Math gurus- NCAA tourney question Topic

jamier2003
Posts: 208 (0)
Hall of Famer
14 172 34 213 2
Prospect
I'm in a pool that give the following points:
round of 64: 2 points per win
Round of 32: 5 points per pick
16: 9 pts
8: 14 pts
4: 20
Championship game: 36

Plus you get bonus points by seed difference in picking the upset.  So if you took the #16 seed over a #1 seed in the opening round you'd get 2 points for the correct pick plus 15 bonus points for the seed difference.  

I took stats in college and know a little about it and can project points per round based on the probability of winning using NYT's calculations.  Here is my question- why wouldn't you pick all upsets in the 1st round with projected higher point totals?  For example-  #7Florida vs #10 UVA.  Florida has a 64.1% chance of winning resulting in 1.282 projected points (.641 X 2) but UVA has a projected point total of 1.795 ((.359 * 2) + ((10-7) * .359)).  There are additional statistical probability questions as you move on rounds but this is a simplistic view.  

Is it me or does this just seem like an upset special bracket?  In the round of 64 there are only 2 lower seeded teams in the West that have fewer projected points than the favorite- the #15 and #16 seeds, simply because their probability of winning is so small.  It just seems like if I were to pick all upsets, or at lest focus on what produces the highest point probability I could win pretty easily because of the seed difference point.  

Does that seem like a fair assumption or is there another way to look at this?

3/14/2012 2:00 PM
cornfused
Posts: 1492 (0)
Hall of Famer
2 235 11
Prospect

That's true, except there are future rounds to consider, right?  For the first round, you should be picking 1-2-14-13-12-11-10-9, but that's only if you get to repick after the next round.  Otherwise, you're SOL when those 3s and 4s make the Elite Eight or whatever.

 

I would suggest you take all of the 10s and 9s, though, and most of the 11s and 12s.  But if there's a 3-4-5-6 seed that you think could go far, you shouldn't be picking them to go out in the first round.

3/14/2012 2:20 PM
jamier2003
Posts: 208 (0)
Hall of Famer
14 172 34 213 2
Prospect
That is correct corn.  Florida actually has a higher Elite 8 point total but their potential 2nd round opponent Mizzou has an even higher point total.  Thus Florida is only considered legitmate for a 1st round win but since UVA has a projected higher total then they are the logical choice, as is Mizzou to move on to the Final 4.
3/14/2012 3:24 PM
cburton23
Posts: 496 (0)
Hall of Famer
1 289 3
Prospect
I understand your situation jamier.  I am in a draft pool tonight where you get seed # X round.  So a 1 seed that wins it all gets you 21 points, whereas a 12 who makes it to the S16 gets you 36 points.  The winner year after year is the guy who gets the upsets like Butler, VCU or a Davidson a few years back.  The problem is which team will be that team each year as I only get 8 teams I have to select carefully and haven't done a very good job as I have been in single digits the last 2 years.
3/14/2012 3:37 PM
mamxet
Posts: 1154 (0)
Hall of Famer
215 48 26 2 436 11
Rookie
Used to have one like that in my office.  The sweet spot in that pool is the bottom of the at large picks.  So, unless there is a real good reason to do otherwise you pick 13, 12, 11, 10 and 9 to win in upsets. (if those are all at large teams).  In effect, you are getting sifgnificant odds in those games.  That also leaves you the 1, 2 and 3 picks to go deep in the tourney.

You dont pick a 13 if it is a non at large

You maybe skip a couple of those 9-13 teams if you are convinced they will lose and you like the higher seed to go deep

With the recipe, you should dominate in this format

3/14/2012 4:51 PM
jamier2003
Posts: 208 (0)
Hall of Famer
14 172 34 213 2
Prospect
there are two instances where the 8's are favored by so much that it is actually better to pick them (Memphis over St. Louis and Kansas State).  I am avoiding the 14-16's all together except for BYU to upset Marquette, although I'm still questioning that one.  I have all 1's and 2's going to the elite 8.  OSU, Mizzou, UNC (toss up with Kansas) and Kentucky in the Final 4.  Kentucky vs. OSU in the final game with UK taking it home.  OSU get's a huge advantage now that Melo is out.  

The one I'm struggling with right now is Belmont vs. Georgetown.  The computer models give Belmont a 45% chance of beating Georgetown, however they give G'town the clear advantage over NCSU but not Belmont.  Tough one there.
3/14/2012 5:26 PM
isack24
Posts: 1566 (0)
Hall of Famer
81 19
Prospect
On a semi-related note, I have no idea why people think BYU (or Iona prior to that debacle) is going to beat Marquette.

