MinnesotaMan's Week 1 Predictions

MinnesotaMan · February 28, 2017, 11:58am

Introduction:
Coming to you from the land of 10,000 robots and no districts in sight is Minnesota Man! I will represent
the great teams of Minnesota by regional and championship predictions and more. Also, as a
bonus, I will say 1 fun minnesota team fact every week.
**
Fun Minnesota Team Fact #1:**
For those of you who don’t know, there will be a Minnesota team playing at every single Week
event. Minnesota is actually everywhere, which isn’t a surprise to me. Little Rock, AR to MN,
Duluth is how far of an extent MN teams will travel this season.
**
Lake Superior Regional History:**
Ever since it’s inception in 2011, the Lake Superior Regional has been a popular event to
attend. From Hall of Fame teams like 359, to attracting teams from the upper midwest area like
537, 706, 1625, 2062, 2530, 2826, 3061, etc. This regional is rich of tradition and holds
memories upon memories under the roof of the DECC. As for the Double DECCer record (an
exhibition match between the Lake Superior Winners and the Northern Lights Winners), Lake
Superior currently trails Northern Lights 1-3.

2017 Lake Superior Insight:
This regional is full of superior teams. Don’t underestimate them because of the lack of
household team names. The perennial powerhouse 2052 has always been a contender since
2011 and have been krawling to the finals of this regional for 5 years in a row. And after
reaching Einstein last year, they are a favorite to win this regional to nobody’s surprise. 4009
also has it in their DNA to snag another blue banner after last season’s previous success of
winning a regional and making it to their division semi-finals at champs. Let’s not forget about
3102 who won the notoriously explosive Wisconsin regional last year either with their simple,
but effective machine. 4539 has been very chaotically competitive (in a good way) since their
rookie season in 2013, coming close to winning their first regional multiple times since then.

Honorable Mentions - 2526, 2977, 3042, 4818

Dark horses that could give teams a run for their totes - 3276, 4480

Chairman’s/EI contenders - 1714, 1816, 2526, 4009
**
Northern Lights Regional History:**
The expanded regional called Northern Lights started in 2013, and has quickly brought many powerhouse teams to their side of the DECC, including Hall Of Fame teams 27 and 359, as well as bringing their own Einstein hungry contenders like 525, 876, 2169, 2175, 2502, 2512, 2883, 3130, and most notably 5172. With so many great teams attending each year, it’s no surprise that they hold the winning record against Lake Superior in the Double DECCer 3-1.

2017 Northern Lights Insight:
Coming all the way from Clarkston, Michigan 27 is coming back to Northern Lights for some pre-district practice as well as rushing to snag a blue banner win, but they have to face some top notch teams to do so. 525 has won at least one regional since 2010 and is going and try to keep that statline by barking up an early week 1 win here. With their interesting three wheel drive last year that got them 3 blue banners and sub-division semi-finalists, it’ll be interesting for what they came up with this year. With their constant praise of scouting and being division finalists last year, 876 is going to try to practice what they preach and show what it takes to thunder their way to North Champs and beyond. With their 3 regional wins in the past two years, 3130’s success is no error. Consistently bringing regional winning bots from their workshop, we hope to see another crowd pleaser. However 5172 is back for its 4th year. This time trying to win Northern Lights after being finalists for the past three years. They finally have experience under their belts after a breakout season last year (winning their first regional, state champions, and division-semi-finalists) and now will be waiting to chomp up the competition this year yet again.

Other powerhouses- 2175, 2502, 2883

Honorable Mentions - **2202, 2823, 2987, 4607 **

Dark horses that could give teams a run for their totes - 1792, 2509, 3883

Chairman’s/EI contenders - 525, 2502, 2512, 4607

**Double DECCer Winners and Finalists (2013 - 2016) **
Winners:
2016 Northern Lights: 2883, 2502, 5232
2015 Northern Lights: 3130, 525, 4215
2014 Northern Lights: 2175, 359, 2502
2013 Lake Superior: 2052, 2062, 4011
Finalists:
2016 Lake Superior: 4009, 359, 6175
2015 Lake Superior: 2526, 4818, 93
2014 Lake Superior: 3018, 2052, 3692
2013 Northern Lights: 2169, 525, 3061

Have a wonderful week 1 Minnesota Teams! - MinnesotaMan

Ryan_Swanson · February 28, 2017, 12:06pm

I’ll just leave this here…

MinnesotaMan · February 28, 2017, 12:11pm

My fellow Minnesota FIRST enthusiast, I appreciate you leaving this here. My post is a different (I hope) way of predicting events for these beautiful MN teams. I took the time and effort into researching teams on my own. Your predictions are very respectable, though, and I wish not to alienate them in any way.

bobbysq · February 28, 2017, 12:12pm

Can’t agree more for 525 as a Chairman’s pick. With the new Shenzhen regional and the Iowa regional last year, 525 gains a lot of credibility for Chairman’s. Being responsible for not just a state’s robotics teams but also a whole country’s can get them a win for Chairman’s here, (edit: or) a repeat at IA, and possibly at the Championship.

