[FiM] Week 2 Recap and Week 3 Analysis
Week 3 is here and you know what that means? The regular competition season is already half over after this weekend! Crazy huh? This week is interesting because we have a clash of still-haven’t-played-yet teams as well as those that are going into what could be their final event of the season. Teams that have their machines well oiled, bugs fixed, autons tuned in, and drivers practiced vs. teams who are just coming out of the bag for the first time. Obviously those teams who are competing for the second time this weekend should have a leg up on those that are just starting out. Lets hope those teams that are playing for the first time know how to figure it out fast because this is the week where we’re really going to see things kick up a notch. As usual, our analysis and discussion is focused on on-field performance and data driven metrics.
Week 2 Recap!
So let’s get going then eh? Let’s start by doing a recap what we saw from Week 2, because there was definitely a lot to see and a lot of impressive robots. Kalamazoo’s Stryke Force team 2767, making their debut didn’t disappoint, utilizing swerve to their advantage again this year with a quick scale bot and reliable buddy ramp climb. Goodrich’s Martians team 494 made a very impressive showing of their own with one of the quicker scale bots we’ve seen so far this season out of Michigan. Sterling Heights’ ThunderChickens team 217 follows very closely in their footsteps with a very similar robot (both had green elevators, duh) and Flint’s Metal Muscle team 1506 was another quick scale robot who joined the 2 cube auto club. Another speedy scale bot was Berrien Springs’ GreengineerZ team 3452 who showed off a successful far scale auto and a quick climb and continued their blue banner streak after getting to Einstein last year. Although these pure scale robots did not having a partner climb, they were quick and effective enough to get everyone’s attention.
Closest match of the week goes to the teams who played at Waterford district’s match #40, where the blue alliance consisting of Pontiac’s Juggernauts team 1, Rochester’s The FEDS team 201, and Monroe’s TEMPEST team 240 just squeaked by against the red alliance consisting of Burton’s Engineers on Fire team 5517, Dryden’s Flock of Nerds team 5662, and Detroit’s Infinity team 7168, winning 285 to 282.
Over the past two weeks, we have been stressing our belief that robots with the ability to climb with a partner consistently for the additional rank point would have an edge over pure scale robots. Now that two weeks have gone by, we wanted to collect more data to see if we could come to any sort of conclusion. Face the Boss vs Pure Scale, who wins? This week, we are going to use Rest of World data, due to the fact that there aren’t many consistent partner climb robots in Michigan alone, although we are expecting this to improve as the weeks go by. When only looking at Michigan we felt the data was skewed and made it look inaccurate due to teams listed solely because of foul induced climb RPs.
Pure Scale - One piece of data we want to start looking at as a metric for ranking scale robots, in an attempt to sort through the overly crowded field, is to look at “scale ownership time”. This stat shows the robots from the past two weeks who had highest average ownership of the scale for their matches in the world, measured in seconds (s). While this metric is obviously subjective and dependent on how many good scale robots there are at the event, it shows where the outliers are, and based on all of the match videos, gives a good indication on some of the top scale bots from the week.
World Team Rankings sorted by Avg. Scale Ownership Time
Team # Climb RP% Scale Ownership Time (s) Rank Result 254 78% 136 1 Gold 2481 0% 129 1 Gold **4003** 0% 126 1 Gold 610 0% 122 1 Gold 59 25% 119 1 Silver 5453 0% 117 2 Gold 148 78% 117 1 Gold 379 44% 116 1 Semi 225 0% 114 1 Silver 1730 0% 114 1 Silver 1538 0% 113 7 Gold 4910 8% 111 1 Gold 829 0% 111 2 Silver **1918** 25% 111 1 Gold 3310 20% 110 1 Gold
Face the Boss - This data was available to us last week, and we have since updated it in the chart below to include both weeks 1 and 2 and Rest of World data.
