolute zero.

I tried to look at the balance between different fields within physics.
1952-1981:   Ast=  7%    Sol=  22%  Par=  39%  AMO=  22%  Img=  2%  QM=  7%
1982-2011:   Ast=  16%  Sol=  32%  Par=  22%  AMO=  16%  Img=  8%  QM=  5%
Here Ast=astrophysics and astronomy, Sol=solid state physics, Par=particle physics, AMO=atomic, molecular and optical physics, Img=imaging and detection physics, and QM=quantum mechanics. Since AMO includes phase contrast microscopy, Imaging includes electron optics, and Solid State includes CCD arrays, these all overlap and I allowed multiple counts in such cases.

For physics, the delay between making a breakthrough and being awarded the Nobel prize is increasing, though there is considerable variability:

The average wait from discovery to prize has more than doubled. Perhaps it takes time to work out what gives impact now, or there are more people queuing up to get them.

For the Nobel in chemistry, this gives in a similar fashion
Chemistry analysis: from 1952 to 2011 (60 years)
1952-1981: S=  23    C=  11    E=  25    T=  10; S=  68%  C=  32%  E=  74%  T=  29%
1982-2011: S=  22    C=  15    E=  30    T=  6;   S=  59%  C=  41%  E=  81%  T=  16%

– This shows it is becoming harder to show breakthroughs in theory in these areas.

The breakdown among sub-fields is
1952-1981:   Org=  31%  Bio=  31%  Ino=    3%  Mech=  15%  Meas=  8%  Phy=  13%
1982-2011:   Org=  37%  Bio=  37%  Ino=  12%  Mech=  10%  Meas=  5%  Phy=  0%
where Org=organic, Bio=biochemistry, Ino=inorganic, Mech=mechanisms (theory), Meas=measurement techniques, and Phy=physical chemistry.

– moved from more simplifiers, to now equal C,S (the same as physics).

– Organic and Biochemistry have started to dominate more, while Physical chemistry disappears and Measurement techniques plummet. Perhaps this is due to proliferation of techniques since few can dominate and most are more specialised.

– the delay in gaining recognition has lengthened less than in physics (but with even more variability).

– perceived utility in chemistry has been strong from the start of this period, perhaps due to the much earlier maturation of industrial chemistry.

Medicine analysis: from 1952 to 2011 (60 years)
1952-1981: S=  27    C=  9    E=  35    T=  0; S=  75%  C=  25%  E= 100%  T=  0%
1982-2011: S=  25    C=  9    E=  32    T=  1; S=  74%  C=  26%  E=   97%  T=  3%

1952-1981:   Mic=  18%  Hum=  34%  Gen=  21%  Cell=  16%  Meas=  8% Ani=  3%
1982-2011:   Mic=  28%  Hum=  19%  Gen=  25%  Cell=  25%  Meas=  3% Ani=  0%
with Mic=microbiology, Hum=human, Gen=genetics, Cell=cell biology, Meas=measurement techniques and Ani=animal/vet science (again with a number of prizes counting under more than one sub-field).

– there appear to be more multiple award winners than in the physics and chemistry Nobels. Are these fields which need more teamwork or with different pieces of the grander puzzle coming from different places to give a breakthrough?

– again medicine had its first phase of applications a long time pre-1950 (`surgery cures’), so has been primed for more recent applications.

– the ratio of Simplifiers:Constructors has stayed the same over time. New understandings are responsible for three quarters of all major advances.

– overwhelmingly experimental advances are rewarded: theory cannot provide the same explanatory powers yet.

– in sub-fields, Human biology success has dropped, while Cell Biology and Microbiology have risen. There are virtually no recognised breakthroughs in the fields of behaviour and psychology.

Where were Nobel advances done?

Total:  Univ=  64%  Institute=  19%  Company=  16%

– more constructors were housed in companies, more simplifiers in universities, with institutes in between.