Is the NCAA tournament a “crapshoot”

[5DollarPitcher]

Was Wisconsin decidedly better than Marquette this year?

I'm not talking about for one game in December, I'm talking about the whole season. Is it your contention that Wisconsin was a superior team to Marquette?

No. Never was.

My contention is teams from an extremely focused top group out of a field of 68 generally make it to the Final Four and win the thing at a statistically significantly higher rate. That fact in and of itself makes the tournament not a crapshoot.

[TSmith34, Inc.]

No. Never was.

My contention is teams from an extremely focused top group out of a field of 68 generally make it to the Final Four and win the thing at a statistically significantly higher rate. That fact in and of itself makes the tournament not a crapshoot.

6, 7, and 8 come up on a roll of a pair of dice at a statistically higher rate. Duh.

[wadesworld]

"I won't give you definitions...but I didn't get the results I expected, once again ended up with egg on my face, so now I am going to tell you you're wrong because of the definition of a crapshoot!"

And you don't even have the actual definition right. Sad.

[5DollarPitcher]

6, 7, and 8 come up on a roll of a pair of dice at a statistically higher rate. Duh.

Those numbers represent 3 possible outcomes in a field of 12 possible outcomes. So 1/4th of the field.

The odds of rolling one of those numbers on any given crapshoot are mathematically 44.3%.

The top 1/4th of the field of 68 (17 teams) wins the tournament 90%+ of the time.

Now explain to me how the tournament is a crapshoot.

[brewcity77]

No. Never was.

My contention is teams from an extremely focused top group out of a field of 68 generally make it to the Final Four and win the thing at a statistically significantly higher rate. That fact in and of itself makes the tournament not a crapshoot.

And yet you said this:

I am a big stats / analytics guy but there comes a point where you have to close the spreadsheet and look at the results.

In order to "look at the results" you have to look at a large sample size, not just a one-off incident. FAU beating FDU, the weakest Round of 32 team in the history of the NCAA Tournament, is not indicative of their superiority. It's looking at all the results from the entire season that tells you that.

It isn't how we crown the champion, but the best team only occasionally wins the Championship. And that's okay. That's what makes it fun.

However anyone with an iota of logic can clearly see Marquette was better than Wisconsin this year, even though Wisconsin won at Fiserv. Because one-off results don't determine the best team, only the best team on that day. Which is why the NCAA Tournament is a product of weighted randomness.

[FartyEightHours]

6, 7, and 8 come up on a roll of a pair of dice at a statistically higher rate. Duh.

You may be on to something here. I'll have to think about that the next time I fill out my bracket.

[brewcity77]

Those numbers represent 3 possible outcomes in a field of 12 possible outcomes. So 1/4th of the field.

There aren't 12 possible outcomes.

[5DollarPitcher]

There aren't 12 possible outcomes.

If you are referring to rolling a specific number, which the original poster of this was, there are only 12 possible outcomes.

[brewcity77]

If you are referring to rolling a specific number, which the original poster of this was, there are only 12 possible outcomes.

No, there are not. There are not 12 outcomes when you are rolling two six-sided dice. There are 11.

[5DollarPitcher]

No, there are not. There are not 12 outcomes when you are rolling two six-sided dice. There are 11.

You are correct.

Actually makes my math more compelling as my denominator to split the field of 68 is now lower, meaning I get bigger groups when I compare apples to apples.

Thanks!

[brewcity77]

Actually makes my math more compelling as my denominator to split the field of 68 is now lower, meaning I get bigger groups when I compare apples to apples.

It really does not. And honestly, saying there are 11 equally possible outcomes is simply not true. In reality, there are 36 outcomes because there is not an equal likelihood to rolling 2 as there is to rolling 7. You keep trying to define crapshoot in the sense that it is completely random. That's not what it is. It's unpredictable, it's a gamble, but it's not purely random. Just like the NCAA Tournament. It's a mix of unpredictable results that still favor certain outcomes over others. Just like shooting craps.

[5DollarPitcher]

It really does not. And honestly, saying there are 11 equally possible outcomes is simply not true. In reality, there are 36 outcomes because there is not an equal likelihood to rolling 2 as there is to rolling 7. You keep trying to define crapshoot in the sense that it is completely random. That's not what it is. It's unpredictable, it's a gamble, but it's not purely random. Just like the NCAA Tournament. It's a mix of unpredictable results that still favor certain outcomes over others. Just like shooting craps.

