Simulated 2020 Tournaments

[Truth&Justise]

A number of different sites are doing simulated 2020 NCAA tournaments. Some will be done by video games. And some others will be done by vote. But most are being done by computer simulation. We've got nothing else going on, so why not discuss some of these?

One such simulation: ESPN used Lunardi's projected bracket and BPI (basketball power index, ESPN's proprietary computer metric) to simulate the 2020 NCAA tournament. But rather than run the simulation many times to create odds (which, FYI, would have given Duke a 17% chance to win it all), they ran it 2019 times and then went with whatever happened on the 2020th simulation. They then did a writeup of every game in the 2020th simulation.

Fairly pointless? Sure. But it's as good as any other distraction.

Here's the link if you'd like to read it.

And if you don't want it spoiled, read first before scrolling below.

*

*

*

*

*

*

*

*

*

*

*

*

SPOILERS BELOW

*

*

*

*

*

*

*

*

*

*

What a roller coaster tournament for Duke! Ups and downs all along the way, hard to know how to feel at the end. Running tally:

Selection Sunday
Good news: UNC is definitely out. And NC State made it, meaning the ACC got five teams in.
Bad news: Duke got a 3-seed in the Midwest and a brutal draw with tournament favorite #1 Kansas and #2 seed Kentucky.

First Round
Duke takes care of business against Belmont. A few heavyweights go home early after shocking upsets--#3 Michigan St., #3 Villanova, #5 Auburn, and #5 Michigan are all done.

Second Round
What a round! Marquette upsets #1 Kansas, removing the top team from the board and saving Wojo's job. And Providence upsets #2 Kentucky, leaving the Midwest wide open for Duke. Duke cruises by Luka Garza and #6 Iowa. Elsewhere #1 Baylor gets knocked out by Saint Mary's, #2 Florida State is upset by underseeded #7 West Virginia, and Yale makes the Sweet-16 with a win over #4 Oregon.

Sweet 16
Duke easily beats #7 Providence. Maryland beats #1 Dayton, and Yale continues its Cinderella run with a wild upset over #1 Baylor. Duke is now the clear best team left in the tournament, particularly after 2-seeds Creighton and SDSU lose. In fact, Duke is now the only team seeded 3 or better. It's ours for the taking boys!!

Elite Eight
#3 Duke loses to #4 Wisconsin by 1 point on a 3-point buzzer beater. What a devastating loss.

Final Four
A season of chaos culminates with two 4-seeds and two 6-seeds in the Final Four. #4 Wisconsin defeats #4 Maryland. #6 BYU defeats #6 UVA. Wisconsin defeats BYU for the title.

Would have been a heck of a wild ride. A third Elite Eight in a row would have been nothing to sneeze at, and we generally outperformed all the other blue bloods. But it would have been a huge disappointment to be the clear best team of the final 8 and then not take advantage. Not to mention being up 2 and losing at the buzzer. This would have joined the all-time list of devastating losses.

So...I guess it's good we avoided that heartache?

[AZLA]

Synergy. I literally thought -- maybe we at DBR can get an online gaming tourney going. Came to DBR and boom. There's the topic. I think we're all wanting some form of sports outlet. OR it would be cool if all of the college teams could select a player on their team and do a March Madness tournament -- but include all schools. Anyhow, was on conference call -- and the thought bubble popped up. Glad to see there's interest. What to do next?

[rsvman]

Yes, that would have been really disappointing, and would've been three straight losses in the Elite Eight that coulda shoulda woulda been wins.

On the other hand, if my alma mater BYU would really have made it to the championship game, that would've been really awesome.



On the whole, though, shouldn't they have told us the tournament outcome via combining all 2020 runs at it and thus give us the most-likely scenario?

[cruxer]

Yes, that would have been really disappointing, and would've been three straight losses in the Elite Eight that coulda shoulda woulda been wins.

On the other hand, if my alma mater BYU would really have made it to the championship game, that would've been really awesome.



On the whole, though, shouldn't they have told us the tournament outcome via combining all 2020 runs at it and thus give us the most-likely scenario?

I think their argument is that tracking a single simulation mimics the crazy single-elimination nature of the actual tourney, whereas you'd expect the odds to play out as predicted if you include more simulations. One wonders why they ran the other 2019 simulations though. Surely they don't really interact with each other?

Congrats to <spoiler averted>. In true kerlina fashion, we should hang the Elite-8 banner.

-c

[dukelifer]

Yes, that would have been really disappointing, and would've been three straight losses in the Elite Eight that coulda shoulda woulda been wins.

On the other hand, if my alma mater BYU would really have made it to the championship game, that would've been really awesome.



On the whole, though, shouldn't they have told us the tournament outcome via combining all 2020 runs at it and thus give us the most-likely scenario?

Lots of people were sad because there was no tournament and almost as many sad after the tournament. Fortunately I ran my own simulation and Duke won every time- and I am sticking to that.

[CDu]

I think their argument is that tracking a single simulation mimics the crazy single-elimination nature of the actual tourney, whereas you'd expect the odds to play out as predicted if you include more simulations. One wonders why they ran the other 2019 simulations though. Surely they don't really interact with each other?

Yeah, the reason of basing it on a single simulation is almost certainly to mimic the crazy single-elimination format. Maybe the argument for the other 2019 simulations is to give a sort of validity check to the results of the "official" result? In other words, you use the other 2019 outcomes to make it clear that the model isn't "flawed", in that you get a reasonable percentage of outcomes to say that the "observed" outcome isn't totally ridiculous.

