[Monte]

lol the never ending topic...
it is so easy to observe. If you take Pinny as example, there is always a time early in the day where the lines move. Then the limits are raised/removed, and the lines usually stay up during day without moving much. Until it gets closer to game time, where they might again move.
So is the line where they take...50k on a mlb ML with the same odds for 5 hours sharp? You bet it fukking is.

[chunk]

if you disagree, you're going against every study done on prediction markets ever

I've never seen one of these studies that you refer to, but fair enough. Frankly, I don't think it's worth much time to worry about beating closers. The bottom line tells the story.

[Pokerjoe]

I've seen Pinnacle move the closing line after the market closed, not once but regularly. And every time they moved it back from shaded to neutral. This happened in NCAAB totals. For those who have studied the Pinny lean (which is not consistent across markets), this could be an indicator that they wiped off their fingerprints.

So it may be more reliable to take the closing line five or ten minutes before the market closes.

If not a little earlier.

Consider the market as a whole.
Team A is widely considered 7 pts better than B. It opens -7.
But it drops and drops, AGAINST the bulk of bettors, to -5, a half hour pregame.
From there it rises back up to -5.5.
I'd say, considering that the bulk of square action comes in the hour before a game starts, that the -5 line might be more accurate than the -5.5, the latter half point backtrack only reflecting the weight of square (obvious) action, which is NOT ignored by almost any book.

Certainly the closer is sharper than the opener. But it is not certain that there isn't an internal line sharper than the close.

As to comments about "all prediction markets," sports betting isn't like the other markets, in that in our market we have the dumbest money coming in at the latest time.

[God1]

but if it is say just Pinn or CRIS that moves right at the end, and that move is what determined the close for a set of data, there is a good chance the pre-move number is more accurate.

I understand your logic but why would pinnacle off of a more accurate number if it was indeed more accurate? Is it possible that pinnacle tries to balance the money more than most of us believe?

[Wrecktangle]

I've toyed with the idea of collecting the opener, "high", "low", as well as the close (like a stock market index) to test my modeling against. But when it comes to backtesting, closers are (almost) always the best (toughest) to test against.

As to Pinny "wiping off their finger prints," I've wondered about that too, and frankly wouldn't put it past them as so many watch their line.

[evo34]

Of course it does, you don't have to be Sharpy Sharp to know that.

Read the whole post. I was refuting Wrecktangle's erroneous claim that, "In College Football, we've have years where the opening line is better on average, and others where the closing line has been. No clear pattern over the last 10 or so years."

[Wrecktangle]

Read the whole post. I was refuting Wrecktangle's erroneous claim that, "In College Football, we've have years where the opening line is better on average, and others where the closing line has been. No clear pattern over the last 10 or so years."

Actually is not my claim, several of the ratings folks we have in college football have noticed this fact. I use closing lines for my modeling tests as it is widely available and extends back to 1976 (for sides) at least. But frankly, this is another of the tempest in a teapot arguments you guys in the tank get into. If your modeling can beat either consistently, you will do very well.