Match Strategy Evolution

Let’s talk about Match Strategy

It’s the middle of the season, and the game’s strategy is shifting from “shoot all of your match loads over and then push them into the goal” to “matchload 1 or 2 triballs into your robot, push them into your goal, repeat”.

For example, watch the Haunted finals matches:

  • Matchloads were launched
  • Most (if not all) of the match loads were used
  • Teams mostly used their match loads closer to the beginning of matches
  • Matchloads were used in “short spurts”
  • Scores seemed to be mostly over 100 points per team
  • More aggressive; No defense

However, if you look at the more recent Sugar Rush Finals:

  • Matchloads were placed into intakes and then scored
  • Not all match loads were utilized
  • There was a slight trend of increased matchload usage near the end of matches
  • Match loads’ use was constant throughout the match
  • Scores seemed to be lower on average
  • More defensive, a lot more back-and-forth

Essentially, the initial “early season” strategy was to try to outscore the other team, while this newer strategy seems to be “gain an early advantage and quit while you’re ahead.”

I also want to talk about how this alters the value of certain mechanisms. For instance…

Elevation :chart_with_upwards_trend:: If scores are lower, elevation’s 20 points suddenly become a more significant portion of the final score, which might encourage more teams to push towards higher tiers of elevation

Launcher :chart_with_downwards_trend:: The reason for this is apparent. If no one is using a launcher, then the draw that a launcher would have, perhaps to be a potential alliance partner, is diminished. However, I still think that launchers will continue to be useful, primarily for skills, but lift launchers (Like 9364C/2775V/9364A) might lose prevalence as the lift is used solely for matches.

Intake :chart_with_upwards_trend:: In the Sugar Rush Finals, intakes seemed to be very important. Specifically, the intake’s abilities to a) quickly and efficiently intake triballs and b) maintain triballs in the intake (even when going over the barrier) seemed to be very important.

Higher power drives :chart_with_upwards_trend:: There seems to be a lot more back-and-forth, so a higher torque/higher motor count drive might come in handy.

Wedged Wings :chart_with_downwards_trend:: Having wedges on wings was initially very appealing, as it’s a very good defensive strategy against the other team launching their triballs. However, as I’ve detailed in this post, I don’t see launching triballs as a major strategy. I still do think these are important, as wedged wings are one of the major reasons for the decrease in launching and a team can simply launch if they realize you can’t do much about it, but not nearly as important as they once were.

Of course, this is all speculation, and the initial strategy of trying to outscore the other team is probably still going to maintain prevalence throughout the year, but it does offer an interesting perspective for higher-level matches.

Furthermore, I want to talk about the leading causes of this change in match strategy to see if we can utilize this to try to predict further Match Strategy changes:

In my opinion, it’s the idea that as launchers get higher, blockers get higher until the launchers are too high to matchload onto. In addition, I also think that the increase in wedged wings plays a factor.

Last, I just want to talk about how this emerging strategy affects the autonomous period. I think that because of the decrease in matchload utilization, the autonomous period becomes much more vital. In my opinion, Gremlin’s 6-triball auton (or something crazy like that) was a leading factor in them winning Sugar Rush.

Do you guys think that this strategy will become more prevalent as the season continues? I’d like to hear different opinions on this strategy evolution; all opinions are appreciated :slight_smile:

38 Likes

I also noticed this trend across other seasons, as the season would start with offensive robots with many variations, but as the season progresses, the game gets more defensive and the robots slowly start to look very similar (the “meta”).
Updating the meta predictions from the over under rules discussion, I think the meta will be, like you said:

  1. Strong drivetrain (88w PTO?)
  2. A strong “hogger” intake meant to keep triballs from being stolen (like Gremlin stealing their opponent’s triballs during Sugar Rush finals)
  3. Either a fast, but prioritized cata, or direct-to-intake systems for matchloads
  4. Higher hangs (I haven’t seen anyone do partner hang :frowning: )
  5. I saw this from MOA, but there was this one robot with a giant net used to catch matchload triballs and shove them directly under the goal. Maybe this bot will make a comeback?
  6. Flatbots maybe? Due to the sheer amount of double-zoning happening, flatbots can discourage double-zoning, forcing the opponent to be cautious, and descore maybe a few precious triballs
7 Likes

I’ve noticed some of these trends too.
How did the net on the bot at MOA work?
Flatbots?

