Wrestleknownothing

So who is the coach he called to personally inform that he was choosing PSU where the conversation was bad? We know it wasn't Brands (see you on the mats).

Damn. I wanted the wire.

Now UFC is accusing USADA of defamation and demanding a retraction: https://www.espn.com/mma/story/_/id/38646276/usada-stands-initial-statement-ufc-plans-legal-action Handbags

It is his company. Other/minority investors have no say. They are just along for the ride.

Singlets are for the puritanically modest. Try a speedo.

Tease

Check out the username of who you responded to.

My consulting contract is in the mail. My logo is "I may be slow, but at least I'm expensive."

Amen. Other areas do use Power BI. So much the worse for them.

Oh, and you should read Born A Crime

The funny thing is I often do that for a different reason. I would call them spurious digits. Since they are not measurements, but integer counts used to calculate percentages I think I am OK on significance, but not so much on spuriousness. I occasionally dig up old stuff I did, like this, where, for some reason, I did not over-ride the default formatting in Tableau (BI software I often use instead of a spreadsheet). As the woman who edits the stuff I write for work can attest, I sometimes (often?) miss inconsistencies in formatting and phrasing in my own work. Feel no conflict. We can all use good editors. And since I cannot be responsible for disturbing Mr. Valentine's eternity, I have reformatted. RIP, good man.

You have also been told how stupid many of your takes are, but you do not seem to quote that. Rationalizing apartheid by implying if it wasn't for us white folks your lives would be worse than under apartheid, is the height of stupid.

Nebraska has been very volatile. From the best performing relative to seed in 2022 to the worst in 2023 among top schools Ok State a little less volatile but also coming off a rough year.

Wrestling, basketball, volleyball, judo, boxing, and weightlifting, especially weightlifting, on ice. Problem solved. You are all welcome. Where do I pick up my consulting fee?

Great tune. Love this version of ZZ Top. Tres Hombres is one ass kickin song after another.

If you are asking me if I ever tortured the numbers until they speak.....my answer is no comment.

The problem comes in much higher than the 16 seed. If all you do is look at the seeds where there is at least a 1 in 3 chance of AAing (top 10) you have major problems with fragmentation at the 3 or 4 seed for all teams, and beginning at the 1 seed for any team outside of the top 6 most entrants seeded in the top 10 between 2010 and 2023. Michigan is outside of the top 6 in spite of finishing 2nd last year. Building a Michigan specific curve would be problematic. But even for PSU, how do you build a curve with only three 4 seeds in a 13 year span?

Check out this post: http://johnnythompsonnum1.blogspot.com/2020/03/gone-lost-forgotten-their-best-notre.html

Another way to look at the over/under performance is to recognize that from a team finish perspective the only thing that matters is finishing top 8. To that end if I reduce the set to just those who have a 1 in 3, or better, chance to AA (top 10 seeds), then seed relative performance looks like this: PSU performs right on seed whereas everyone else falls short by 1 to 4 spots on average.

That translates as blah, blah, blah. Feel free to read through my prior opuses on the topic.

Ideally, yes, but where that gets problematic is you wind up over-Balkanizing the data. If a team has a 10 seed and they have never had a 10 seed before, what do you do with that? I am thinking of playing with ranges instead, but stay tuned.

Correct. I have thought about doing something with that, but have been a bit conflicted on it. I will give it some more thought.

This one is a little tougher to read, but here goes. Based on 2010-2023: The first three columns show the percentage of times a team beats it seed, matches its seed, and does worse than its seed The second set of three columns shows what their average seed is when they beat it, match it, or do worse than it. The third set of three columns shows by how much a team beats their seed, matches their seed (always zero), or does worse than their seed. An example of the way to read this would be: Ohio State beat their seed 35% of the time from 2010-2023 and when they beat their seed they were seeded at 12.5 and beat that by 5.8. Said that way, Iowa is pretty impressive in that they beat their seed more often than anyone else (36% of the time) and did it while having a very high starting point (9.3 seed on average - second highest). And they do it by 4.1 spots. And when they underperform it is usually because they have a very high seed (5.6 - highest) and they only underperform by an average amount (5.9 spots).

Here is how each of the expected top 15 teams has performed relative to their seeds since 2010: One thing to keep in mind is that these teams are likely to have higher seeds. This means they have more room to miss by than room to outperform by. So if 50% perform worse than seed, it is likely that they will underperform their expected score rather than match their expected score.