FiveThirtyEight Blog- So now..... | The Boneyard

FiveThirtyEight Blog- So now.....

Fairfield Fan said:
According to that blog, UConn has a 74% chance of winning it ALL. The erstwhile odds-on favorite to win it ALL? Baylor?

As Rosanne Rosanna-Dana used to say: "Never mind."

2017 March Madness Predictions
Expand ...

I'm not following your point. Are you suggesting that Baylor was the favorite (they were, very briefly, and very appropriately).
 
Phil said:
I'm not following your point. Are you suggesting that Baylor was the favorite (they were, very briefly, and very appropriately).
Expand ...

Just tweaking Fivethirtyeight Blog's nose a bit. They kept putting UConn's chances of winning lower and lower. Until the Elite Eight games, Baylor was the favorite, and UConn less than even money. Now a day or two later, we're the CAN'T-MISS team in the tournament. Just sort of funny how they've jumped around. 538 is sponsored by ESPN, so these guys are supposed to be the pros.

Just for fun....

Wonder where Massey is going to put UConn? That site has been extraordinary in calling the wins, and also calling the final scores really well!!
 
Phil said:
I'm not following your point. Are you suggesting that Baylor was the favorite (they were, very briefly, and very appropriately).
Expand ...
I disagree with that "appropriately" part because UCONN beat a healthier Baylor team earlier in the year. Baylor did nothing to justify being the favorite in a rematch.
 
A quick look at Massey shows UConn about 83% to win semis and 68% to win it all.
SCar is 52% semis and 12% finals!
 
All 538 does is calculate probabilities based on scores of games. They're not making predictions. Baylor was putting up scores that rivaled UConn all season. It's not designed to pick upsets unless the selection team screwed up the seedings in the first place, which happens more on the men's side.
 
CocoHusky said:
I disagree with that "appropriately" part because UCONN beat a healthier Baylor team earlier in the year. Baylor did nothing to justify being the favorite in a rematch.
Expand ...
They weren't a favorite in a rematch. The brief period was after Baylor won their second game and had to win four more but before UConn won their second game so had to win five more. The probability of beating Baylor was still in UConn's favor but they had to win one more game, which slightly reduce their overall probability of winning the whole thing.
 
Phil said:
They weren't a favorite in a rematch. The brief period was after Baylor won their second game and had to win four more but before UConn won their second game so had to win five more. The probability of beating Baylor was still in UConn's favor but they had to win one more game, which slightly reduce their overall probability of winning the whole thing.
Expand ...
Still not buying it, not even briefly. I only got a B+ in probability and statistics but...
The probability of an undefeated team winning five games in a row has got to be better than the probability of 3 loss team winning 4 games in a row. Especially with a sample size of > 30.
Can we please have an retired or current math professor check my logic?
 
And "odds-on" favorite doesn't just mean the favorite. It's a favorite with a less than even odds. UConn at 1-10 is odds-on. UConn at 6-5 is not odds-on.
 
Phil said:
They weren't a favorite in a rematch. The brief period was after Baylor won their second game and had to win four more but before UConn won their second game so had to win five more. The probability of beating Baylor was still in UConn's favor but they had to win one more game, which slightly reduce their overall probability of winning the whole thing.
Expand ...
Actually, Baylor was a favorite in a rematch with UConn according to 538, after both teams had won their Sweet 16 games. UConn had slightly better overall odds to win the tournament, but only because Oregon was viewed as a much easier opponent than Mississippi State.
 
CocoHusky said:
Still not buying it, not even briefly. I only got a B+ in probability and statistics but...
The probability of an undefeated team winning five games in a row has got to be better than the probability of 3 loss team winning 4 games in a row. Especially with a sample size of > 30.
Can we please have an retired or current math professor check my logic?
Expand ...
Phil crunches numbers and work with statistics for a living. I'm a financial analyst and I can't do what he does. Guy's a numerical brainiac.
 
HuskyNan said:
Phil crunches numbers and work with statistics for a living. I'm a financial analyst and I can't do what he does. Guy's a numerical brainiac.
Expand ...
lol. Well my little B+ is probably not going to hold up against all that brain power so I'll yield back the balance of my time. :D:D
 
More than anything, this probably shows that using score differentials start to produce questionable results when teams are blowing out opponents by 30-60 points. A few days ago, 538 had this to say about the model's results:

The huge disparity between our model and the markets suggests that it’s worth having some skepticism toward our model results. In general, models are at their worst when it comes to extreme outliers, which UConn certainly has been. For one thing, it can be hard to tell their true strength, because so many of their games are fairly noncompetitive. For another, you would normally expect a team that has performed like UConn this year to be a little worse than it looks — because of regression to the mean — but in UConn’s case, we have a lot of corroborating evidence. Namely, that this season’s performance isn’t even that much of an outlier, for them.
 
And Massey now has a prediction for the score of Connecticut versus Mississippi State:

Massey Ratings

77-67: Connecticut wins!

If it were 538, that plus two bucks would get you a cup of coffee. But Massey has been surprisingly good.

Let's hope his track record remains strong!
 
Fairfield Fan said:
Just tweaking Fivethirtyeight Blog's nose a bit. They kept putting UConn's chances of winning lower and lower. Until the Elite Eight games, Baylor was the favorite, and UConn less than even money. Now a day or two later, we're the CAN'T-MISS team in the tournament. Just sort of funny how they've jumped around. 538 is sponsored by ESPN, so these guys are supposed to be the pros.
Expand ...
When a team whose probability of winning the tournament was in the 40-50% range is eliminated from the tournament (and thus has 0% chance to win), then of course the remaining teams are going to see their win probabilities go up dramatically. I don't see anything surprising or discrediting about that.

In the NFL, if a team with a 10% chance of winning the Super Bowl doesn't make the playoffs, then of course the other teams will see their chances increase (that 10% that no longer exists has to go to somebody). Why should you expect anything else?
 
JoePgh said:
When a team whose probability of winning the tournament was in the 40-50% range is eliminated from the tournament (and thus has 0% chance to win), then of course the remaining teams are going to see their win probabilities go up dramatically. I don't see anything surprising or discrediting about that.

In the NFL, if a team with a 10% chance of winning the Super Bowl doesn't make the playoffs, then of course the other teams will see their chances increase (that 10% that no longer exists has to go to somebody). Why should you expect anything else?
Expand ...

Yeah, sure. But the team that was eliminated was the one that 538 had as the favorite to win the championship! So, yes, now it makes sense to raise the odds.
 
Phil said:
I'm not following your point. Are you suggesting that Baylor was the favorite (they were, very briefly, and very appropriately).
Expand ...
The proof is in the pudding---the proof of what you see---the proof is in the loss---that the apparent experts (one who is given the job of expert)--missed the boat--- I think it was the Wizard of OZ ---the only thing you need to be a great thinker was a certificate--and he handed them out. A piece of paper is only really needed for one use. If you can do it -you do it.
 
Fairfield Fan said:
And Massey now has a prediction for the score of Connecticut versus Mississippi State:

Massey Ratings

77-67: Connecticut wins!

If it were 538, that plus two bucks would get you a cup of coffee. But Massey has been surprisingly good.

Let's hope his track record remains strong!
Expand ...
Massey has Stanford vs. USCr 66 - 67 almost even.
Massey Ratings
 
You must log in or register to reply here.

Online statistics

Members online
278
Guests online
1,490
Total visitors
1,768
Totals may include hidden visitors.

Share This Page

Forum statistics

Threads
164,065
Messages
4,380,736
Members
10,177
Latest member
silver fox


.
..
Top Bottom