[TotallyTilt]

Wondering if anyone knows the value of 2 +ev handicappers liking the same game. Let's say they both hit at 55%, does that mean there's a > than 55% chance of winning? What if one hits at 55% and the other 60%?

[Justin7]

It depends. The more they disagree (or fail to agree on a play), the more valuable it is.

[TotallyTilt]

I'm not sure I understand, could you elaborate a little bit?

[Justin7]

Consider 2 pairs of 55% cappers.

Pair 1: They agree 100% of the time. The obviously hit 55% on all agreed plays.
Pair 3: They agree 50%, and they oppose each other 50% of the time. When they disagree, they both hit 50%. When they agree, they both hit 60%.

[TotallyTilt]

Gotcha, thanks Justin.

[yak merchant]

From an anecdotal perspective, I know if I see too many good cappers on the same side of a game, I usually become suspicious. If everybody that is good, or has a model is on one side of a game and the line isn't moving, somebody knows something that isn't factored in and Vegas is taking a stand. If you see one or two good cappers on a pick, great. If you see 20 good cappers on a pick be wary.

[Justin7]

I think a more relevant question is: how can you identify cappers likely to hit 55% going forward? Or even 54%? There are not many of them that post.

[Wrecktangle]

Having worked this both ways, these methods (many ways to combine up selections) work much better with modelers (it seems) as the underlying algorithms are much more consistent. However if you have a lot of selectors (15+) this tends to even out.

Using the inverse (fading) does not work either as the faders either tend to either quit, or change methods.