I hope for you that Wickmayer wins, but if you watched those girls play you'd know that this match is total coin-flip, odds should be 1.85:1.85 or something. Lol, ok i'll stop annoying you with my stupid comments.
I don't mind it at all! The whole reason I made a thread instead of just a word document is to hopefully have conversation, and learn about what I'm doing right/wrong and share it.
First off a coin flip is 50/50. A bookmaker would not give you 50/50 odds on the flip. Bookmakers charge you an amount for taking your bet. This commission for running the sportsbook is called the vigorish, vig, or juice. Instead of the coinflip being 2.00/2.00 odds the sportsbook will charge 10% vig and the odds will become 1.9/1.9. We want to bet at sportsbooks with the best odds to get the highest payout possible on our bet. If the same odds are being offered on an event across all sportsbooks (which is not the case usually) then the best odds will be the place with the least vig. Since there is not an even amount of money being bet at all books, and books adjust their lines independently (some being slower or quicker to adjust for changes in odds), there's often a huge difference in the odds between events.
Anyway, considering all this above... if Wick-Hercog was 50-50 the true odds would be 2.00/2.00. If a sportsbook had a 5% vig then it'd be 1.95/1.95. Looking at the available odds prematch for Wick-Hercog:
http://www.oddsportal.com/tennis/usa/wta-indian-wells/hercog-polona-wickmayer-yanina-trfpSjn3/At Pinnacle Wick opened around @1.61 and closed @1.55; Hercog opened @2.45 and closed @2.61. If the true odds of either winning was 50/50, then Hercog would be the better play hands down. Bet Hercog every time if this is the case, and in the long run over many games this would be profitable.
The whole damn tricky thing is determining what the true odds of an event occurring are. Wick-Hercog can't be 50/50. Tennis can be a sport where it's one person versus another, so it should be easier to examine than team sports. If two people are equal in every way then they will be 50/50 to win. The problem is that no one's equal, and their abilities change over time, and through different conditions. People have flaws, advantages and disadvantages. I suppose you would have to think about every possible variable in a match, and then test to see if those variables really matter in the outcome.
I know that people create models to predict the outcome of a match, but I don't really know what goes into making one. It would be cool to learn how to make one, and check if it is effective or not. If you made one you would need to feed it data all of the time that you get from somewhere. It would be a huge hassle to manually enter it all so you would need some way of automatically getting it, or hiring someone to get it. At some point the model is going to be like the mouse--if you give a mouse a cookie, it's going to ask for some milk as well. Then he'll ask for a straw. He'll finish and ask for a napkin, then a mirror to check if he has a milk mustache... the last thing you need is a model with a milk mustache. Irregardless of potential challenges it's something that would be awesome to look into.