1 Like

Bots that are able to go underneath the goal completely to de-score (also called pancake bots) (Robokauz early season bot)

Imagine, the match begins, and this robot from the top flips out a giant square-shaped net funnel as a vertical/horizontal expansion. Their partner would fling the triballs and the net bot would go to a location and “catch” the triballs. It was extremely top-heavy and looked very clumsy, but was a very creative idea.

2 Likes

We’re a flat bot and it absolutely destroys the meta. All defensive play goes out the window.

3 Likes

I watched the elims at Sugar Rush and noticed the same management of Triballs. There was very, very careful management of the triballs being introduced. Retention was key, there were some impressive body slams to dislodge the triball and then grab/score.

Winning Auto was a factor in most of the matches. Gremlin’s multi-triball auton is a thing of beauty, it’s worth watching their matches from the event.

I agree to on the hang. With a one or two tri-ball lead, the hang has got to happen.

@9080Sympathizer - any match or reveal videos we can watch? I can see where the flat bot would compel teams to change strategy.

Sugar Rush was very impressive. For early December all the robots were very strong contenders. The top 4 robots were amazing.

Good post @Money and your analysis was great to read.

12 Likes

Alright, there was a slight oversight on my part. I do think that robots will push towards higher tiers of elevation, but more so I think they’ll go for more consistency on their hangs as well. For example, an F tier hang that works 50% of the time is less valuable than a B-tier hang that works 80% of the time imo. My reasoning behind this is that the gap between 15 points and 20 points isn’t that substantial. Let’s say that the F-tier hang will get all 20 points when it works, while the B tier hang will get 15 points. Over 10 matches the B tier hang would get a total of 120 points, while the F-tier hang poll get 100 points. Even that is a bit of an exaggeration as most teams don’t have higher than b-tier hangs.

1 Like

Finally… a thread that’s worth camping with my popcorn… :slight_smile:

Mostly agree with your anaylysis and predictions.

but i don’t quite agree with the effectiveness of “flatbot” though (normally we will just call it low-profile robot). There are easy ways to counter flatbot.

and interestingly, just yesterday I was discussing with one of my students and I reminded him of the importance of defensive plays and asked him to analyse how 8059A and 169A won in 2019 - gained a headstart and then defend like your life depends on it.
But this approach requires a lot of control (of the game), e.g. the ability to take back the lead whenever the opponents managed to score 1 or 2 balls. So in a way, retention is one way of having control of the game.

I still think lanucher will still be important. But it is about how and when you use it. And also how should you position your alliance when you are launching.
maybe I shouldn’t reveal too much of my thoughts at this point (Singnat is not over yet :stuck_out_tongue: ).

But please keep the meta-gaming going… :slight_smile:

18 Likes

From my experience, a lot of robots do have shooters, but not a lot have blockers. My team made a blocker in a couple of hours and it uses no motors and no pneumatics. We use rubber bands that are connected from C-channels on either side of the blocker to inside the gears of the drivetrain. This allows the blocker to start out folded up and then at the beginning of the match it pops up and was able to block every shooter we have gone up against since adding it. This mechanism is the reason we went 6 and 1 at our last comp.
I recommend this for everybody.

Unfortunately we haven’t dove into the realm of reveals or media, so I can’t exactly show you the bot other than a picture. However, I’ll be glad to give some insight on strategy

Since flatbots make it risky to double zone, alliance partners and I double zone shoot, splitting them 10-12 or so. We start on the far side, and transition to the left of the field to shoot while our teammate takes the close side of the field and immediately starts. What usually happens is when the match starts, they shoot all of their triballs immediately and start scoring whilst we shoot and defend. If the opponents double zone to defend or attack, our alliance de-scores. This forces opponents to adapt and it makes it difficult to outscore us. It hasn’t failed us against metabots and we have other strategies for individual cases such as Qualifying Matches too, but it’s a bit complicated case by case.
I’ll give a small picture of the bot below.

