Sunisa Lee on balance beam

Data Deep Dive: Simulating the 2022 Postseason

It’s been a whirlwind of a college gymnastics season, starting out with a wave of COVID-related cancelations and postponements and ending with some of the whackiest scoring we’ve ever seen in the sport. We can’t be the only ones who can’t believe postseason is already here! That means it’s time to run our postseason simulator again to see who the title favorites are and where to look for the most likely upsets along the way.

Methodology

We collected all of the scores from the 2022 season and removed those from gymnasts who were known to have season-ending injuries. We wrote a script that randomly selects scores from the remaining pool to simulate a meet, with gymnasts who competed more often during the season being more likely to make the lineup in the simulated meet. We also accounted for the fact that there are more judges in the postseason by adding or subtracting 0.0125 from some of the scores, determined randomly in each simulation. Finally, we expanded the script to simulate the entire bracket, starting with the play-in rounds and finishing with the national final. We ran this script 10,000 times.

This year we added in a couple of variations to the script to see how the data would change as a result. In previous years all scores throughout the season were weighted equally, but that doesn’t take into account which teams are performing at their best going into postseason. These teams are often the ones to make more noise at regionals and nationals, so we decided to make late-season scores count more than early season scores in this year’s run. However, we still ran the numbers using last year’s method as well to compare the results.

We also made a change in how we handled non-season-ending injuries. Previously, gymnasts were excluded from the simulation if they had a known injury and hadn’t yet returned to lineups. This year, as long as the injury wasn’t known to be season-ending, we included scores from these gymnasts but gave them a significantly lower chance of being chosen for a lineup compared to everyone else.

Results

First we’ll present the results using the newer method of weighing later-season scores more heavily. The following table shows how often each team qualified to each round of the postseason across the 10,000 model runs in this simulation.

Region Rank Team 2nd Round 3rd Round Nationals Finalist Top 3 Top 2 Champion
Norman1Oklahoma100.00%99.42%96.14%84.61%74.18%57.08%30.32%
Norman16Arizona State100.00%60.44%4.96%0.17%0.01%----
Norman18Arkansas100.00%36.24%2.83%0.06%0.01%----
Norman31West Virginia45.18%1.37%0.01%--------
Norman32Arizona54.82%2.53%0.04%--------
Norman8Minnesota100.00%89.82%55.79%13.42%5.40%1.94%0.30%
Norman9California100.00%92.32%39.73%5.07%1.37%0.24%0.04%
Norman24Boise State100.00%9.80%0.27%--------
Norman25Utah State100.00%8.06%0.23%--------
Auburn2Florida100.00%99.93%99.08%86.31%78.45%61.94%37.57%
Auburn15Denver100.00%39.50%1.49%0.01%------
Auburn17Ohio State100.00%54.75%3.77%0.04%------
Auburn28Iowa State56.31%3.91%0.02%--------
Auburn36Western Michigan43.69%1.91%----------
Auburn7Auburn100.00%96.61%82.23%23.45%12.93%4.94%1.13%
Auburn10Kentucky100.00%77.51%12.62%0.50%0.12%0.02%--
Auburn23Georgia100.00%17.16%0.63%--------
Auburn27Southern Utah100.00%8.72%0.16%--------
Raleigh3Michigan100.00%99.32%90.64%63.80%49.55%32.52%15.79%
Raleigh14UCLA100.00%89.90%23.52%2.35%0.59%0.08%0.02%
Raleigh20Maryland100.00%9.27%0.13%--------
Raleigh33Towson62.92%1.04%----------
Raleigh35North Carolina37.08%0.47%----------
Raleigh6LSU100.00%97.25%75.69%23.13%13.14%5.51%1.33%
Raleigh11Missouri100.00%87.79%9.71%0.41%0.08%----
Raleigh22Iowa100.00%12.54%0.30%--------
Raleigh30N.C. State100.00%2.42%0.01%--------
Seattle4Utah100.00%98.68%93.81%72.92%53.98%32.48%12.81%
Seattle13Oregon State100.00%62.35%9.18%0.47%0.06%----
Seattle19Illinois100.00%28.08%2.12%0.03%------
Seattle29Stanford62.08%8.09%0.22%--------
Seattle34San Jose State37.92%2.80%0.03%--------
Seattle5Alabama100.00%94.42%75.50%22.08%9.94%3.22%0.69%
Seattle12Michigan State100.00%76.23%17.31%1.14%0.19%0.03%--
Seattle21BYU100.00%22.49%1.64%0.03%------
Seattle25Washington100.00%6.86%0.19%--------

Overall this bracket is much more balanced than last year, with almost no upsets predicted by the simulator. This is because the distribution of non-seeded and non-play-in teams was much improved by the selection committee, and the regional semifinals were finally arranged in a true snake format. Outside of No. 28 Iowa State being sent to a play-in and a couple of strange geographical choices, we really can’t complain about this bracket.

The Norman regional shows Arkansas as having a fairly strong chance of upsetting Arizona State in the second round while Minnesota and California are expected to easily make it through to the regional final. Oklahoma is of course a heavy favorite to make it to Fort Worth out of its home regional, but the race between the Gophers and the Golden Bears for the second qualifying spot is closer than the contenders in any other regional. The 5% chance of making nationals given to Arizona State will rise if the Sun Devils get past Arkansas, so don’t count them out either.

Auburn presents us with the only scenario where a seeded team is not expected to make it to the regional final, with Ohio State a small favorite over Denver. All model runs for Denver included only five gymnasts on vault and bars, since all other athletes who have competed on those events for the Pioneers are out for the season, so that is a huge factor in this simulation. This should be a very close race between the two teams in the regional semifinal. Florida and Auburn are heavy favorites to make it to nationals, of course, but don’t count out Kentucky with its season high of 197.650 coming only two weeks ago.

