ADVERTISEMENT

2021 Probabilities (SPOILER ALERT)

Unbiased_football_fan

Well-Known Member
Aug 18, 2006
77
741
1
Frederick, MD
DO NOT SCROLL DOWN IF YOU DON’T WANT TO KNOW

I got a chance to run WrestleSim with the actual seeds. Not much has changed from the estimates using my projected seeds. Iowa is a clear favorite. If they don’t win it would be an epic failure.

These estimates are based on 3000 realizations (replicates) using Monte Carlo sampling. I had to run more than normal to pick up the low probabilities of winning. The good news is Penn State is a solid favorite for second. I think we should be happy our young team is that good. Imagine how we would feel if a bad year for the football team is #2. The bad news is there is quite a gap between Penn State and Iowa. There is a large group of teams that will be in trophy contention, so that will make the tournament interesting if you are a fan of one of those schools.

Here are the schools, probability of first, second, third, and projected team score. If you use this information and lose and need to get a hawkeye tattooed on your butt I am not responsible. However if you wager and win I will take the customary 5% cut. Numbers may not add to 100 due to roundoff.

Iowa, 98.1%, 1.8%, 0.1%, 126 pts
PSU, 0.9, 51.6, 24.5, 84
Missouri, 0.6, 15.6, 16.9, 71
NCSt, 0.4, 13.4, 19.0, 72
ASU, 0.1, 7.5, 14.8, 68
VT, 0.1, 6.4, 11.9, 66
Michigan <0.1, 3.5, 7.6, 63
Nebraska, <0.1, 1.8, 5.4, 58

I will add that this year will be LESS predictable than normal because the seeds are likely to be off. The committee primarily went by records and conference finish. This penalizes Big Ten wrestlers who may have many more ranked wrestlers in conference and didn’t get to pad their results with out of conference results. This also means that some branches of brackets may get totally blown up. I like the beauty of chaos. Only 14 points separates 3rd from 8th!

Penn State can make up the 42 point gap but it won’t be easy. I ran a variety of scenario analyses (conditional results) to see how PSU could get there. Below are the keys:
  • Need to beat Iowa at bonus points. The model says based on seeds Iowa should score about 21 and PSU 13. I think PSU can actually outscore Iowa in bonus or break even. Yes even with Spencer Lee bonusing everyone. Iowa does not score a lot of bonus points against Top 16 wrestlers (after the first round), but pretty much nobody does. Watch for the hawks bonus (outside of Lee) to drop way off after round 1. (+10 to 12) QUIZ: What is the average bonus rate of a tournament wrestler against a Top 16 opponent? Answer: 6-7%
  • Need Lee to beat Eierman. Of course Lee has to get there first which is no easy task as both Parker and Rivera will be tough. (+8)
  • Need 1 or 2 from Berge, Howard, Kerk, Beard, J. Lee to step up. This group is only expected to earn about 15-20 points. One or two AA’s out of this group would net points (+10-12)
  • Need 1 or 2 from Warner, Young, DeSanto, Marinelli to underperform. I didn't include Lee, Cass, and Kemerer because they are unlikely to underperform. Warner and Young are Warner and Young, meaning you wrestle 3-2 matches they sometimes come back to bite you. DeSanto looked like he may have some hurt fingers. I have to say DeSanto showed great sportsmanship in the loss to RBY. Marinelli has a horrible draw IMHO. Bullard is tough, Griffith is very tough, and Lewis is fantastic when healthy. If he survives those landmines the championship will be a cakewalk. Also, if guys are cutting too much I think it is day 2 where it gets them. (+10-12)
So that is a lot that needs to go right (wrong), probably not as low as the 1% numerical model result (maybe more like 6-8%). It is certainly possible but not likely.

Here is your viewing guide (related to above, if you are like me all of the tourney is on your viewing guide!):

Round 1 (not a lot)
Howard – Heinselman. You can do this Robbie! You are so close.
No PSU upsets
Iowa upsets

Round 2 (now we get more interesting)
Murin-Lovett
Young-Berge (bigger for Iowa, not a huge deal if Berge loses)
Brands-Bolen
Warner-Coleman/Davison
Beard-Schultz
Kerk-Wood
Howard or J. Lee if they win their first. Wittlake is nasty. If Howard can beat Heinselman that part of the bracket is ripe for a deep run.
No PSU upsets
Iowa upsets

Things could change after Round 2. That tiny crack may slam closed or open up. Stay tuned, I will update each night. Once again, don't read this thread if you don't want to know!
 
Love these two lines

If you use this information and lose and need to get a hawkeye tattooed on your butt I am not responsible. However if you wager and win I will take the customary 5% cut.

Isn't that kind of like heads I win, tails you lose? Just kidding, great stuff! I always look forward to reading your posts.
 
