Game Predictor vs. the Spread (1 Viewer)

[Da Beers]

So I never really paid attention to the Game Predictor until just now. I was curious as to how the predicted final scores would compare with the Vegas lines. It surprised me that every single game (except 1) was within 1.5 points of the line (used wsex.com). The 1 game that has a significant discrepancy is the Green Bay - New Orleans game. Wsex.com has the Saints favored by 2.5, while Game Predictor has the Pack favored by .7. As I was going through all of the games, I started making the assumption that you were using the Vegas lines to set your game scores and then working top-down to come up with the passing/rushing breakdowns. But perhaps I made a false assumption?

Also, if you are not using the Vegas lines as input, do you keep records of how Game Predictor does against the spread?

 

[radballs]

Yes, Dodds has historically kept track of how the games do against the spread. Or at least he did going back two years ago and earlier. And it's done well enough that you would have won a bit of money and covered the juice.

I was in one of those Yahoo pools one year where you actually pick against the spread (this was three or four years ago). My system was to take the lines and then subtract Dodds' Game Predictor spreads from those lines. Then rank them based on the biggest point spread in my favor (in other words in this pool you not only picked against the spread but also had to set your confidence level of each game.) I won the pool that year with about 40 people in it. The projected points seem to be very close to the casual observer so they are conservative, but people underestimate how many dogs cover or outright win their games.

 

[Jene Bramel]

Dodds posts some thoughts here.

 

[Da Beers]

Dodds posts some thoughts here.

Thanks -- I actually did run a search (+Game +Predictor) but must have missed this thread.

 

[MikeReagan]

I emailed him about tracking how it does SU and ATS... he said:

"I build it so I can accurately make fantasy projections for the players. I really do not care how successful it is against the spread."

 

[David Dodds]

I don't look at the point spreads at all when I do this. They are not an input. But that said, it does not surprise me that my data comes up close to the point spread people. I am sure they run algorithms as well to come up with their initial lines.

The game predictor has traditionally gotten a lot better after about 4 games since it relies so heavily on good data to drive it. I no longer track it against the spread because I don't want to be responsible for people losing their houses, etc betting. I created it so that I could create accurate projections.

 

[Michael Brown]

I no longer track it against the spread because I don't want to be responsible for people losing their houses, etc betting.

Yeah, that's my job now.

 

[BassNBrew]

I don't look at the point spreads at all when I do this. They are not an input. But that said, it does not surprise me that my data comes up close to the point spread people. I am sure they run algorithms as well to come up with their initial lines. The game predictor has traditionally gotten a lot better after about 4 games since it relies so heavily on good data to drive it. I no longer track it against the spread because I don't want to be responsible for people losing their houses, etc betting. I created it so that I could create accurate projections.

Dang...you should run for Congress and shut down internet poker.P.S. You don't think people have money at risk who are using your projections and cheatsheets?

 

[QUEZILLA]

Used it last year for my office pool and won