Big Ten Season Forecast

#1      

danielb927

danielb927
Orange Krush Class of 2013
Rochester, MN
danielb927
Looking forward to the start of Big Ten play this afternoon! I've got my program dusted off (nothing broke during the off-season this year, amazingly!) and just ran 100k simulations of the full Big Ten season. I'll try to keep these updated at least twice a week throughout the year so we can see what changes as the results play out. If anyone has suggestions for new wrinkles to add, feel free to post or PM me!
 
#3      

danielb927

danielb927
Orange Krush Class of 2013
Rochester, MN
danielb927
C0sjRzWXEAAsuFb.jpg:small


First, the overall forecast. As a reminder, I use "0%" to indicate scenarios that did happen at least once, but <0.5% of the time. Noteworthy this time is that every team landed in every seed at least once, even though I only ran 100k iterations! Last year, with a million simulations, Rutgers never finished in the top 3 and Purdue/MSU never finished last. I wouldn't lay much cash on Rutgers to win the conference, but it appears to be at least an order of magnitude more likely than last year. The parity in the conference shows up other ways, too--there are four teams (tOSU through MSU) with a 1% or better chance of finishing first OR last!

The forecast for the Illini is similar to last year, but a little sunnier. We entered both conference seasons 10th in the Big Ten by Pomeroy ranking. Last year, breaking down our finish chances into the big ten tournament bye categories (double bye 1-4, single bye 5-10, and no bye 11-14) our outlook was 1%-52%-47%. This year that comes out to 9%-60%-31%.

C0skgTuXUAA3RkT.jpg:small


The other thing I started looking at last year was the effect of tiebreakers on the Big Ten tournament seeding. That's not really of huge interest in terms of important things like making the NCAA tourney, but can be important for making the cutoffs for byes or getting to own the top seed. Tiebreaker effects start to emerge as results are finalized between teams, but this early they're pretty minimal. The chart shows the difference in percent chance of a seed between a case with real tiebreakers and a case where all tiebreakers are a coin toss.

The only thing that's noteworthy is that tiebreakers give Purdue a bump for the top seed--almost certainly a result of their schedule, which includes only one game against Wisconsin (in West Lafayette). And it's not a huge effect, only 1% or so.
 
#5      

danielb927

danielb927
Orange Krush Class of 2013
Rochester, MN
danielb927
Simulations are run based on KenPom's current data. So, big upsets can have a double outcome: limiting the scenarios that still remain possible AND changing the strengths of the teams in the next simulation.
 
#7      

GameOfThrows

G
GameOfThrows
Good stuff. My predictions for the BIG season are in line with that and probably similar to everyone else:

Tier 1
Purdue
Wisconsin
Indiana

Tier 2
Michigan
Ohio St.
Michigan St.
Maryland
Illinois
Northwestern
Iowa
Rutgers

Tier 3
Nebraska
Penn St.

After the top 3, the conference is wide open and schedule will have a huge impact on the final standings. Our ceiling is 4th. We have as much size, depth and experience as anyone. We lack elite players (future NBAers) with all due respect to Malcolm. Gonna be a lot of close games. Let's hope our experience and depth help us win more than we lose. We have a decent chance to defend our home court this season. If we can win 8 or 9 of our home games, then we would only have to steal 2-3 road games to get to 11-7 which coupled with one BIG tourney win, should get us in the Dance.
 
#8      

Fighter of the Nightman

Fighter of the Nightman
Chicago, IL
Fighter of the Nightman
Thanks for this! Though I tend to be a more mathematical/left-brained person in a lot of aspects of my life (finance major, work in commercial banking), I tend to look at sports in a less "brainy" way. :D My guesses:

Tier One - Clearly the top three teams in the Big Ten
Wisconsin
Purdue
Indiana

Tier Two - Has the horses to compete with Tier One but also an obstacle
Ohio State
Michigan State
Illinois
Michigan
Maryland

Tier Three - Can beat teams in Tier Two easily, but I think they'll have worse records
Minnesota
Iowa
Northwestern

Tier Four - Have no business beating the above teams, and if they do it's a major upset
Penn State
Rutgers
Nebraska

I can't see us getting up to third ever, but I am cautiously optimistic that we can be in the upper half of tier two if we keep playing well, improving and hopefully get a confidence-boosting win at UMD today! :thumb: That would put us around 4th or 5th in the Big Ten, which I imagine we'd all take gladly!
 
#9      

danielb927

danielb927
Orange Krush Class of 2013
Rochester, MN
danielb927
Updated stuff going into this weekend's games. The first chart is the full effect based on the results so far and most recent KenPom rankings. Since the tiebreaker effect is still pretty small, the second chart shows the difference between simulations with and without updating the strengths of the teams based on their first results.