Round 2 (now we get more interesting)
Murin-Lovett
Young-Berge (bigger for Iowa, not a huge deal if Berge loses)
Brands-Bolen
Warner-Coleman/Davison
Beard-Schultz
Kerk-Wood
Howard or J. Lee if they win their first. Wittlake is nasty. If Howard can beat Heinselman that part of the bracket is ripe for a deep run.
No PSU upsets
Iowa upsets

Things could change after Round 2. That tiny crack may slam closed or open up. Stay tuned, I will update each night. Once again, don't read this thread if you don't want to know!
Always enjoy reading these!

While this might be what the simulation said, not quite sure I can buy into the sentence above. IMO this is also a big match for Berge to win. Two reasons:
1. Good shot at reaching the semis -- of the 4 top seeds, Dellavecchia is easily the most favorable matchup. If he gets this far, 4th place isn't out of reach -- the Hidlay/Carr loser would be in the other consi semi.

2. Risk of getting upset in consis. Berge wrestles a lot of low-scoring close matches, always risky at nationals. Consis means 2x as many matches, thus 2x the risk.

Borrowing from Roar's thread: if Berge beats Young and Dellavechia, he's clinched 6th place. If he loses to Young, clinching 6th requires 4 consi matches with only 1 of them being relatively easy.

To be clear: the potential consis opponents are guys Berge can beat -- but the risk doesn't shrink with more matches.
 
I think we are on the same page and I agree 100% with what you wrote.

What I meant was Berge as a #12 isn't expected to earn many points so if he loses a match to a much higher ranked opponent it doesn't change the outcome much (this is expected), I was going by how the computer looks at this result. He would still need to win other matches against other more highly seeded opponents to place and get a bunch of points. A #12 may beat one better seeded guy but they don't usually beat numerous, that is why many guys seeded #10 to #33 don't usually place. But as you note, if Berge (or Young) loses he would need to go through a much more difficult path to earn his points, and for a guy that wrestles close matches (either guy) that is the kiss of death.

Don't get me wrong I think Berge can absolutely win this match. Full strength Berge is better than #12. There are only three guys I don't think he could beat and they aren't in his part of the bracket. Young is the type of opponent that Berge has better success with. I would like to see Berge attack and see what happens. When he pulls the trigger he can be very quick. I'd rather see a guy lose 7-2 because he took educated chances. Can't wait for the perfect chance (although maybe with Young you can) or with 45 seconds left against a guy with good defense.

Dellavechia can be beat by either guy, so from that standpoint it is important to both teams. If Berge went from 2 to 12 points that is a big chunk of those 40.

Guys seeded #1 to #8 do not place about 25% of the time (averaged over all seeds). For guys seeded #7 and #8 it is closer to 50%. This makes sense as a #7 has to beat a #10 and a #8 has to beat a #9 (usually) and in theory those guys are similar. But overall it is hard for a guy seeded #11 to #33 to AA (these guys should really be applauded no matter who they wrestle for).

A #12 has to beat a #5 and likely somebody else higher seeded than he or get lucky and have another upset happen to a higher seeded guy in his path. (this is assuming seeds are correct, this year I think we will find seeds may reflect what was earned but not ability/strength)
 
I think we are on the same page and I agree 100% with what you wrote.

What I meant was Berge as a #12 isn't expected to earn many points so if he loses a match to a much higher ranked opponent it doesn't change the outcome much (this is expected), I was going by how the computer looks at this result. He would still need to win other matches against other more highly seeded opponents to place and get a bunch of points. A #12 may beat one better seeded guy but they don't usually beat numerous, that is why many guys seeded #10 to #33 don't usually place. But as you note, if Berge (or Young) loses he would need to go through a much more difficult path to earn his points, and for a guy that wrestles close matches (either guy) that is the kiss of death.

Don't get me wrong I think Berge can absolutely win this match. Full strength Berge is better than #12. There are only three guys I don't think he could beat and they aren't in his part of the bracket. Young is the type of opponent that Berge has better success with. I would like to see Berge attack and see what happens. When he pulls the trigger he can be very quick. I'd rather see a guy lose 7-2 because he took educated chances. Can't wait for the perfect chance (although maybe with Young you can) or with 45 seconds left against a guy with good defense.

Dellavechia can be beat by either guy, so from that standpoint it is important to both teams. If Berge went from 2 to 12 points that is a big chunk of those 40.

Guys seeded #1 to #8 do not place about 25% of the time (averaged over all seeds). For guys seeded #7 and #8 it is closer to 50%. This makes sense as a #7 has to beat a #10 and a #8 has to beat a #9 (usually) and in theory those guys are similar. But overall it is hard for a guy seeded #11 to #33 to AA (these guys should really be applauded no matter who they wrestle for).

