ADVERTISEMENT

Bad Seed

STAND with PRIDE

Well-Known Member
Sep 9, 2018
6,875
10,985
1
How does everyone feel now about how guys were seeded?

Wonder if there is a quick way to find out the percentage of matches won by the higher/lower seed.
 
How does everyone feel now about how guys were seeded?

Wonder if there is a quick way to find out the percentage of matches won by the higher/lower seed.

I'll leave that to someone else. My general sense though is that the seeds were pretty fair

Overall there were only 3 top-4 seeds who failed AA. #2 Eierman, #4 Scott, and #4 Brucki.

The lowest seeds to AA were #21 Hoffman, #15 Clarke, #17 H. Willits, and #14 Bulsak. After that, it was 5-8 seeds sometimes getting replaced by 9-12 seeds.
 
How does everyone feel now about how guys were seeded?

Wonder if there is a quick way to find out the percentage of matches won by the higher/lower seed.
I hear you but would propose a different metric -- final placement vs seed, or maybe expected points vs actual points.

I think I could do that fairly easily in Excel ... after work.

Plus it would even out the individual upsets -- for example, if 25 beat 8 in Round 1 and then lost the next 2, he still went 1-2, close to expectations. If 8 came back to R12, that's not far from his seed either.
 
A bad seed would be someone who is clearly a 6 or 7 seed and because he has 3 or 4 guys at his weight in his conference that in the top 6, then gets a much lower seed based on those losses that he hits a top 4 guy in 1st or second round. Not a perfect matrix they use. Should be more weight on quality of opponents.
 
A bad seed would be someone who is clearly a 6 or 7 seed and because he has 3 or 4 guys at his weight in his conference that in the top 6, then gets a much lower seed based on those losses that he hits a top 4 guy in 1st or second round. Not a perfect matrix they use. Should be more weight on quality of opponents.
Or it's another thread about Downey.
 
I know that 59 top 8 seeds AA’d. That’s 73.4%. I know because someone picked “the top 8 seeds at every weight” in my AA pick’em contest. Glad that entry didn’t end up winning. :)

I’m guessing @RoarLions1 might be able to offer where 59 / 73.4% stands historically.
 
I hear you but would propose a different metric -- final placement vs seed, or maybe expected points vs actual points.

I think I could do that fairly easily in Excel ... after work.

Plus it would even out the individual upsets -- for example, if 25 beat 8 in Round 1 and then lost the next 2, he still went 1-2, close to expectations. If 8 came back to R12, that's not far from his seed either.
@STAND with PRIDE Ran the data thru Excel, here's what I came up with:

Seed RangeAvg Delta
1
0.4​
2
0.4​
3
-1.1​
4
-1.4​
5
-0.6​
6
0.2​
7
-0.2​
8
1.4​
9-12
3.5​
13-16
1.1​
17-24
0.9​
25-33
0.4​
Overall Avg
0.9​


- Delta = Actual Points - Expected Points for that seed.
- Actual Points include bonus. Expected Points do not.
- Expected Results do include pigtails: 1 pt for 32, and 0.5 pts for 30.
- Table above shows Seeds by ranges, since those clusters have the same Expected Results. I have the data by individual seed, but that table is way too long.
 
  • Like
Reactions: mcpat
The 2 individual seeds with the widest Delta:
#10 = +7.9
#11 = +3.4

This makes some sense -- they're in the range next up to the podium, where the real points are.

However, a rarity: of these 20 wrestlers only 1 underperformed -- and he only slipped by -0.5 pts. Meaning that 19 of those 20 guys either made the podium, met expectations by reaching R12, or made up the difference with bonus.

These were also the seeds of several of the top overperformers: Lovett, Andonian, Robb, and Bastida.
 
Last edited:
Biggest individual underperformances to seed:
- 141 Eierman, Iowa: 2 --> R16 = -14.5
- 197 Brucki, Michigan: 4 --> R16 = -11.0
- 149 Wilson, NC State: 2 --> 7 = -8.5
- 157 Scott, NC State: 4 --> R12 = -8.5

Top 10 biggest individual overachievers vs seed:
- 149 Lovett, Nebraska: 10 --> 2 = +16.5
- 141 Clarke, UNC: 15 --> 2 = +15.0
- 157 Robb, Nebraska: 10 --> 4 = + 15.0
- 149 Andonian, VT: 11 --> 3 = +14.0
- 197 Bastida, Iowa St: 10 --> 5 = +10.5
- 141 Willits, Oregon St: 8 --> 4 = +9.5
- 141 Bergeland, Goofers: 10 --> 7 = +9.0
- 197 Hoffman, Ohio St: 21 --> 6 = +8.5
- 157 Willits, Oregon St: 17 --> 7 = +8.0
- 197 Warner, Iowa: 6 --> 2 = +8.0
 
Our guys:
- 125 Hildebrandt = -0.5
- 133 RBY = +3.0
- 141 Lee = +3.0
- 149 Bartlett = 0.0
- 157 Berge = +1.0
- 174 Starocci = +2.0
- 184 Brooks = +7.0
- 197 Dean = +1.5
- 285 Kerk = +3.0

Good recipe for success -- get high seeds and overperform them. Cael might be onto something there.
 
Our guys:
- 125 Hildebrandt = -0.5
- 133 RBY = +3.0
- 141 Lee = +3.0
- 149 Bartlett = 0.0
- 157 Berge = +1.0
- 174 Starocci = +2.0
- 184 Brooks = +7.0
- 197 Dean = +1.5
- 285 Kerk = +3.0

Good recipe for success -- get high seeds and overperform them. Cael might be onto something there.
You should let Pyles know.
 
How does everyone feel now about how guys were seeded?

Wonder if there is a quick way to find out the percentage of matches won by the higher/lower seed.

0ad80c6be0d4871c05d035e8b8bf21e7.gif


(This is what popped up when I searched "bad seed")
 
640 bouts wrestled, half (320) in the championship bracket, half in consolations.

-- 146 upsets (based on seed)
-- That is 23% of the bouts (low by historical standards - by memory only)
-- 63 (20%) of championship bracket bouts were upsets
-- 83 (26%) of consolation bouts were upsets
-- 149 was the highest (17 upsets)
-- 174 was the lowest (10 upsets)
-- Only one upset in the championship finals (184)
-- 3rd round of championships, 2nd/3rd/6th round of consis all were > 30% upsets
-- 174 had the highest number of first round upsets (6, along with 125), but there were no more upsets in the championship bracket
-- Only two weight classes had the top-4 seeds in the semis, 125 & 174
 
Last edited:
  • Like
Reactions: hlstone and mcpat
ADVERTISEMENT
ADVERTISEMENT