By the Numbers Preseason Projections

[CLC FAN]

Well, the season is only 17 days away, so it's high time to get some scouting done.  Tom will have this posted on the website soon, but I thought I'd share here as well.  So here's my little write up about the projections with links to what you need.

https://docs.google.com/document/d/e/2PACX-1vTZTFPd3UujXBf9YRnrIj0qhSj6xK_sCT1i19li3vWAWuYrgVU0Gpab9azUuTQU5VWTtSkjbKbGgP5L/pub


If you want to skip my pontifications, you can just go right to

Projections for 2018
https://docs.google.com/spreadsheets/d/e/2PACX-1vQLXGajdQb_ZPCkh8fV0ipL1i0kEPB5SmVjMJUhZoa7Dfuw371iDspS8UVZkiZTIb5-Qa1EgP22XkIR/pubhtml

Scouting Reports for All Wisconsin Teams
https://docs.google.com/document/d/e/2PACX-1vRUG3eaHu0OZ45ZcFQYDM39xwO5YaxxHp1vYJS9MJKtJga2jPZcvI7ACvM-socH3W5vQLDt4NpVIbMU/pub

2018 NOPIN Dual Meet Projections
https://docs.google.com/spreadsheets/d/1RnXKa3CSOFACDtly6_WJdtWaLnLsCFuv6U9SG-E5uRA/edit?usp=sharing

2018 Odds for Top Contenders
https://docs.google.com/spreadsheets/d/e/2PACX-1vTPIGP5csVD_EitHHfKJ8N1fw_eGLxM1D3g-VpxmsARd4S_LwlO7v2U6_rTkMG8z4ZaypKQsuuMk8P3/pubhtml

[FinalWord]

I can't open any of the documents? Message says " may have been deleted"

[Troy Grindle]

I always like reading the projections but sometimes I just scratch my head and cant figure out for the life of me how a computer inputting data can come up with what it does.  I know that it takes into account teams history and filling lineups but a wrestling fan can look at teams graduating the large majority of their lineup and other teams graduating very few and realize who is most likely a better team.  I know this is all for fun but sometimes the computer seems way off.  Good luck to all the teams this year it will be a fun ride as always!

[bolsen]

I love this information. Thanks for compiling it. We were one of the unlikely to win regional teams last year. It is entertaining watching upsets play out. Plus it raises the interest level. Keep it up.

[CLC FAN]

I always like reading the projections but sometimes I just scratch my head and cant figure out for the life of me how a computer inputting data can come up with what it does.  I know that it takes into account teams history and filling lineups but a wrestling fan can look at teams graduating the large majority of their lineup and other teams graduating very few and realize who is most likely a better team.  I know this is all for fun but sometimes the computer seems way off.  Good luck to all the teams this year it will be a fun ride as always!

Troy - I like to respond to head scratching moments like this not because I feel the need to defend but because I think it can bring out good points or ways to improve the model if people understand the logic behind these numbers.  So this is not necessarily a way to agree with the projection simulations, but it is maybe a way to understand where they are coming from.
When you click into the scouting report for a specific regional, take a look at the PPSR box for each of the two teams you are talking about.  PPSR stands for Points Per Starter Returning.  The simulation bases a team's estimated regional score on two factors - how full is the team expected to be and how well are the team's starters expected to score at regionals.  A casual fan usually only notices the "how many do they bring back" factor.  Even a fan looking for who owns the talent edge can overestimate the point scoring ability of super studs.  Regional points are capped at 26.  Doesn't matter if you pin your way through a weak regional weight and then lose first round of sectionals or if you pin your way through a tough weight and then go on to dominate your way to a state finals - you earn 26 regional points either way.  So PPSR cuts through a lot of that noise.  An example from last year:

Coming into the 2017 season, I gave Germantown a 59% chance of winning their regional.  They had graduated 8 out of 14 and brought back 0 state qualifiers.  Competing against them were teams bringing back 11 out of 14 and 10 out of 13, among others.  Germantown, though, had 8 starters who averaged 14.5 points per starter at regionals the previous year.  The teams with 11 and 10 had PPSRs of 10.7 and 9 respectively.  I bring it up of course because Germantown did win :-), but the point is that PPSR helped see a talent edge that wasn't immediately apparent, and it correctly saw the low number of starters returning as less predictive of struggles filling out a lineup than most people would assume.  In 2017, Germantown scored 201.5 points at regionals with a full lineup of 14 starters.  Neither of the teams with more starters returning had a full lineup.

Again, not sure this will cause you to agree with who is the better team - the system may well just be in error quite obviously.  I hope, though, that the explanation maybe shows where the simulations get the idea of which teams are likely to be stronger at the end of the year. 

[bigG]

Loved it. Thanks for posting this.

[bigoil]

I love this, wonder if you should wait til after fat testing (not for weights but for roster changes and size).

Example, Kaukauna has 22 freshman. Stoughton has only 7-8 freshman but looks to have a transfer from Milton. Obviously the transfer would be weighted more heavily because it comes with returning points.

[bronco]

Like many before said, thanks for doing this...we are getting close guys, good luck and stay healthy everyone!