Geo's Predictions

[lindend]

Date	Prediction	Outcome	Verdict
9/8	Expectation for Cole is UGA

 

[lindend]

I've thought about the rule preventing last second flips/changes in predictions in all the prediction threads. Rather that counting the original prediction only, I think if there are conflicting predictions and the last prediction happens at least 48 hours prior to the committment, we'll rule it a Push if the last prediction is correct.

 

[lindend]

Accuracy change after adjusting to the new rule on conflicting predictions being a push. (10/14 compared to 10/16)

Hit Rate: 71% (10/14) [Sample size too small, don't draw conclusions from this percentage]


2023 Predictions

Date	Prediction	Outcome	Verdict
9/8	Expectation for Cole is UGA	Picked UGA	Push
6/29	Schmo isn't coming to Miami	Picked Pitt	Push
6/29	Would be shocked if Cole isn't a Cane	Picked UGA	Push
6/27	Cole on commit watch	Picked UGA	Push
6/27	Geo see Schmo committing	Picked Pitt	Push

 

[Da Jumbo Mutombo]

I've thought about the rule preventing last second flips/changes in predictions in all the prediction threads. Rather that counting the original prediction only, I think if there are conflicting predictions and the last prediction happens at least 48 hours prior to the committment, we'll rule it a Push if the last prediction is correct.

Kind of lets them off the hook for a previous incorrect prediction tho?

Why not just count both predictions? 1 hit, 1 miss.

 

[calinative umstudent]

Kind of lets them off the hook for a previous incorrect prediction tho?

Why not just count both predictions? 1 hit, 1 miss.

I supposed 1hit-1miss =push/0

 

[iamB_rock]

I supposed 1hit-1miss =push/0

Disagree. 1-1 is 50% a push doesnt bring down (or up %)

 

[Da Jumbo Mutombo]

I supposed 1hit-1miss =push/0

Yea guy below you answered how I see it.

A push means they're off the hook for the initial wrong prediction.

1 miss and 1 hit means they get recorded accurately. Which is 50%.

 

[grover]

Yea guy below you answered how I see it.

A push means they're off the hook for the initial wrong prediction.

1 miss and 1 hit means they get recorded accurately. Which is 50%.

I see your point for sure. On the one hand, kids change their mind. On the other hand, coming up with a SWAG early on and then changing your crystal ball is a cop out. Should just say "leans" not "predict", so it doesn't count either way.

 

[Da Jumbo Mutombo]

I see your point for sure. On the one hand, kids change their mind. On the other hand, coming up with a SWAG early on and then changing your crystal ball is a cop out. Should just say "leans" not "predict", so it doesn't count either way.

Yea, listen nobody is going to bat 100% in recruiting predictions bc its fluid like you said.

But if you're confident enough to make a prediction, then it should be counted.

Otherwise, this is like what Wiltfong does with his crystal balls. Always changes them last second and keep his % high.

 

[P-Tizzle]

Accuracy change after adjusting to the new rule on conflicting predictions being a push. (10/14 compared to 10/16)

Hit Rate: 71% (10/14) [Sample size too small, don't draw conclusions from this percentage]


2023 Predictions

Date	Prediction	Outcome	Verdict
9/8	Expectation for Cole is UGA	Picked UGA	Push
6/29	Schmo isn't coming to Miami	Picked Pitt	Push
6/29	Would be shocked if Cole isn't a Cane	Picked UGA	Push
6/27	Cole on commit watch	Picked UGA	Push
6/27	Geo see Schmo committing	Picked Pitt	Push

these aren't pushes. They are bad predictions. He predicted in June that Cole would be a Cane, and he committed to UGA. That's not a push. Lindend, do Geo and Flo have something on you that you don't want to get exposed?

 

[lindend]

I've thought about the rule preventing last second flips/changes in predictions in all the prediction threads. Rather that counting the original prediction only, I think if there are conflicting predictions and the last prediction happens at least 48 hours prior to the committment, we'll rule it a Push if the last prediction is correct.

