Best book to learn probability - Poisson, binomial, regression, etc [duplicate]

What is the best book to learn probability - Poisson, binomial, regression, etc.

I am working as an odds adjuster at a bookmaker and need to advance my skills to an odds compiler level.

-

marked as duplicate by Gavin Simpson, Peter Flom♦, gung, COOLSerdash, JohnSep 16 '13 at 23:41

Perhaps few here know what the mathematical level required of an odds adjuster is, or that of an odds compiler: it would probably help you to get better answers if you elaborated a little. – Scortchi Jan 10 '13 at 20:05
I understand. Hence, I stated poisson, binomial, etc. As these are what is required in the field. Most basic probability books cover these distributions at a very basic level. I am looking for, not an advanced but a complete book which would cover say "Poisson Regression" and not just "poisson" or "regression", separately. Again, appreciate all the help as at other forums I havent been getting many answers. – O P Jan 10 '13 at 20:18
Saying 'Poisson' doesn't help all that much - Poisson processes are covered in graduate level monographs & in 'Statistics for Dummies'. – Scortchi Jan 10 '13 at 20:44
you are right. How about a probability book which will arm me with tools to deal with such a question "Wigan scores after 30 minutes. Calculate home, away and draw in percentage terms with poisson regression, assuming pre-match odds were 5.00 - 4.00 - 1.80 for Win/Draw/Loss respectively" – O P Jan 10 '13 at 20:58

Grinstead & Snell is very good (& free):

http://www.dartmouth.edu/~chance/teaching_aids/books_articles/probability_book/book.html

(Note it doesn't cover statistical topics e.g. regression, just probability.)

As you mention Poisson regression in a comment:

Dobson, An Introduction to Generalized Models

-

I think your question is worded a bit too "loosely". In any case: A number of excellent books are mentioned in this (aggregated) thread that deals mostly with Machine Learning books. From those mentioned probably The Elements of Statistical Learning is your best shot for a free book. All of them offer a general treatment for probability and regression.

Talking about non-free books I have found Statistical Models by Davison and Van Kampen's Stochastic Processes in Physics and Chemistry extremely helpful "to learn probability - poisson, binomial, regression, etc." but this is a quite open-ended approach...

-
I am sure there are books out there that deal purely with probability topics without delving into "machine learning" or "chemistry, physics". appreciate the help nevertheless. – O P Jan 10 '13 at 20:15
What's your level of Stats/Probability experience? Maybe Papoulis' Probability, Random Variables and Stochastic Processes is a better introduction. And don't be fooled by the titles and labels, "machine learning" books might as well contain all the information you need about regression. Finally, Van Kampen's book is a excellent read for Stochastic Processes, albeit coming from a physicist's perspective. – usεr11852 Jan 10 '13 at 20:28
In that case, I won't be fooled and will definitely check it out. Again, like i posted above, do u think those books will help me answer questions like "Wigan scores after 30 minutes. Calculate home, away and draw in percentage terms with poisson regression, assuming pre-match odds were 5.00 - 4.00 - 1.80 for Win/Draw/Loss respectively". – O P Jan 10 '13 at 20:59
" "Wigan scores after 30 mi ... " Could you please elaborate a bit more? which sport is this, for instance? What is the time for a full match. If this is football, for instance, I guess I would like to assume something about rate of scorings, too ... Then the pre-match odds could be related to the different teams different scorings intensitys, and your problem woulod be an updating using bayes. – kjetil b halvorsen Jul 2 '13 at 14:14

You could find some of the following of interest:

a and b

The following might be too advanced for you, but maybe still useful:

c and d

-

We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.

I added the post notice because this answer is unintelligible without following the links. Please at least indicate what you are recommending. Much better would be to include the reasons for your recommendations. – whuber Aug 31 '13 at 18:39