1 Like

Oh boy. Before I start my post, please note that these are my opinions based on my observations. People that have watched different events than me may have different opinions. (I haven’t watched sugar rush. Just moa, haunted, speedway, and many smaller local Quals.)

First of all, I will analyze some aspects of robots we have seen this far:

  1. Lifted shooter. This was one of the main strategies of recent. Teams lift up their shooter and shoot match loads effortlessly over opponent blockers. My personal opinion is that this will die out before worlds. A smart opponent could just put out their wings and push all the thrown triballs right back over the long barrier.

  2. Giant blocker. These are pretty sweet. The ability to lock up your opponents match loading is a game changer. If your opponents cannot get triballs to their side of the field, ain’t no way they gonna win.

  3. Descoring abilities (flatbot). Sigh. I wish I could say this is the new meta, but it’s not. The goals are difficult to descore from, the rules on descoring are airtight, and teams aren’t stupid. If your robot can fit in a 5 inch tall bot, might as well, but in any other cases this is not optimal.

  4. Wings. Wings are awesome. They are a necessity for any team at the top level. You can control triballs in the masses and amp up your skills score. We will see these at worlds.

  5. High elevation. I really don’t know about this one. A vertical pole is difficult to climb on with a 10 pound bot. I assume teams will figure it out, but gosh it’s tough.

Now I will look at some strategies:

  1. Throw all your match loads and hope for the best. This strategy is already beginning to dwindle. Teams are becoming more decisive on when to use their thrower and when to just hold the triballs.

  2. Double on the defense. Both robots play on the defensive side until late game. They minimize opponent scoring and save scoring for the late game. I think this strategy is the way to go. Especially if you have good wedged wings.

With all this in mind, here are my predictions for worlds; Teams will have 6 motor drives pto’d to their elevation mech. Yes, they will get to the top of the pole. Teams will have intakes. On top of their high elevation mech, they will have a kicker that can match load from up high. Teams will use their elevations as blockers. Teams will have wings. However, the winner of worlds will be the team that can lift their alliance partner up to a j-level hang. Or maybe not. Who knows…

5 Likes

While, yes, flatbots can discourage double-zoning, I agree with @meng in that they’re just too easy to counter. If a team notices that their opponents have a flatbot, they can say: “Oh, they have a flatbot. Let’s just not double zone” end of story. And a lack of double zoning doesn’t seem to correlate with more losses, as is apparent from Sugar Rush Finals with 9364H, Hailstorm, being an extremely competent flatbot. And I do think that Hailstorm’s success is correlated with them being a flatbot, but solely because there’s not much space for other mechanisms, so they had an 8-motor 600rpm drive. It is not because of the threat of double zoning. Just an opinion, but I’m willing to listen :slight_smile:

3 Likes

Sounds about right. Just that we are expecting all these to happen during SingNat in early Feb and not in April :sweat_smile:

Actually, i would lure the flatbot to come underneath my goal and then un-double-zone and trap it underneath for the rest of the match. It will become 2v1.

6 Likes

I would recommend watching these videos: Sugar Rush Finals, Sugar Rush Highlights, and for the strategy that I’m talking about, Gremlin’s Recap.

Ima try to rate ur takes in my opinion:

6/10; I agree with the fact that the lifted shooter isn’t THE BEST design, but it’s useful against some types of teams. In addition, I just feel like the lift serves too many purposes to just die out (I mean, it’s elevation, a blocker, and a lifted shooter mech all in one. But still a good take.

7/10; Review those videos I talked about earlier and see if you still think this take applies. Nobody launched as you said, so they are useful but certainly situational. If a team does do a blocker, I think it should double as an elevation like 21417A’s Mid-season robot Sparky. Lifted shooters can also act as blockers. The reason I rated this take a 7/10 is because I think that the reason people match-loaded was a result of blockers and wedged wings, so if the other alliance doesn’t have either one of those things I think it’s safe to launch.

10/10; I think you already know how I feel about flatbots. While they certainly can be effective (9364H), I don’t think it’s because of their descoring abilities, but rather their 8m 600 direct drive.

