Quote Originally Posted by jgilmartin View Post
It depends what you mean by 'consistently' I suppose. No one can do it every time, but people can do it two thirds of the time; do you consider that 'consistently'?
Yes, I would think that someone who can predict the line movement correctly 66% of the time has a big advantage. I just don't know anyone who can do that; maybe you do.



Quote Originally Posted by MadTiger View Post
Great post, but don't underestimate the need to at least have somewhat of a clue as to what the line will do in the remaining time to the start of the event. You have to at least have a reasonable justification for why you are pulling the trigger on the bet now instead of later.
I respectfully disagree. For example, I've developed a model for MLB that spits out probabilities; if the predictions of the model are significantly different from the current line, I bet it. I really don't have any clue whether the line I'm betting on is going to stay where it is, move toward me, or against me. But as I said, the prediction from my model has to be SIGNIFICANTLY different from the line, so small moves one way or the other won't drastically affect my estimated EV. Admittedly, I haven't (yet) carefully looked at the subsequent line moves to see if there's anything systematic that I can find relative to what my model is saying -- that's on my to-do list.

Having said all that, I agree that IF you happen to have some basis for knowing which way the line is likely to move on a given game, then yes that's very valuable information. I just think it's probably a very tough thing to do consistently.