NYT's Nate Silver Handicaps the Field

Donate To GopherHole

Love, love Nate Silver's work. Read his latest book "The Signal and the Noise" in January, and thoroughly enjoyed it - but then again I love things like: statistics, sports, the natural sciences, politics, odds . . .

Thanks for sharing.
 

Probably the most interesting 1st round probability there is that he gives Michigan a 90.6% chance of beating South Dakota St, quite a bit more than the other 4/13 games. If I were a Michigan fan, I sure wouldn't feel that confident.
 

mattw1067 said:
Probably the most interesting 1st round probability there is that he gives Michigan a 90.6% chance of beating South Dakota St, quite a bit more than the other 4/13 games. If I were a Michigan fan, I sure wouldn't feel that confident.
Click to expand...

Auburn Hills makes it a virtual home game for the Wolverines...I love Wolters and the Jackrabbits are a nice team, but they like to play somewhat up-tempo and I think it plays right into the hands of the Wolverines. I will be surprised if Michigan doesn't explode for 90 with an array of dunks for Hardaway and GR3.
 

ZebH82 said:
Auburn Hills makes it a virtual home game for the Wolverines...I love Wolters and the Jackrabbits are a nice team, but they like to play somewhat up-tempo and I think it plays right into the hands of the Wolverines. I will be surprised if Michigan doesn't explode for 90 with an array of dunks for Hardaway and GR3.
Click to expand...

There will definitely be more Michigan than Jacks fans. But you need to remember that Michigan State is also playing in the same building , and there is a thread on the Spartan board about how they can get SDSU gear. If the game stays close for awhile , the noise level might even favor the Jacks.
 


Well, I guess I found someone to fill my bracket for me.
 


According to Silver, there are not going to be many upsets.

-The only first round upset is MN over UCLA (9 over 8 doesn't count).
-Sweet 16 is 1-4 except for 5 Wisconsin
-Elite 8 is 1 and 2 except Florida
-3 out of 4 #1 in Final Four

I can see the reason for this being that it doesn't use historical statistics of NCAA wins and loses. For example, a 12 almost always beats a 5. This is practically a rule for any bracket maker. I think taking this as a guide along with historical stats will help. ie Look for the 12 that has the closest % chance to beat a 5. Which would be Oregon according to Silver.

Just helping out.
 

scher215 said:
Thinking the same thing lol
Click to expand...

Or maybe not.

Looking deeper at the bracket he filled out, the only lower seed he has winning in the first round is the "U". Every other game is chalk, including every 8-9 matchup.

I think the only other deviation is the 4-5 matchup between KState and wisky, where he has the red weasels going to the sweet 16 and losing.
 



I love the graphic with the width of the lines!

My calcs are very similar to Silver's. I came up with the following to win the national title:

Louisville 24.7%
Indiana 13.1%
Florida 10.6%
Kansas 8.7%
Gonzaga 6.4%
Duke 5.2%

For the Gophers...
Round of 32: 56.7%
Sweet Sixteen: 12.7%
Final 8: 4.8%
Final 4: 1.5%
National Semi: 0.5%
National Title: 0.1%
 

scher215 said:
Thinking the same thing lol
Click to expand...

Or maybe not.

Looking deeper at the bracket he filled out, the only lower seeds he has winning in the first round is the "U" along with Missouri over Colorado St. in a 8-9 matchup.

I think the only other deviation is the 4-5 matchup between KState and wisky, where he has the red weasels going to the sweet 16 and losing.
 

Very neat. I will have some question marks answered using this. Nate Silver does some great work. His electoral college stat predictions are unreal.
 

Michigan is about the worst possible matchup for South Dakota. It's a rare instance where the Jack Rabbits won't have the best/most dynamic guard on the court and Michigan's athleticism will be overwhelming.
 




GopherGold said:
I can see the reason for this being that it doesn't use historical statistics of NCAA wins and loses. For example, a 12 almost always beats a 5. This is practically a rule for any bracket maker. I think taking this as a guide along with historical stats will help. ie Look for the 12 that has the closest % chance to beat a 5. Which would be Oregon according to Silver.

Just helping out.
Click to expand...

Technically speaking, if you're going for overall accuracy you should pick all the 5's to beat the 12's. Odds are if you choose one 12 to upset a 5, you will pick the wrong one. Then you'll have two wrong games instead of just one.
 

Gopher07 said:
Technically speaking, if you're going for overall accuracy you should pick all the 5's to beat the 12's. Odds are if you choose one 12 to upset a 5, you will pick the wrong one. Then you'll have two wrong games instead of just one.
Click to expand...

True, but who doesn't like saying they called that upset?
 

It seems like 13 over 4 is the new 12-5 upset.
 


GopherGold said:
I can see the reason for this being that it doesn't use historical statistics of NCAA wins and loses. For example, a 12 almost always beats a 5. This is practically a rule for any bracket maker. I think taking this as a guide along with historical stats will help. ie Look for the 12 that has the closest % chance to beat a 5. Which would be Oregon according to Silver.
Click to expand...

From a statistically nerdy perspective, Silver gives only a 21.9% chance that all of the 5 seeds win.

Oklahoma State at 60.5%
Wisconsin at 75.6%
VCU at 71.1%
UNLV at 67.2%

Multiply all those together for 21.8%.
 

If you guys could not use this prediction method, I'd appreciate it. It's basically what I use to fill out my bracket, save any computations. And while it's never right, it seems to get you 75th percentile or better every time.
 

Silver admits (by way of this academic article, if you're interested) that a higher-risk strategy than simply picking the statistical favorites is a better way to go in pools because you're specifically picking teams that others aren't.

Rather, I suspect it’s too easy to get carried away with looking at the little stuff and that such efforts will lead you astray unless you’re in the 99th percentile of basketball dorks, at which point you might consider moving to Macao.

Instead, if you’re really serious about winning your tournament pool, there’s probably a lot more to be gained by thinking strategically. Because most tournament pools only give prizes to the top couple of places, high-risk strategies tend to be rewarded, and specifically picking teams that other people aren’t picking.

If, for instance, our forecast says that a team is a 40-60 or 45-55 underdog, it may nevertheless be the correct pick provided that other people aren’t making it (as opposed to a “trendy” upset pick, which can be a trap).
Click to expand...

http://fivethirtyeight.blogs.nytimes.com/2011/03/14/how-we-made-our-n-c-a-a-picks/

So definitely use this as some sort of guide, but not as some sort of bible. It will work on a long-term basis but not within a single given year as well.
 

Gopher07 said:
Silver admits (by way of this academic article, if you're interested) that a higher-risk strategy than simply picking the statistical favorites is a better way to go in pools because you're specifically picking teams that others aren't.



http://fivethirtyeight.blogs.nytimes.com/2011/03/14/how-we-made-our-n-c-a-a-picks/

So definitely use this as some sort of guide, but not as some sort of bible. It will work on a long-term basis but not within a single given year as well.
Click to expand...

That's the ticket - you have to intentionally pick some upsets, especially because you know some will occur. The question, of course, is where they will occur. It would be interesting to research the statistical percentage of games that are upsets and then target that proportion in your picks, selecting games that are close to going either way.
 

You must log in or register to reply here.



Top Bottom