Hoop-a-nator


2021 Hoop-a-nator Game Changer

For the 2021 NCAA tournament the Evolytics team used our custom built, Predictive Analytics application—Hoop-a-nator—to enter bracket contests and bet on the NCAA tournament games in Las Vegas. The new and improved tool included a new calculation model and interface, and a new name and branding. We were so confident in it’s predictions, or maybe we were just crazy, that we put up $1000 of our own cash to bet on the games, with the proceeds going to a really deserving local charity, a no-kill animal shelter, the Great Plains SPCA.

In addition, for every game we predicted right from the Sweet 16 to the Championship game, we sponsored one pet adoption, with the goal of helping to find a new home for up to 15 dogs and cats.

See all the results below!

split-block-image-21669

DOMINATE YOUR BRACKET WITH AI

Whether you’re a sports fan or not, now you can dramatically improve your odds of picking winning teams in the NCAA tournament by putting 30+ years of game data, math, and predictive analytics on your side. Just pick two teams to see predicted winners, point spreads and total points from two different models.

Sponsored Adoption Results

With a 66% success rate, we’re pretty happy to have made a BIG Difference with Data for the pets at the Great Plains SPCA. Here are the pets adopted through the game-winning predictions of Hoop-a-nator.

APRIL

BOOMER

CAROLINA

HORTON

MABLE

TAIL SPIN

TABBY

MACY

NENA

UNO

Betting Results & Analysis

FINAL 4 SUMMARY

Competitive Bracket Challenges

FINAL TOURNAMENT RECORD

35 / 64

FINAL CONTEST STANDINGS

ESPN

2,283,090 / 14,700,00

CBS

121,248th Place

YAHOO

62,632 / 466,937

CAPITAL ONE

209,000th Place

Vegas Betting

Games Won

32

Games Lost

30

Pushes

1

Starting Balance

$1,147.61

Ending Balance

$1,006.18

Final 4 Bracket and Betting Analysis

Our strategy in this final round was to minimize our risk to keep our initial $1,000 investment and donation to the Great Plains SPCA. We did try to increase the total amount as much as possible so we used our $146 earnings to date to bet with. In the end we basically broke even, but we learned a lot along the way.

Our biggest success was in picking the outright winners of the Sweet 16, Elite 8 and Final 4 games, and sponsoring a pet adoptions every time we won. We went 10/15 and sponsored each pets adoption fee to help find a new home!

This was a great experiment to see how well our new Bayesian model would hold up to a real world test by betting against the Vegas point spreads. We had a lot of fun picking strategies, watching the games and analyzing the results. 
We learned that it’s incredibly hard to pick the winner of one game with accuracy. There are just too many variables in a single game. However across a large set of games, we were able to beat the average and come out slightly ahead. Next year, we can use the data we gathered to help inform our betting strategies and are considering including the tournament games in our calculations, and putting more weight on recent years game data.


Set your marketing channel mix next quarter, based on last year’s data.

Data Science and Predictive Analytics can do that too.

LET’S TALK.

ELITE 8 SUMMARY

Competitive Bracket Challenges

ELITE 8 WINS

2 / 4

SWEET 16 CONTEST STANDINGS

ESPN

2,871,474 / 14,700,00

CBS

148,848th Place

YAHOO

94,508 / 466,937

CAPITAL ONE

372,000th Place

Vegas Betting

Games Won

30

Games Lost

29

Pushes

1

Starting Balance

$1,139.36

Ending Balance

$1,147.61

Elite 8 Bracket and Betting Analysis

This round was tough, we could have easily won all four games if it wasn’t for a last second hail-mary 3 pointer by Oregon State to beat the spread against Houston. Then Arkansas came back to beat their spread after being way down the entire game, up until the last few minutes. All-in-all we’re happy to have broken even at 2-and-2 and retain our initial investment with a decent return that can still be donated to a good cause, the Great Plains SPCA. We also continue to move up in the rankings in all our competitive brackets, and have 2 of the 4 teams in the final four remaining, including both teams we selected in the championship game.

At this point, overall we’re at a 50% win rate with a 14.7% return on our investment. The combination of our machine learning model plus our betting strategy has created the positive results. One mistake we commonly see people make is turning the decision making over to their models, and not using it a tool to inform. You have to do both.

