Alex and Brendan, 

  This is a very good paper on an interesting topic where there is still a lot of disagreement.
  The evidence from different sports (like bowling) is valuable because it highlights
  how differences in the game play probably influence whether "hot hand" situations
  actually come up or not.  And I think you are exactly right to distinguish games (like
  slot machines in casinos) where the player has no direct control over the outcome and
  "hot hands" demonstrably don't exist from sports like bowling (or most other sports) where
  the skill and actions of the player do influence the outcome directly.
  There are also lots of interesting questions about the proper statistical methods to apply 
  to these questions.  As you say, the basketball results seem to come down to this: even 
  if shots within actual basketball games behave essentially as though they were close to
  being independent events from a statistical point of view, there is still a real 
  psychological phenomenon underlying the ways that players, coaches and fans perceive what 
  happens in games.  One can argue whether this is a "perception" or a "misperception" in 
  the light of the statistics, but it's real as a pattern in human psychology.  We are 
  "wired" to perceive patterns in the events around us and to invent explanations for 
  our experiences based on those perceived patterns.  And in the much more consistent 
  and controlled enviroment of bowling, the evidence seems to point in the other direction(!)  

  The only really substantive criticism I have is that I was hoping that all of the projects 
  would involve doing some actual analysis of data.  You essentially decided to report the 
  results from the Gilovich, Vallone, and Tversky article and others, plus the bowling data 
  from the Dorsey and Palmeteer paper and others.  You presented a really good range of 
  different viewpoints and results.  But it would have been possible to do some more original
  work with things like what GVT discuss on page 311-312 of their article.  And that 
  is what the slate.com article by Jordan Ellenberg and the the Miller and Sanjurjo article I 
  pointed out in the final project assignment sheet discusses as well.  

  Specific comments:

  Page 1: "31 years later, in this paper we will discuss that although historically debated 
  not to be found in sports, the ‘hot hand’ phenomenon is exhibited in Bowling and not in 
  Basketball."  This is an awkward sentence because you are trying to cram so much into it:
  the fact that these questions have raised debates over a long time period (31 years 
  since the GVT paper), the fact that different sports have been considered (bowling vs.
  basketball), and finally that you are going to concentrate mostly on the new article 
  about bowling. When you find yourself in this situation, try breaking your ideas up into
  several sentences and introduce the points one at a time(!) Also, since you have mentioned 
  the GVT paper earlier in the paragraph, it would be better to give an explicit reference 
  to the bowling sources you are drawing on at this point as well.   

  Page 1:  "This is not a far conceived preconception."  Do you mean "far-fetched" or
  something like that?

  Page 2:  "Therefore there must be no ‘hot hand,’" and later "Gilovich originally disproved 
  the notion of a hot hand in basketball in the 1980’s."  Still later on p. 6 "to try and 
  prove the ‘hot hand’ phenomenon." These statements are all too strong.  
  In science, the results of a single experiment are *never* taken as definitive proof of 
  something!  Say something like "the data did (or did not) provide evidence to support 
  the 'hot hand' hypothesis."  

  Page 2:  The formula for serial correlation looks *really messy* (and almost unreadable,
  at least within the Google document viewer).   I don't know if you were using MS Word to 
  prepare this, but for the future, you may want to look into the *Equation Editor*.  That 
  lets you set up formulas like this and insert special characters like a real Greek 
  lower-case "mu" for the mean.  The results look much better(!) 

  Pages 2 - 3:  I think it might have been better to put the intuitive explanation of
  serial correlation before the technical discussion of the formula.  The example 
  of weather patterns (rain) is a really good way to introduce the idea, and having
  that to refer to would make it easier to write about the mathematical formula for 
  doing the computation of serial the correlation coefficient.

  Page 5:  "This 2-3% does not seem like significant evidence to reject the null hypothesis."
  It all depends on how many shots are included (the n).  See the class lecture notes on
  testing for proportions, or Chapter 10 in "Naked Statistics" for the standard error
  formula used to compare proportions.  A 2-3% difference could easily be statistically
  significant if n is large enough.  Whether or not that difference is practically significant
  is a different question, of course.  

  Page 5:  "There seems to be too many external variables like psychological factors even
  at the free-throw line and controlled experiment level. If a player hits two shots in a 
  row, he will psychologically think that he will make the next one. This cannot be recorded 
  in statistics.

  Page 6:  "Dorsey-Palmateer argues that"  The discussion you are quoting is their criticism
  of why they think the GVT paper "got it wrong" in basketball.  You should probably say that
  to set up the next bit, which explains how bowling is different and possibly a better
  test of whether the "hot hand" exists or not.  

  Final Project: 

  Annotated Bibliography -- 10/10
   
  Presentation -- 33/35

  Paper -- 49/55

  Total -- 92/100  (letter: A-)