The Raleigh regional has been labeled by some to be the most intriguing, mainly due to the presence of UCLA, a team that has exceeded the 198 mark this season, but also because of Missouri’s two wins over LSU this year. Our model likes UCLA’s chances of pulling an upset more than Missouri’s, and it puts the total probability of an upset over LSU at nearly 25% overall.

The Seattle regional is an interesting one due to the injury of Luisa Blanco on the very last event of SEC championship. She was not removed from the model completely, but she was given a 25% chance of making lineups compared to others on the team, and as a result Alabama is only given a 75% chance of making nationals, with Michigan State being slightly more favored to pull the upset than Oregon State. However, we would be remiss not to point out that both of the second round meets have potentials for upsets, with Illinois and BYU both having over a 20% chance at advancing to the regional final.

Once we get to nationals, Florida and Oklahoma are the heaviest favorites to make it to the final, followed by Utah and Michigan. Auburn, LSU and Alabama’s chances are not insignificant, and the second qualifier out of the Norman regional (likely either Minnesota or California) will probably have around the same chance as well. Finally, the model picked Florida as the champion in the most number of runs (38%), which is a lower total than in the 2021 simulation, but it is enough to consider the Gators the favorites for the title. Oklahoma is second at 30% ahead of Michigan (16%) and Utah (13%). It may be surprising to see Michigan’s chances so low compared to Florida and Oklahoma given its No. 1 ranking for much of the season, but the Wolverines’ chances in this simulation were affected by the adjustment to weigh later-season scores more heavily. However, don’t underestimate how tight this race is, as all three teams posted identical 198.200s at their respective conference championship meets last weekend.

Let’s also take a look at the results using last year’s method of weighing all meets from the season equally.

Region Rank Team 2nd Round 3rd Round Nationals Finalist Top 3 Top 2 Champion
Norman1Oklahoma100.00%99.18%94.78%83.25%70.15%52.02%27.60%
Norman16Arizona State100.00%50.74%3.84%0.13%0.01%----
Norman18Arkansas100.00%45.27%3.96%0.17%0.04%----
Norman31West Virginia45.76%1.73%----------
Norman32Arizona54.24%3.08%0.02%--------
Norman8Minnesota100.00%89.83%57.15%16.82%6.51%2.28%0.46%
Norman9California100.00%90.56%39.55%7.12%1.77%0.39%0.03%
Norman24Boise State100.00%10.38%0.31%0.01%------
Norman25Utah State100.00%9.23%0.39%--------
Auburn2Florida100.00%99.93%98.94%84.24%75.23%58.08%33.21%
Auburn15Denver100.00%48.91%2.44%--------
Auburn17Ohio State100.00%45.50%2.44%0.03%------
Auburn28Iowa State48.32%2.83%0.02%--------
Auburn36Western Michigan51.68%2.83%0.02%--------
Auburn7Auburn100.00%96.18%79.21%16.52%8.89%3.09%0.63%
Auburn10Kentucky100.00%79.77%15.97%0.57%0.13%0.02%--
Auburn23Georgia100.00%15.60%0.79%--------
Auburn27Southern Utah100.00%8.45%0.17%--------
Raleigh3Michigan100.00%99.70%94.52%74.92%63.98%46.59%24.76%
Raleigh14UCLA100.00%87.94%15.90%0.90%0.24%0.07%0.01%
Raleigh20Maryland100.00%9.95%0.07%--------
Raleigh33Towson64.94%1.78%----------
Raleigh35North Carolina35.06%0.63%----------
Raleigh6LSU100.00%97.49%78.36%22.38%13.42%5.59%1.38%
Raleigh11Missouri100.00%87.23%10.94%0.44%0.13%----
Raleigh22Iowa100.00%13.06%0.21%--------
Raleigh30N.C. State100.00%2.22%----------
Seattle4Utah100.00%98.60%93.93%74.08%53.26%30.42%11.70%
Seattle13Oregon State100.00%68.93%16.10%1.08%0.13%0.02%--
Seattle19Illinois100.00%20.66%1.47%0.02%0.01%----
Seattle29Stanford66.68%9.69%0.42%--------
Seattle34San Jose State33.32%2.12%0.03%--------
Seattle5Alabama100.00%92.70%69.01%16.25%5.94%1.46%0.22%
Seattle12Michigan State100.00%73.11%16.52%1.03%0.16%----
Seattle21BYU100.00%27.28%2.38%0.04%------
Seattle25Washington100.00%6.91%0.14%--------

The most noticeable difference between the two sets of results is Michigan’s 9% higher chance of winning the national title, but there are other significant changes as well: UCLA is given less of a chance of making a run to nationals, and Denver has a higher chance of making the regional final.

Want to try this simulation yourself?

You're in luck! Adjust the settings and click the simulate button below, and tables will appear showing the results of both semifinals and Four on the Floor! We've also added in individual event winners this year using the scores from the eight teams that make nationals in each run. Click the button again to run a new simulation. Counts of wins will show up under the "Win Count" header. Be sure to share your results on social media!

Settings










Calculating...

Win Counts

Semifinal 1

Semifinal 2

Four on the Floor

Individual Winners


Article, simulation script and data by Jenna King; on-page simulation by Izzi Baskin

Like what you see? Consider donating to support our efforts throughout the year!

6 comments

  1. It took until the 7th sim for Oregon State to make nationals, but on the first sim that the team was there Jade Carey won multiple titles. While the upgrade to the script is very nice, it feels unrealistic to sim the individual events without including the individual qualifiers.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.