Caleb_Sykes · February 28, 2017, 12:35pm

Awesome, I’m glad to see a MN specific prediction account.

For another perspective, here are the current Elo ratings of the Northern Lights and Lake Superior teams.

Northern Lights:

Lake Superior:

Northern Lights is stacked, the 8th highest Elo at Northern Lights is approximately equal to the 3rd highest Elo at Lake Superior.

Jon_Stratis · February 28, 2017, 12:41pm

Not to nitpick, but if they win Chairman’s in Duluth they are not eligible for it in Iowa. From the awards page, “Teams will be restricted to earning this award once at each level of competition. In other words, once a team has earned Chairman’s Award at a Regional or District event, they may not earn it again that season at a later Regional or District Event.”

I certainly believe they are deserving, and will win it at one of their regionals.

A_Mountain_Man · February 28, 2017, 1:00pm

Yay! A Minnesota account! I’ll be sure to follow you throughout the year. It’s exciting to see more people stepping up to cover different regions in the FRC.

Great post, by the way!

ndp · February 28, 2017, 1:04pm

Interesting profile pic…

A_Mountain_Man · February 28, 2017, 1:17pm

Also, I’ll say it here too.

If you are thinking about stepping up to cover PNW, NC, ONT, Israel, or another area, you totally should! It’d be awesome to see more quality analysis in other regions of the FRC. If you want to jump into the predictions business, I will support your efforts. I know how hard it can be to get a prediction account off the ground.

mjc49 · February 28, 2017, 1:30pm

Caleb, I remember reading about Elo last year. Can you provide a refresher what goes into Elo.

sidagarwal · February 28, 2017, 1:54pm

Yes please!!! I would love an analysis on some NC events. I would support 100% and I know many other teams here will also.

bobbysq · February 28, 2017, 2:01pm

I believe it’s from the https://frc.divisions.co/charts representation map, where MN was greatly underrepresented at Worlds due to… certain circumstances. that[sup]may[sup]or may not[sup]be because[sup]they aren’t on districts[/sup][/sup][/sup][/sup]

Oh, and they added the communism symbol.

Caleb_Sykes · February 28, 2017, 2:02pm

Sure, Elo models make predictions for matches based on the difference in Elo between two opposing entities. The greater the difference, the higher the likelihood that the entity with the larger Elo will win. For each FRC match, my Elo model sums the Elos of each robot on both alliances and compares this difference for the match prediction. After the match, the prediction is compared to the results, and every participating team has their Elo adjusted accordingly. My model looks at the winning margin of the match and compares it to the predicted winning margin. My model is primarily concerned with qualification matches, but it also factors in elimination performance to a lesser degree. It only counts official FRC events. There is more detailed information in my 2008-2016 Elo workbook if you want it.

Let me know if you have any more questions.

MinnesotaMan · February 28, 2017, 2:44pm

Thanks for posting the elo stats Caleb.

When I post these insights I hope to create a discussion every time and other users providing their insights or statistics like these are more than welcome here.

Thanks for the support my mountainous brother!

TheMilkman01 · March 1, 2017, 3:57pm

While I appreciate the sentiment of this post, please be considerate and capitalize the state that you so greatly represent. This, along with many other minuscule grammatical blunders, could improve the overall quality of your posts. Of course, I do support such an endeavor and being a Minnesota man myself, I gladly offer my assistance in language the next post you submit to this glorious form.

This is quite a generous description, although I would question your alliterative use of the word “perennial.” Despite five years being lasting, it is hardly an apparently infinite time. Thus, I would humbly suggest you consider substituting it for a more appropriately descriptive word, such as “dependable” or “persistent.”

Do not take this as me badgering you, I simply desire the best for my state, as I’m sure you do as well. With all this said, I wish you the best of luck on your journey ahead in the Minnesota Nicest way possible.

BumblingBuilder · March 1, 2017, 9:58pm

See this thread

Caleb_Sykes · March 2, 2017, 11:09pm

With the schedule released, I can make better predictions, here are my LS Elo ranking predictions.

team, predicted win ranking points

Caleb_Sykes · March 2, 2017, 11:28pm

Also, here are what Elo predicts to be the best (close with strong teams) LS matches to watch:
19, 41, 42, 53, and 67

Caleb_Sykes · March 5, 2017, 12:15am

Elo ended up doing reasonably well, but not spectacular. My start of event Elo predictions explained 18% of the variance in ranking results. Not great, but still decent considering the predictions did not update over the course of the event and that they used exclusively data from pre-2017.

LS graph.PNG