A #12 has to beat a #5 and likely somebody else higher seeded than he or get lucky and have another upset happen to a higher seeded guy in his path. (this is assuming seeds are correct, this year I think we will find seeds may reflect what was earned but not ability/strength)
Yes, this, all of this.

IMO going fwd you may have to exclude 2021 results in order to simulate the likelihood of each seed's progression. Too many wacky seeds not commensurate with talent or even in some cases achievement. 5 Bolen, 8 Griffith, 9 Kerk, 10 Wittlake, 21 Woods, 26 Traxler, 32 Alirez (before he withdrew) ... just to name a few.
 
On bonus points, Iowa scored 17 at Big Tens ... zero came from 133, 149, 157, 165 and 184 ... PSU scored 16.5.

We've seen the Hawks wrestle tight before, you never know.
When you wrestle tight or play not to lose in any sport...it seems the likelihood of losing increases. Here is hoping for a tight Hawkeye team. Much different being the hunted versus the hunter.
 
Nice to see another simulator, we are (not surprisingly) seeing pretty similar results (and I share your skepticism in this year's seeds giving similar performance to prior years).

Back when flowrestling used to write articles using my simulator, I convinced them to use graphs showing the probability distributions of outcomes (which seems like a very intuitive way to show the uncertainty to me), but they never really liked the idea. If I wasn't so lazy, I'd figure out how to insert a figure and show this year's distributions (PSU's likely points are in a pretty narrow range).
2018 NCAA Team Championship: Simulated (flowrestling.org)
 
Nice to see another simulator, we are (not surprisingly) seeing pretty similar results (and I share your skepticism in this year's seeds giving similar performance to prior years).

Back when flowrestling used to write articles using my simulator, I convinced them to use graphs showing the probability distributions of outcomes (which seems like a very intuitive way to show the uncertainty to me), but they never really liked the idea. If I wasn't so lazy, I'd figure out how to insert a figure and show this year's distributions (PSU's likely points are in a pretty narrow range).
2018 NCAA Team Championship: Simulated (flowrestling.org)

throw it into a googlesheet and share the url, we would love to see it!
 
  • Like
Reactions: hlstone
Lions had a great first session! I love that Joe Lee went for it, keep on wrestling Joe!

Penn State's probability increased to 4.6%, that is a big jump. Still small, but could have gone the other direction. Need a similar session tonight........
 
  • Like
Reactions: PSUer1989
Lions had a great first session! I love that Joe Lee went for it, keep on wrestling Joe!

Penn State's probability increased to 4.6%, that is a big jump. Still small, but could have gone the other direction. Need a similar session tonight........
I like the idea of going 10 - 0 tonight (Joe will wrestle twice, right?)
 
Everyone wrestles once.

Those that lose in the championship bracket (hopefully none from PSU!), and Joe when he wins tonight, will wrestle twice tomorrow, Session 3.
thanks Roar... to lazy to check if they were going to go two consi's tonight ... so I like 9-0!!!
 
After Round1:

Iowa sits at 96% chance to win (134 pts) and PSU is 4% (105 pts). PSU closed part of the initial 42 point gap but has more work to do. The gap has closed because PSU has wrestled up rather than the Hawks wrestling down. To have a chance PSU needs to continue to wrestle up but need a few slip ups by the Hawks.

PSU's chance for second is now 88% (up from 52% initially). NCSt is the favorite for third at 47%, but that also means the field is a slight favorite over them.

Beard - Just go out and win 197. You got this, you can beat anyone in this bracket.
Berge - Great chance to AA, top 3 really tough at this weight though.
Howard - Path to AA is there, going to need to beat one good guy.
Kerk - Dang he looked good against Wood. How about going full old-lady-lifting-a-car and access that part of your brain that we don't know how to get to and pin Gable?

Need no PSU top guy upsets and somebody (or two) from Iowa to get upset. Great day of wrestling ahead!
 
After Quarterfinals:

Iowa is now at 88.6% chance to win (127 pts) and PSU is 11.4% (111 pts). Hawks projected total went down with Marinelli losing (primarily). These numbers do not take into account Berge's injury. He looked seriously hurt to me but didn't want to remove him until it is official. Perhaps if they put the headgear on Vincenzo nobody would notice? Trying to think outside the box.

The difference between the median and 95th percentile score for both Penn State and Iowa is about 16 points. As we run out of matches the variance decreases.

PSU still at 88% for second. NCSt is now 57% for third but a whole bunch of teams can still get third or fourth. Penn State needs a (6-1) or (5-2) round tonight combined with a (5-4) or (4-5) type round for Iowa to still be in contention going into Saturday. It would be great if PSU were within striking distance Saturday Night and had some PSU vs. Hawk finals.

Regardless anxious to see some great wrestling tonight! Proud of this very young team.
 
ADVERTISEMENT
ADVERTISEMENT