I didn't expect such a firestorm of disagreement after not a single person complained about the proposed change. Given the ferocity of the disagreement, clearly, further refinements are needed! Hopefully, we can come up with a consensus that allows a more accurate view of the hits/misses.

 

[lindend]

Kind of lets them off the hook for a previous incorrect prediction tho?

Why not just count both predictions? 1 hit, 1 miss.

That equally lets them off the hook (both solutions have the same problem). One Hit, one miss ends up a push.

Here's the thing. Recruiting is fluid. If a prognosticator knows 48 hours prior to committment that the original prediction is wrong, how do we handle it? We said the original prediction was the only one that counted. Seems rather harsh not to allow new intel to come in.

My modification this weekend, gives them no credit for either a hit or miss so it doesn't help or hurt (unlike Wiltfong's last second switches which count as hits for him).

 

[lindend]

Disagree. 1-1 is 50% a push doesnt bring down (or up %)

You have a point here. It actually does lower their accuracy rates (although not as much as the original model which only counted the miss). You and @Da Jumbo Mutombo might have the correct model.

 

[lindend]

1 miss and 1 hit means they get recorded accurately. Which is 50%.

this is the solution we'll go with until someone proposes something better.

 

[lindend]

Otherwise, this is like what Wiltfong does with his crystal balls. Always changes them last second and keep his % high.

Yea, the Wiltfong model that allows the last second flips is bogus and should be avoided at all costs. We will only take changes 48 hours prior to the commitment. Otherwise, the original prediction is the only one that will count.

 

[JD08]

I didn't expect such a firestorm of disagreement after not a single person complained about the proposed change. Given the ferocity of the disagreement, clearly, further refinements are needed! Hopefully, we can come up with a consensus that allows a more accurate view of the hits/misses.

My take: Your thread, your criteria. If someone doesn't like it, tough. It's not like you're skewing to get a desired result.

 

[grover]

I didn't expect such a firestorm of disagreement after not a single person complained about the proposed change. Given the ferocity of the disagreement, clearly, further refinements are needed! Hopefully, we can come up with a consensus that allows a more accurate view of the hits/misses.

I think @Da Jumbo Mutombo has it right. You make a "prediction" then you stick by it. You change it then you get a 0 for the prediction that was wrong, and two 0s if both the old and new predictions are wrong. Saying "lean" or "likely" or similar doesn't count either way. But once the CB goes up, it's either a 1 or a 0.

 

[lindend]

That's not a push. Lindend, do Geo and Flo have something on you that you don't want to get exposed?

Why is it that everything always has to have an ulterior motive? I put the change out there multiple days beforehand and everyone could've commented on it if they had issues with the change. But no one one did.

Its not like I did this in the dark and no one was aware of it. Nevertheless, lesson learned. For full transparency, in the future, I'll conduct a public poll for criteria changes and adopt the majority's preference.

 

[lindend]

Accuracy change after the second adjustment to the new rule on conflicting predictions being a push. (13/17 compared to 10/14)

Hit Rate: 68% (13/19) [Sample size too small, don't draw conclusions from this percentage]

2023 Predictions

Date	Prediction	Outcome	Verdict
9/8	Expectation for Cole is UGA	Picked UGA	Hit
7/20	I see Patterson in Miami's recruiting class	Picked Miami	Hit
6/29	Schmo isn't coming to Miami	Picked Pitt	Hit
6/29	Would be shocked if Cole isn't a Cane	Picked UGA	Miss
6/27	Cole on commit watch	Picked UGA	Push (already counted as a miss)
6/27	Geo see Schmo committing	Picked Pitt	Miss

 

[JHallCanes]

OP is the hero of this recruiting board and I would never want to demean his efforts here

However these prediction tracker threads have proven to be a futile exercise. It was a good thought but I think it accomplished a goal by showing insider predictions mean jack ****