World Team Rankings sorted by Avg. Climb RP
Team # Climb RP% Scale Ownership Time (s) Rank Result 254 78% 136 1 Gold 148 78% 117 1 Gold 2839 75% 71 2 Quarters 5172 73% 73 1 Gold **2767** 67% 98 1 Gold 179 63% 100 2 Gold 118 56% 93 2 Gold 1678 56% 103 1 Gold 1619 55% 97 1 Gold 272 50% 39 11 Quarters 1425 50% 99 1 Gold 359 44% 91 2 Gold 379 44% 116 1 Semi 3489 44% 50 33 Quarters **67** 42% 58 5 Silver
Our initial review of the results we believe show a couple key items. First, while all the teams on the lists above did very well, it seemed the scale robots finished better than the climb robots. This makes sense, as climbing is less valuable in elims as it is in qualification matches. It continues with the theme that controlling the scale is above all the most important game objective. While some top climb robots are also great scale robots, the initial results seem to point to top tier scale robots having more consistency at ending up Rank 1 and making it to the Finals. However, as the weeks go on and we get closer to MSC this may change as more teams make their climbing mechanisms more consistent. Climbing this year seems similar to the 4 rotor RP from 2017. We did not see it very often in the early events, but towards the end of competition season it was happening quite a bit. Moving forward, we believe teams who are both very consistent partner climb robots AND pretty good at scoring on the scale will consistently take the top ranking spots and will be favorites to win their events, while pure scale robots might drop down a tiny bit at the district level. Only time and data will tell.
Weekly Updated Rankings
Once again we have done some reshuffling of the weekly rankings. As we get a better feel of this game, see new teams play, and identify what’s the most important metrics in this year’s game leads us to continuously move teams around. We anxiously await for the final 3 teams on the original preseason list to play this week so that we can get a more accurate idea of where everyone falls.
- 2767 (↑1)
- 1918 (↑7)
- 67 (↓2)
- 910 (↓1, 4 Wins, 2 Finalist, 10 Points)
- 33 (↓1)
- 3357 (Added to List)
- 494 (Added to List)
- 5505 (↓2, 3 Wins, 2 Finalist, 8 Points)
4391 (↓2, 3 Wins, 1 Finalist 7 Points)
11.1506 (Added to List)
- 217 (Added to List)
- 3452 (Added to list)
- 5114 (↓2)
- 3538 (↓1)
Week 3 Analysis:
A number of high profile and already proven michigan teams are playing their second event, so expect the intensity of this week to increase compared to the first two weeks. This is the first look we get to see at robots who have gone through the debug process and now should start to be competing at the highest levels. If we don’t see a 2 or even 3 cube auto out of some of these michigan teams playing this week I would be very surprised.
Traveling to tangle with some of the most northern of the michigan teams, Rochester’s Adambots team 245 who is coming off their most successful start to a season since 2013 will be looking to do it again with a fast scaling capability, scale auto and quick single hang. A total of only 5 robots at this event will be playing their second michigan event. Two of our last 3 high potential teams based on our preseason rankings, Gladstone’s BraveBots team 4391 and Alpena’s V2 Robotics team 5505 who each brought home two blue banners last year also makes their debut, and we are curious to see if they can continue their run of high performance. Finally, Grand Rapids’ D Cubed team 904 will be debuting at this event who surprised us last year, bringing home 2 blue banners since 2013. Lets see if they can keep their momentum going. Our favorite heading into this event is 245 as we have seen their robot perform very well, dominating the scale during qualifications, but we’ve had lots of nice surprises so far this season and I wouldn’t put it out of the question to see another very effective robot to come out of this event.
Well, this should be a moderately familiar event for some, considering 11 teams competing here competed together at Traverse City in week 1. A total of 16 teams here will be playing their second Michigan event, matching the Lincoln event for most second event teams. Only three of those 16 made it to the finals of their first event, Canton’s Lightning Robotics team 862 at Gibraltar, Traverse City’s Raptors team 1711 at their own district event in Traverse City, and last but definitely not least, Fenton’s Titanium Tigers team 5114 at Kettering#1. None of these teams were able to come home with the gold, but look for one of them to be pushing to capture it this weekend. This should be a great competition if all of these second-event teams have gotten the bugs worked out and have some new auto modes ready to go. 5114 is our favorite going into this event, as they had a good scale scorer and a promising partner climb that hopefully helps them get to rank 1. Shout out to 862 for having a fast scale auto, cube scorer, and quick drivetrain.