But it's not *just like shooting craps* because the most UNLIKELY outcomes in the field of craps outcomes is far more likely to occur than the most unlikely outcome in the field of possible NCAA tournament outcomes.

It's analogous to shooting craps in that both have risks, uncertainties, probabilities, etc. But it's far more predictable than rolling dice at the end of the day.

I won't perpetuate with you anymore. I actually respect your opinions and your logic. You present arguments in a logically sound manner and don't sink to personal insults immediately. And I usually agree with your takes, to be quite honest! Cheers!

[brewcity77]

And considering your preferred result got absolutely pasted in the poll you created, maybe just time to take the loss on this one.

You refused to provide a definition, so it was defined for you in the replies. The arguments made led to a decisive result that went against you, indicating that while you may not like it, Scoop on the whole agrees with the crapshoot terminology. Better luck next time.

[Judge Smails]

I agree: "But it's not *just like shooting craps* because the most UNLIKELY outcomes in the field of craps outcomes is far more likely to occur than the most unlikely outcome in the field of possible NCAA tournament outcomes."

[TSmith34, Inc.]

"Guys, guys--the standard deviation of the normal distribution in craps is greater than the standard deviation of the tournament outcomes--so your analogy is wrong!"

[5DollarPitcher]

Can't win them all

[jesmu84]

Except it's not at all statistically random because a top 6 statistically ranked team wins it 86% of the time.

Try to take my whole statement in context when responding

[Scooter22]

I answered yes, and I think so because it's a single elimination format, as many have said.  Thus, specific matchups are very important, and there is the luck factor (see KenPom).  If Purdue and FDU play 100 times, Purdue probably wins 98-99 of those.

[panda]

Is the tournament a crapshoot or are analytics wrongly evaluating certain teams?

[FartyEightHours]

The butterfly effect cannot be ignored too at the micro level.

[PointWarrior]

Is the tournament a crapshoot or are analytics wrongly evaluating certain teams?

Analytics are never wrong, at least on Scoop. 

[wisblue]

Is the tournament a crapshoot or are analytics wrongly evaluating certain teams?

If these are the two choices, then "crapshoot" is the easy answer.

The metrics evaluate teams based on their performances over several games and project scores and probabilities of winning based on those metrics. The metrics might tell us that Team A has an 80% chance of beating Team B (which would reflect a pretty large point spread). But, if Team B wins, that doesn't mean the metrics were wrong, it's just that the lower probability resulted. I'd compare it to saying that the probability of rolling a 5 or higher on a pair of dice is 83.3%. If I proceed to roll a 4, that doesn't mean the probability was wrong.

[Dr. Blackheart]

To get the definitive answer, I asked ChatGPT:

Open sidebar
NCAA Tournament: Random or Not?


Dr. Blackheart
Is the NCAA basketball tournament a crapshoot?

There is certainly an element of unpredictability in the NCAA basketball tournament, also known as March Madness. The tournament features 68 teams from across the United States, many of whom have not played against each other during the regular season. Additionally, the tournament is single-elimination, meaning that one loss can knock a team out of the competition.

With so many teams and so many games, it is difficult to predict with certainty which team will win the tournament. Upsets are not uncommon, and lower-seeded teams have been known to knock off higher-seeded teams.

That being said, the NCAA basketball tournament is not a complete crapshoot. The teams that make the tournament are typically the best teams in the country, and they have earned their spots through their performance during the regular season and conference tournaments. Additionally, teams that perform well in the tournament often have strong coaching, experienced players, and a cohesive team dynamic.

In summary, while there is an element of unpredictability in the NCAA basketball tournament, it is not completely random. The best teams still have a better chance of winning, but upsets and unexpected outcomes are always a possibility.


ChatGPT Mar 14 Version. Free Research Preview. Our goal is to make AI systems more natural and safe to interact with. Your feedback will help us improve.

[FartyEightHours]

The robots are getting smarter.  I'm not totally convinced 5dollar wasn't a bot.

[wisblue]

I don't think anyone has come close to suggesting that the NCAA tournament is completely random.

But saying that upsets and unexpected results are a ""possibility" is the understatement of the year.

There are always upsets and unexpected results, even when high seeds end up in the Final Four. Part of the crapshoot is that you never know where those upsets are going to strike.

And when you get to games involving two teams in the top 32 seeds in this age of parity, there are not many true upsets, even though they are described that way when a lower seed wins. Results like that are quite common during the regular season, but they don't eliminate the losers from competition.