Example of how to read this: Purdue jumped from a 41% to 46% chance to win the top seed. The second chart tells us that one fifth of that jump (1%) is due to Pomeroy thinking they're a better team now, while the other part (4%) is because they already have a conference win under their belt (in addition to the other 5 results from Tuesday and Wednesday).

Other noteworthy bits
- IU's chance at a top 4 seed dropped from 75% to 47%. About half of that drop was due to their ranking change, the other half due to the game result itself (plus other wins by top teams).

- Illinois bottom 4 chance jumped from 31% to 48%, with over half of that change being due to our raking drop after the blowout loss.

- Iowa had a similar increase in bottom 4 chance (37% to 48%), but only a quarter of that was due to the game result itself. Getting blown out by 24 may tell you to update your outlook by 3x or 4x more than just losing to a top team.

This does strike me as interesting because we talk a lot about how much one loss really means. Obviously any one particular loss isn't devastating, as the Big Ten champion usually finishes the season with at least 3. This quick analysis seems to bear that out, indicating that the most important thing about a loss or win could be what it tells you about the strength of a team, not what it does to the team's conference record.

C09FogYUkAAIKRd.jpg:small


C09GsJDUQAAnlmI.jpg:small
 
#10      

DragonIlliniFan

D
DragonIlliniFan
Fighter of the Nightman said:
Thanks for this! Though I tend to be a more mathematical/left-brained person in a lot of aspects of my life (finance major, work in commercial banking), I tend to look at sports in a less "brainy" way. :D My guesses:

Tier One - Clearly the top three teams in the Big Ten
Wisconsin
Purdue
Indiana

Tier Two - Has the horses to compete with Tier One but also an obstacle
Ohio State
Michigan State
Illinois
Michigan
Maryland

Tier Three - Can beat teams in Tier Two easily, but I think they'll have worse records
Minnesota
Iowa
Northwestern

Tier Four - Have no business beating the above teams, and if they do it's a major upset
Penn State
Rutgers
Nebraska

I can't see us getting up to third ever, but I am cautiously optimistic that we can be in the upper half of tier two if we keep playing well, improving and hopefully get a confidence-boosting win at UMD today! :thumb: That would put us around 4th or 5th in the Big Ten, which I imagine we'd all take gladly!
Click to expand...

Have you gotten to see Northwestern play this year?
As much as I hate to say, they could be a dark horse to contend for a top-2 or 3 seed when all is said and done. I think they finish 4th.
 
#11      

Konnie

K
Western Suburbs
Konnie
DragonIlliniFan said:
Have you gotten to see Northwestern play this year?
As much as I hate to say, they could be a dark horse to contend for a top-2 or 3 seed when all is said and done. I think they finish 4th.
Click to expand...

I have been to a couple of Northwestern games and watched them play MSU. Whether you like Collins or not, he has them ready to play and has been able to coach them up. Right now I would suggest that Northwestern will beat Illinois.
 
#12      

alexander58

A
alexander58
Konnie said:
I have been to a couple of Northwestern games and watched them play MSU. Whether you like Collins or not, he has them ready to play and has been able to coach them up. Right now I would suggest that Northwestern will beat Illinois.
Click to expand...

Nebraska won at Indiana, too. This isn't your father's big ten. Any conference win will be hard fought and all the teams really compete.
 
#15      

EfremWinters84

E
S. Carolina
EfremWinters84
With the benefit of now having seen each team play (roughly) two games......

1. IND
2. WISC
3. MSU
4. PURD

5. MD
6. MICH
7. OSU
8. ILL

9. NW
10.IOWA
11.MINN
12.NEB

13.PSU
14.RUT
 
#16      

TEYPAY

T
Springfield
TEYPAY
Nebraska wins at Indiana (1) and at Maryland (5) and they are 12....????
 
#17      

Illini1221

I
Peru,IL
Illini1221
EfremWinters84 said:
With the benefit of now having seen each team play (roughly) two games......

1. IND
2. WISC
3. MSU
4. PURD

5. MD
6. MICH
7. OSU
8. ILL

9. NW
10.IOWA
11.MINN
12.NEB

13.PSU
14.RUT
Click to expand...

I'd also have Minnesota higher too. I know they lost to MSU, but they're a good team. Mason is a stud.
 
#24      

danielb927

danielb927
Orange Krush Class of 2013
Rochester, MN
danielb927
Here's the latest, before the Wisconsin-Indiana game tonight.

C1SUaxxXgAAF5GH.jpg:small
 
#25      

illinifan4249

I
Space Coast, FL
illinifan4249
I imagine that if Wisconsin can also take care of Purdue then they will be greater than 50% to get the 1 seed. Wisconsin is coming on strong.