Two teams who have already won blue banners will compete at this event to try and bring home the double gold between both of their michigan events. Ferndale’s IMPI Robotics team 1025 won at Center Line as the first overall pick with an effective scale scorer and our personal favorite Auburn Hills RoboJackets team 3538, who was the #1 seed alliance captain at the high powered Southfield event with a quick and consistent switch auto and super effective scale and switch scorer. Two Other contenders that teams need to watch out for at this event are Mattawan’s WiredCats team 5675, who won two blue banners last year, and East China’s Flurb team 5843 who was rank 4 at Southfield and also had an effective scale scorer and scale auto. This event should prove to have a high level of competition, with 13 teams who are competing in their second event of the season.
This event hosts a handful of quality teams who typically field quality robots, two of which who are debuting will be Livonia’s Warriors team 2832, and Carleton’s Cow Town Robotics team 5050. Temperance’s Bedford Express team 1023 is coming in off of their win in Gibraltar, and looking to add more hardware to their showcase with their smart drive team and effective scale / single climb robot. Also coming from the Gibraltar district is Brownstown’s Goon Squad team 3604, who ranked third at the event with an effective scale scorer and a history of strong robots. Debuting with extremely high potential is team 910 who looks to make some noise of their own after an impressive 4 wins in the last two years. Unfortunately, they were only 15 points away from from making it to Einstein last year, but hopefully that pushed them to create an even tougher robot this year. They have a long history of strong performing robots and we definitely expect them to impress the crowds with another one. The competition will be intense for all of the teams here this weekend, and we would put this as our #2 event of the week. Our favorite for this event goes to 1023 based on their performance in week 1 and their competitive history, but we are definitely excited to see what these other teams are bringing to the table (especially 910).
Undoubtedly the Michigan event of the week to watch. Milford has potential to be one of the most competitive events this Michigan season outside of MSC, with Auburn Hills’s Killer Bees 33 who won already at the Southfield event with an impressive scale scorer, Highlands HOT team 67 who is Michigan’s #2 partner climb robot, Linden’s HAZMATs team 2145 who was the #1 seed at the Kettering #1 event, Hartlands Electro Eagles team 3536 who won at Gibraltar with an impressive 2 cube auto, Brighton’s GEMS team 4362 who took a blue banner home to michigan from the Miami Valley regional with their fast scale scorer and auto, Clarkston’s RUSH team 27 who showed strong scale and auto capability as well as (hopefully a fixed) buddy hang in New York, Rochester’s FEDS team 201 who was rank 6 at Waterford, and Pontiac’s Wings of Fire team 51 who had some difficulty but still a strong showing with scale and switch autos at Traverse City. I would expect at least one, if not more of these teams to show off a consistent multi-cube scale auto during this event because that might be what’s needed to put them over the top. All of these top teams are known for their continuous improvements over the season, so expect these teams to be much more competitive than at events they’ve competed in so far this season. Also debuting is Lapeer’s Strike Zone team 5460 and Birmingham’s Automation Nation team 2960. Don’t count out team 503 who will probably show significant improvement from their competition in Indiana as well. Surprisingly, mixed in with this high powered crowd are 10 rookies making their first ever appearance, more than any other event so far this season. It will be interesting to see what these rookies are made of and how the veterans help them along, if needed, in order to secure that auto rank point.
Note to the Readers:
We are always looking for ways to improve this weekly review, so please, give us feedback in the forum thread. Let us know your thoughts on what data you would like to see next week, or any questions you have hanging out there that you would like us to touch on. We see all comments as an opportunity for us to improve, so feel free to let us know that we are a waste of internet space, as long as you give us some suggestions on how to improve as well.