[npc]

[ QUOTE ]
I have read some good stuff about sports betting and since USA is supposed to be the biggest sports betting market in world, im surprised there if there isin't a good book?

[/ QUOTE ]

There are some good books, Wong's "Sharp Sports Betting" being the best I've read, but there is a lot that could be written on the subject that isn't, probably mostly for the reasons that King Yao mentions.

[ QUOTE ]

And are these things included in books(as if they are im very intrsted to hear what's the even bigger key for winning sports betting):

1) Kelly-criteria

[/ QUOTE ]

This is mentioned everywhere. Wong's book, for instance, discusses it. So does J.R. Miller in his book, although he derides it. (This is one of the things that I greatly disliked about his book.) The blackjack literature is replete with discussions about Kelly betting. Of course, the trick is to accurately determine your edge.

[ QUOTE ]

2) Over odds(+ev)


[/ QUOTE ]

I don't know what this means.

[ QUOTE ]

3) Power-ratings and how to make them


[/ QUOTE ]

The best information I've seen on this in a book is the aforementioned J.R. Miller's "How Professionals Beat the Pro Football Point Spread". His methods have some flaws, but it's a start. Also, some information on this can be found in the academic literature. For example, Dr. Hal Stern of Iowa State has written a couple of papers that contain information that someone working in this area might find useful.

[ QUOTE ]

4) Calculating the odds


[/ QUOTE ]

I'd say this is mostly subsumed in other areas.

[ QUOTE ]

5) Making computer programs to calculate the odds


[/ QUOTE ]

Once you've got a model for the games themselves, this should be a straightforward application of statistics. I'd think nearly any book on computer statistics would serve here.

[ QUOTE ]

6) Poisson
7) Arbitrases


[/ QUOTE ]

Wong covers these two topics quite well, IMHO.

[ QUOTE ]

8) How much does every single thing effect to your %-estimations.


[/ QUOTE ]

Some can be said about this, but the effects on some things (such as weather) may be time variant. However, it's pretty straightforward to write simple programs to examine the data to look for variable effects such as home field, weather, etc.. Part of the trick is knowing what variables to look for, although, honestly, I think most sports betting pundits make a bigger deal of these factors than is warranted.

Most of these things are at least discussed in existing literature. I'm still looking for a better published power ranking algorithm than the one Miller describes. (If someone knows of one, let me know. I've got one I use, but it could likely use some work. Heck, I don't even see any discussions of these issues on the 2+2 Sports Betting forum, although maybe I just don't know what I'm looking for.) Most of the math stuff is out there or straightforward for a modestly skilled programmer to adapt. Wong's book provides a lot of good information here.