[BostonCard]

[quote author=Boston Card link=topic=12578.msg125628#msg125628 date=1440623749]
First, the results correlate reasonably well with others, based on "gut feelings" and subjective analyses by both us and professionals;

I am sure you realize that there is a certain irony that proceeding through a substantial statistical analysis is justified because it agrees with gut feelings.


But, if you find it useful then it is useful to you.
It's just that, as you know, this is one of my pet peeves - a professional hazard.
[/quote]

Sorry, I did not mean to suggest that the statistical analysis was useful because it agreed with my gut feelings and that it would not have had it disagreed with what my gut feeling.  I am very attuned to the possibility of epistemic closure (which is a professional hazard for me), and have joined you in the past of being critical of relying too much on the "eyeball test" and being overly dismissive of data when it disagreed with what the "eyeball test" was saying.  Were there a discrepancy between what ESPN's stats said and what I thought, it would have been a good reason to to re-examine my assumptions about the team, not proof that the statistical analysis was useless.  Even with the agreement between a semi-statistical method, gut feel, and talking heads, I am concerned about the possibility of "herding" (that is, that the various different methods are overly correlated, say, because my "gut feeling" is unduly influenced by preseason votes or by the math).

As an example, having looked at the results of the data, I have come to the conclusion that OutsiderFan's projections of Cal, while on the optimistic side, are not wild and crazy.  While my gut feeling was that I was not convinced that Cal would get their sixth win, let alone their seventh or eighth, looking at things more closely, I have come around to believing that OutsiderFan is more likely to be right than Dodd, though I don't think it would be fair to label Dodd's predictions as "crazy".

BC

[BostonCard]

[quote author=Boston Card link=topic=12578.msg125628#msg125628 date=1440623749]
First, the results correlate reasonably well with others, based on "gut feelings" and subjective analyses by both us and professionals;

I am sure you realize that there is a certain irony that proceeding through a substantial statistical analysis is justified because it agrees with gut feelings.


But, if you find it useful then it is useful to you.
It's just that, as you know, this is one of my pet peeves - a professional hazard.
[/quote]

Sorry, I did not mean to suggest that the statistical analysis was useful because it agreed with my gut feelings and that it would not have had it disagreed with what my gut feeling.  I am very attuned to the possibility of epistemic closure (which is a professional hazard for me), and have joined you in the past of being critical of relying too much on the "eyeball test" and being overly dismissive of data when it disagreed with what the "eyeball test" was saying.  Were there a discrepancy between what ESPN's stats said and what I thought, it would have been a good reason to to re-examine my assumptions about the team, not proof that the statistical analysis was useless.  Even with the agreement between a semi-statistical method, gut feel, and talking heads, I am concerned about the possibility of "herding" (that is, that the various different methods are overly correlated, say, because my "gut feeling" is unduly influenced by preseason votes or by the math).

As an example, having looked at the results of the data, I have come to the conclusion that OutsiderFan's projections of Cal, while on the optimistic side, are not wild and crazy.  While my gut feeling was that I was not convinced that Cal would get their sixth win, let alone their seventh or eighth, looking at things more closely, I have come around to believing that OutsiderFan is more likely to be right than Dodd, though I don't think it would be fair to label Dodd's predictions as "crazy".

BC