Live by the Three, Die by the Three — Part 2 — the Deuce

Before the game last weekend against BYU (a high volume 3pt shooting team), Kansas coach Bill Self stated, “If they hit 13 three’s and we hit 3, they will win.  It’s a 30 point advantage.” That turned out to be prophetic because those were the exact numbers that each team put up and the Cougs won at Phog Allen.  This is notable because it’s something only 17 other conference opponents have been able to do in the time Self has been the boss.  That cat has won 17 conference championships in that same time.  Dang. The 3 is a deadly weapon and a giant killer if used appropriately.  

But this incident leads us back to our original question, does shooting the three win NCAA tournament games?  Not anecdotally, and not based on opinion, but when you examine the data over the last 20 years does being a high volume or high 3pt percentage team give you an advantage and an opportunity to win more games than average in the tourney?

Super Quick Recap of Part 1:

3pt shooting is a thing now.  The math bares out that shooting tons of 3’s will get you wins in the regular season, but neither high volume nor high 3pt shooting percentage alone shows a significant difference to teams advancing further than expected in the tournament. 

If you haven’t read Part 1, click here.  Then come back.

So if high volume doesn’t lead to more wins and neither does a high percentage, what does? With their powers combined, would these two criteria show us something hidden that does work?

Sort of.  And this is where it gets cool.

Theory

If you shoot a high volume of 3’s and you shoot a high percentage, together, you should be unstoppable in the tourney.

So what we did was examine the data from 2003 to 2023. First let’s look at what should be expected from each seed. 

Now, in order to make this next observation, we needed to increase the number of teams we are looking at.  For instance, if we just observe 1 seeds we have 80 data points in that time.  To make it statistically significant we tried blocking together bins with 10 seeds at the same time.  And then shifted the window from 1-10 to 2-11 to 3-12, etc. so that we could adjust and see if there was a trend across the whole field.  

And there was.  But it wasn’t what we were expecting.  

The Reality

So I’m going to intentionally make these next charts really small and I’m wondering if by just looking at the color coding, you can pick out the one important trend they show us. Please humor me and just ignore the labels for now.  Here they are:

Is it popping out to you yet?  The dark green block in the lower center right.  That block tells us that teams in that sweet spot tend to win more tournament games than they ought to, compared to their seeding historical average.  It’s like a glaring green light saying hey dummy, look at me, I’m the only way that you use 3pt shooting to win more games.  

That green light are the teams that shoot more 3’s than half of the field while also shooting a higher 3pt% than half of the field.  But notably, they are not the top 25% of 3pt attempters or top 25% or 3pt FG% teams.  They are good and they shoot more than most but they are not the huckers and chuckers of the college basketball world.

Notably, that green block is not in the far bottom right corner, like you would expect.  Down there it would be the highest volume and highest 3pt % teams—that doesn’t win you more games.  

It’s like the Goldilocks zone: they are better than most but not solely reliant on the 3 ball to win either. 

WAG’s (Wild Ass Guesses)

This data leads me to believe that the “live by the three, die by the three” philosophy doesn’t give you an edge, or competitive advantage, if you intend to outperform your seed expectation in the tournament.  If you have been a reliably high 3pt% shooting team though out the season, such as Kentucky at 40.6% or high volume lofting 32 treys per game like BYU, that alone doesn’t guarantee success.  

The tournament is a high stress neutral/road-ish environment.  That stress does affect performance and seems to not work out for teams that are reliant on this idea.

What does work?  Being good at more than just one thing.  If you shoot more threes at a better rate than others, you will probably win more games in the tourney.  But you’ve got to have a solid inside game and the ability to shift your offensive production if the three is not hitting.  

So launch away guys, just don’t put all your eggs in one basket; Cinderella is libel to upturn it.