[dukelifer]

Yeah, the reason of basing it on a single simulation is almost certainly to mimic the crazy single-elimination format. Maybe the argument for the other 2019 simulations is to give a sort of validity check to the results of the "official" result? In other words, you use the other 2019 outcomes to make it clear that the model isn't "flawed", in that you get a reasonable percentage of outcomes to say that the "observed" outcome isn't totally ridiculous.

I also have to wonder if it factored in Duke’s new three point weapon and shot blocker Justin Robinson. No way did the computer know what he was about to do.

[moonpie23]

#3 Duke loses to #4 Wisconsin by 1 point on a 3-point buzzer beater. What a devastating loss.

except, Jrob BLOCKS that BS and DUKE WINS!!!! (i am there)

[CameronBlue]

Lots of people were sad because there was no tournament and almost as many sad after the tournament. Fortunately I ran my own simulation and Duke won every time- and I am sticking to that.

And I bet not once did you have to endure Bill Rafferty's "How 'bout those ONIONS!"

[kako]

Hmmmm. Wisconsin winning it all is not an incredible stretch, but unlikely. But I would be more apt to believe it except that this same bracket has Yale in the Elite 8 with no #1s or #2s, St. Mary's beating Baylor (I've watched SMU several times, they would get killed by Baylor), and two #4s and two #6s in the Final Four. Vitale would call this Upset City. I'd refer to this as Upset Alternative Universe. I'd put up either of my kids' brackets (teenage girls who pick teams kind of randomly while looking at seeds) against ESPN's.

9F

[proelitedota]

Hmmmm. Wisconsin winning it all is not an incredible stretch, but unlikely. But I would be more apt to believe it except that this same bracket has Yale in the Elite 8 with no #1s or #2s, St. Mary's beating Baylor (I've watched SMU several times, they would get killed by Baylor), and two #4s and two #6s in the Final Four. Vitale would call this Upset City. I'd refer to this as Upset Alternative Universe. I'd put up either of my kids' brackets (teenage girls who pick teams kind of randomly while looking at seeds) against ESPN's.

9F

R/cbball is running the games at actual game time. Should we open the chat for it?

[proelitedota]

Yes, that would have been really disappointing, and would've been three straight losses in the Elite Eight that coulda shoulda woulda been wins.

On the other hand, if my alma mater BYU would really have made it to the championship game, that would've been really awesome.



On the whole, though, shouldn't they have told us the tournament outcome via combining all 2020 runs at it and thus give us the most-likely scenario?

That'll be all chalk with no upsets.

[Bluedog]

Hmmmm. Wisconsin winning it all is not an incredible stretch, but unlikely. But I would be more apt to believe it except that this same bracket has Yale in the Elite 8 with no #1s or #2s, St. Mary's beating Baylor (I've watched SMU several times, they would get killed by Baylor), and two #4s and two #6s in the Final Four. Vitale would call this Upset City. I'd refer to this as Upset Alternative Universe. I'd put up either of my kids' brackets (teenage girls who pick teams kind of randomly while looking at seeds) against ESPN's.

9F

Well, ESPN agrees with you and their inputs only gave Wisconsin a 1% chance to win it all. So, it's not a likely outcome, but was simply a possible outcome that was randomly chosen. They were pretty clear on their method from the onset, which I personally kinda like. It wasn't meant to indicate a likely outcome, but rather one outcome based on thousands of simulations and randomly choosing one - just as the tournament only occurs once.

[Truth&Justise]

Well, ESPN agrees with you and their inputs only gave Wisconsin a 1% chance to win it all. So, it's not a likely outcome, but was simply a possible outcome that was randomly chosen. They were pretty clear on their method from the onset, which I personally kinda like. It wasn't meant to indicate a likely outcome, but rather one outcome based on thousands of simulations and randomly choosing one - just as the tournament only occurs once.

Usually. As we're all learning, sometimes it occurs less often than that

[duketaylor]

Evening games would be tipping about now...
Gonna have to go do a fireball, or can use to disinfect something😜

[dukelifer]

I also have to wonder if it factored in Duke’s new three point weapon and shot blocker Justin Robinson. No way did the computer know what he was about to do.

Still in it. Tre with 19 points, 10 assists and nine rebounds. That will boost his NBA stock. Wow

https://nypost.com/2020/03/19/march-...-in-a-classic/

[Truth&Justise]

Still in it. Tre with 19 points, 10 assists and nine rebounds. That will boost his NBA stock. Wow

https://nypost.com/2020/03/19/march-...-in-a-classic/

An overtime Sweet 16 win over Villanova would have been sweet!

It looks like the NY post has the following as its Elite Eight:

East: #3 Duke v #4 Maryland (note, this simulation had #16 Winthrop beating #1 Dayton and then making the Sweet 16)
South: #3 Seton Hall v. #4 Louisville
Midwest: #3 Michigan St. v. #12 East Tennessee St.
West: #2 SDSU v. #4 Oregon

[dukelifer]

YouTube with a live stream of Duke Hofstra 2K-8 simulation. Jack White playing well. Tre is not.

[proelitedota]

https://www.youtube.com/watch?v=wS2R...ature=youtu.be

[dukelifer]

YouTube with a live stream of Duke Hofstra 2K-8 simulation. Jack White playing well. Tre is not.

Too close for comfort. 3 point win. White with 13 and 13. Stanley with an impressive performance. There was black Matthew Hurt in this one and Carey bore a remarkable resemblance to Zion. All that said- I got to yell at the TV. Tre hit a late shot and a free throw but must have been hurt- played poorly.