Going into the next round we recognize that we’re moving into more of a betting situation than a data informed one since there are only two games. We’re in a tough spot because this will be like trying to pick a winning stock using data from an index fund. We plan on betting on those games, but our wager will most likely be fairly conservative unless the odds change dramatically away from what our model is predicting.

SWEET 16 SUMMARY

Competitive Bracket Challenges

Sweet 16 WINS

5 / 8

SWEET 16 CONTEST STANDINGS

ESPN

3,438,133 / 14,700,00

CBS

172,040th Place

YAHOO

110,565 / 466,937

CAPITAL ONE

410,000th Place

Vegas Betting

Games Won

28

Games Lost

27

Pushes

1

Starting Balance

$989.00

Ending Balance

$1,139.36

Sweet 16 Bracket and Betting Analysis

If feels great to be in the black after this weekend’s wins and making money to help the Great Plains SPCA. In addition, we picked 6 of the 8 outright winners and are sponsoring 6 cat adoptions! We gained a little ground in all our bracket contests but it still looks unlikely that we’ll win those—time will tell.

For this rounds betting strategy, we decided to lower our bankroll a little to $500. We didn’t see any games that we felt we had a competitive edge over Vegas. Our largest margin was on the Baylor vs. Arkansas game but it’s only 2.6 points and we’re not feeling overly confident based on recent game outcomes.

Our strategy for this round is to bet evenly across the four games based on the Bayesian Model and let the chips fall where they may. We do feel good that the underdog has covered the spread 48% of the time so far, and we have 2 underdogs picked to do it in this round. Fingers crossed that that continues!

ROUND OF 32 SUMMARY

Competitive Bracket Challenges

ROUND OF 32 WINS

10 / 16

ROUND OF 32 CONTEST STANDINGS

ESPN

5,746,630 / 14,700,00

CBS

267,947th Place

YAHOO

179,254 / 466,937

CAPITAL ONE

565,000th Place

Vegas Betting

Games Won

23

Games Lost

24

Pushes

1

Starting Balance

$922.70

Ending Balance

$989.00

Round of 32 Bracket and Betting Analysis

We went into the round of 32 having learned a lot from the previous round, but still treated this as a learning and data gathering opportunity. We kept our bets conservative and spread across all the games to give ourselves enough data, and hopefully funds, to maximize winnings in future rounds for the Great Plains SPCA.

This year has produced some of the biggest upsets in recent memory. We knew there were still a lot of variables our model couldn’t account for, like COVID illness, that the Vegas odds were adjusting to. So, we updated our betting strategy to focus on the teams in the middle of our point spread variation against the Vegas odds. Putting more money into those that fell into the 1.5 to 3.5 point variations. See the visualization above for more details.

In the end our strategy proved effective compared to the previous round, and we made almost all of our money back. In addition, we moved up in the rankings for three of our four competitive brackets. Knowing that each round moving forward will have fewer and fewer games, we may have to consider adjusting our strategy and placing bigger bets where we see an advantage, and smaller bets where there’s not.

ROUND OF 64 SUMMARY

Competitive Bracket Challenges

ROUND OF 64 WINS

19 / 32

ROUND OF 64 CONTEST STANDINGS

ESPN

10,212,642 / 14,700,00

CBS

550,746th Place

YAHOO

174,641 / 466,937

CAPITAL ONE

649,000th Place

Vegas Betting

Games Won

16

Games Lost

15

Pushes

1

Starting Balance

$1000.00

Ending Balance

$922.70

Round of 64 Bracket and Betting Analysis

For this first round, we decided to treat this as an experiment to test and learn from. To gather the most data, we bet on all the games, but we also wanted to make money for the Great Plains SPCA. So we built a betting calculator, the one you see above, to scale our bets based on the difference in the Vegas odds vs. our Bayesian Model’s point spread prediction. We knew we couldn’t account for this year’s unusual challenges and COVID issues, like teams dropping out, or missing players due to illness. So we decided to ignore all that and just include it in our analysis and future betting strategy if they influenced the results.

Overall this was one of the wilder first rounds in many years. There are always surprises and this year was full of them. From #15 Oral Roberts upsetting Ohio State to Virginia dropping out of the tournament. Some of our big bets paid off, but we really struggled in the closer point spread-difference games.

At the end of the Round of 64 we may be out of the running to win the bracket contests, but we still feel good about our chances to make money betting against the point spreads from Vegas. Better yet, we now have betting data to inform our strategy in the next round.