2018 NCAA Team Championship: Simulated
2018 NCAA Team Championship: Simulated
We simulated the 2018 NCAA Wrestling Tournament 10,000 times based on our preseason rankings. Check out the results!
The college wrestling preseason is a time for speculating. And while no championships are won on paper -- or computer screens -- there is no harm wondering who has the early leg up on the national competition.
We've already calculated the potential team scores using a simple scoring rubric based off our individual rankings. But thanks to our friend Dr. Irv O'Nian, an engineering professor at the university with the best wrestling program in the SEC, we've souped up our projecting power with an NCAA tournament simulator that he kindly created.
O'Nian took the last four years of tournament seeds and results and built a model that could simulate how a team would perform based off its seeds. In this case, and as we will continue to do throughout the season, we are substituting our No. 1-16 rankings for seeds.
Before we look at the simulator results, a quick caveat: These numbers and rankings are to the best of our knowledge as of right now. They will inevitably change as wrestlers switch weight classes, suffer injuries, and jockey for starting spots on their respective lineups. For instance, Jason Tsirtsis is unranked for Arizona State right now, due to the Sun Devils having two returning national qualifiers ranked at 149 and 157 instead. Should Tsirtsis establish himself as a starter, that will surely change.
Prospective line up looks: Oklahoma State | Virginia Tech | Ohio State | North Carolina | Iowa | Penn State
And now for the graph.
The results are based off running a simulated tournament using our preseason rankings as seeds 10,000 times.
To read the graph, locate a team by the color of its line (we used only the top five teams to keep things simple and legible) and find the peak of the curve. Then follow that peak down the the horizontal line to see the most likely number of team points that school will score at the 2018 NCAA Tournament, based on our preseason rankings and simulation model.
We can also calculate the average team score of all 10,000 of our simulated tournaments. Those average scores for each team can be found in the table below.
Not only is the race for the team title razor close -- which is to be expected when the top two teams both have five wrestlers in the top 20 pound-for-pound rankings -- but the battle for third is also shaping up to be be a barn burner as well.
Another thing you can calculate with our simulator is the range of likely scores each team will end up with in March. By breaking things down by quartile, you can estimate that 25 percent of the time a team will score a certain amount or lower, and 25 percent of the time the team will score a certain amount or higher. These are your worst and best case scenarios, basically.
And what's left in the middle will be a range of scores you'd expect to see half the time. A chart of the percentile breakdown for our top five teams is below.
So, for example, if the 2018 NCAA Tournament was held simultaneously 10,000 times in 10,000 alternate dimensions, you've expect Penn State to score between 119 and 136 points in 5,000 of those extra-dimensional championships.
Finally, according to our model, Penn State will win the title approximately 57% of the time, while Ohio State will be crowned champs approximately 42% of the time. Which means that of those 10,000 pretend tournaments, Penn State and Ohio State does NOT win in only about 100 instances.
Hopefully you found this preseason perspective useful; however, it can't be stressed enough that this is a super-early premature preseason guesstimate about how the 2018 NCAA Tournament will play out. It's as good as we can do in mid September, but your real best bet is that by late April, our best guess will be something much different.
We've already calculated the potential team scores using a simple scoring rubric based off our individual rankings. But thanks to our friend Dr. Irv O'Nian, an engineering professor at the university with the best wrestling program in the SEC, we've souped up our projecting power with an NCAA tournament simulator that he kindly created.
O'Nian took the last four years of tournament seeds and results and built a model that could simulate how a team would perform based off its seeds. In this case, and as we will continue to do throughout the season, we are substituting our No. 1-16 rankings for seeds.
Before we look at the simulator results, a quick caveat: These numbers and rankings are to the best of our knowledge as of right now. They will inevitably change as wrestlers switch weight classes, suffer injuries, and jockey for starting spots on their respective lineups. For instance, Jason Tsirtsis is unranked for Arizona State right now, due to the Sun Devils having two returning national qualifiers ranked at 149 and 157 instead. Should Tsirtsis establish himself as a starter, that will surely change.
Prospective line up looks: Oklahoma State | Virginia Tech | Ohio State | North Carolina | Iowa | Penn State
And now for the graph.
The results are based off running a simulated tournament using our preseason rankings as seeds 10,000 times.
To read the graph, locate a team by the color of its line (we used only the top five teams to keep things simple and legible) and find the peak of the curve. Then follow that peak down the the horizontal line to see the most likely number of team points that school will score at the 2018 NCAA Tournament, based on our preseason rankings and simulation model.
We can also calculate the average team score of all 10,000 of our simulated tournaments. Those average scores for each team can be found in the table below.
| SCHOOL | AVG SCORE |
|---|---|
| Penn State | 127 |
| Ohio State | 123 |
| Michigan | 81 |
| OK State | 78 |
| Missouri | 61 |
Not only is the race for the team title razor close -- which is to be expected when the top two teams both have five wrestlers in the top 20 pound-for-pound rankings -- but the battle for third is also shaping up to be be a barn burner as well.
Another thing you can calculate with our simulator is the range of likely scores each team will end up with in March. By breaking things down by quartile, you can estimate that 25 percent of the time a team will score a certain amount or lower, and 25 percent of the time the team will score a certain amount or higher. These are your worst and best case scenarios, basically.
And what's left in the middle will be a range of scores you'd expect to see half the time. A chart of the percentile breakdown for our top five teams is below.
| SCHOOL | 25TH PERCENTILE | 75TH PERCENTILE |
|---|---|---|
| Penn State | 118 | 137 |
| Ohio State | 111 | 135 |
| Michigan | 70 | 93 |
| OK State | 67 | 89 |
| Missouri | 51 | 71 |
So, for example, if the 2018 NCAA Tournament was held simultaneously 10,000 times in 10,000 alternate dimensions, you've expect Penn State to score between 119 and 136 points in 5,000 of those extra-dimensional championships.
Finally, according to our model, Penn State will win the title approximately 57% of the time, while Ohio State will be crowned champs approximately 42% of the time. Which means that of those 10,000 pretend tournaments, Penn State and Ohio State does NOT win in only about 100 instances.
Hopefully you found this preseason perspective useful; however, it can't be stressed enough that this is a super-early premature preseason guesstimate about how the 2018 NCAA Tournament will play out. It's as good as we can do in mid September, but your real best bet is that by late April, our best guess will be something much different.