Yes, I'm an MU fan, but I'm also a pessimist.  Yes, BYU has some nice size and Davies/Hartsock might present some matchup problems, but MU is going to run BYU right off the court.

Edit: To add to that, MU will crush BYU's man defense, and they are one of the best penetrating teams in the country, so the zone that frustrated Iona isn't going to work.  If MU isn't completely off, they'll win by a bunch.
3/14/2012 5:29 PM
zbrent716
Posts: 1970 (0)
Hall of Famer
90 1 58 77
Prospect
Posted by jamier2003 on 3/14/2012 5:26:00 PM (view original):
there are two instances where the 8's are favored by so much that it is actually better to pick them (Memphis over St. Louis and Kansas State).  I am avoiding the 14-16's all together except for BYU to upset Marquette, although I'm still questioning that one.  I have all 1's and 2's going to the elite 8.  OSU, Mizzou, UNC (toss up with Kansas) and Kentucky in the Final 4.  Kentucky vs. OSU in the final game with UK taking it home.  OSU get's a huge advantage now that Melo is out.  

The one I'm struggling with right now is Belmont vs. Georgetown.  The computer models give Belmont a 45% chance of beating Georgetown, however they give G'town the clear advantage over NCSU but not Belmont.  Tough one there.
Belmont over NC State.
3/14/2012 6:35 PM
dmurphy104
Posts: 2299 (0)
Hall of Famer
18 1 137 12 101
Prospect
i won one of those brackets last year. I just picked it without considering the seeds. It did help that I was only 1 of 4 to pick UCONN to win it all. I have Kentucky in this one, so I picked alot of upsets in the first round. I still worry it will come back to bite me. But with just a few upsets I should be in shape if UK does win it.
3/15/2012 9:39 AM
emy1013
Posts: 2731 (0)
Hall of Famer
899
Prospect
Seems to me that the best strategy would be to simply pick the winners.
3/15/2012 9:52 AM
theploww
Posts: 151 (0)
Hall of Famer
303 2 27
Prospect
As an earlier poster mentioned, you have to look past just the first round. Picking a 14 seed may have a higher expected value in the first round, but once you extrapolate out to the later rounds the overall expected points will probably favor the 3. It gets to be pretty complex after a round or two, but if you know what you're doing with excel you can work up a nice spreadsheet that will calculate all the expected points given the probabilities of each time winning.
3/15/2012 11:02 AM
dmurphy104
Posts: 2299 (0)
Hall of Famer
18 1 137 12 101
Prospect

I re-read the original post. I was thinking that it was points for the seed + the points for the round. Didnt realize the bonus points were only for underdogs. So, in this case, it makes alot of sense to pick the dogs.

Look at the 11-6 games. Assuming that you only pick the winner to win one game..As long as there is just one upset you come out ahead

If you pick all dogs:
1 upset = 7 points
2 upsets=14 points
3 upsets=21 points

if you piick all favorites
1 upset = 6 points
2 upsets = 4 points
3 upsets = 2 points

12-5 matchups
1 upset = 9 points if you pick all dogs, 6 points if you stick with favorites.

I would agree with picking all 10-12 seeds, and picking the rest as you normally would.

 

3/15/2012 11:18 AM
ryandaniel
Posts: 188 (0)
Veteran
22
Prospect
Posted by emy1013 on 3/15/2012 9:52:00 AM (view original):
Seems to me that the best strategy would be to simply pick the winners.

Unless you really aren't sure who will win, in which case you may as well pick the upset?

 

3/15/2012 11:19 AM
jamier2003
Posts: 208 (0)
Hall of Famer
14 172 34 213 2
Prospect
This chalk is killing me.
3/15/2012 5:29 PM
isack24
Posts: 1566 (0)
Hall of Famer
81 19
Prospect
Posted by isack24 on 3/14/2012 5:31:00 PM (view original):
On a semi-related note, I have no idea why people think BYU (or Iona prior to that debacle) is going to beat Marquette.

Yes, I'm an MU fan, but I'm also a pessimist.  Yes, BYU has some nice size and Davies/Hartsock might present some matchup problems, but MU is going to run BYU right off the court.

Edit: To add to that, MU will crush BYU's man defense, and they are one of the best penetrating teams in the country, so the zone that frustrated Iona isn't going to work.  If MU isn't completely off, they'll win by a bunch.
Like I said...
3/15/2012 5:51 PM
Math gurus- NCAA tourney question Topic
△Top△Bottom

Search Criteria

Terms of Use Customer Support Privacy Statement

© 1999-2026 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.