9/10; This is a great take, but I did decide to dock you a point because I do think that the old meta was “shoot all of the triballs over to the other side and then use wings to push the triballs into the goal”, but now it’s more of a “Use wedged wings to get some triballs over the barrier.”

8/10; I’d put more emphasis on the consistent part (look at my last post on this thread), but in that take, I do assume that a B-tier hang will get 15 points every match. As the season progresses and teams push towards higher hangs, a B-tier hang won’t get very many points.

Now for the strategical analsysis

I was thinking more of the “gain an early lead (have a tuff auton like Gremlin) and quit while you’re ahead” Basically score the triballs on the field, maybe matchload a couple into your intake, and push them straight in the goal.

Alright, first of all, the 6 motor drive PTO’d to their elevation mech seems good in theory but I haven’t seen many truly competitive robots at sigs that utilize PTOs for the sake of simplicity. I could get behind having a 66W elevation though, and this was one of the major designs I was considering for my team’s rebuild. I think that the team who wins worlds is going to have an E-F-tier pole hang and then a C-tier horizontal hang. I think that a team that can partner hang a J-tier alliance will be able to do nothing more than that. Yes, kickers on elevation mechs, utilizing elevations as blockers, and wings all seem to be good strategies.

Also, @meng your strategy seems to be very risky. It’s certainly feasible, but I feel like you would need to have perfect coordination with your alliance partner (one robot pushes in the triballs to lock the flatbot in, and the other un-double zones. Also, flatbots seem to have very fast and powerful drives so they might be able to descore a lot of triballs before you’re even able to un-double zone.

5 Likes

pretty sure my bot got itself stuck under a net by itself, but technically this could limit the number of triballs you fit under the goal, which could still allow the outnumbered bot to be able to win (likely only going to happen if that driver is really good)

1 Like

Like what I always tell my teams - if you can’t win in a 2v1 situation, then you really don’t deserve to win :sweat_smile:

2 Likes

It planning to lure the flatbot has to be worked out with the alliance before the match and it has to be intentional.

Honestly, it is not difficult to see the flatbot driving towards your goal and start to get ready to trigger the un-double-zoning.
And also, I only intend to lure the flatbot in only when the goal is at least half filled. This will make it more difficult for the flatbot to manoeuvre underneath and also more difficult for the flatbot to push the balls out.

But caveat - this is just how I am visualising it. Will need to get my teams try it out during their practice matches.

observations

  1. tunneling is low efficiency as of now. this is why match scores are more than half what it was early season. this will not remain the case forever, as teams get better at bowling, scores will increase (china as a case study)

  2. in any blocker vs shooter arms race, blockers will ALWAYS win the arms race. the reason is twofold
    a. blockers are so much easier to make than raised shooter
    b. shooters have a theoretical ceiling (which is how high matchloaders can
    reach). blockers have no such limit.
    this isn’t even considering the new gdc rulings

  3. elevation remains pointless. every team that knows what they’re doing has either a barrier hang or at least a passive a tier hang. once your opponets have an a tier hang, your hang doesn’t remotely matter anymore, as “close” as matches are getting. what we tell ourselves is, “if we’re winning by less than 5 points, we don’t deserve the win anyway”

  4. wings are completely unaffected by the meta shifts, as wings are primarily a tool for skills anyway.

meta predictions :

  1. flatbots will never be the meta. in the matches i’ve played, whenever i’m against a flatbot, i almost hope they attempt to descore. unless the rules with double zoning changes, i don’t see how on earth flatbots will EVER be meta. triballs are too difficult to push out the goal under 1-2 seconds, and the rules with double zoning is way too disadvantageous for the double zoning team.

  2. shooters will never be meta again. two-stage blockers have been more common than snail bots in CU since haunted, and blockers will ALWAYS win the arms race. shooters are further nerfed by the more recent GDC changes. I don’t see them reverting this anytime soon, cox said he liked the current meta a lot.

  3. match scores will get higher as teams get better at bowling. china has been bowling since summer, and their match scores are all still 160+. as we get better at bowling, scores will increase. i think at least until kala, matches in mainstream regions will look like sugar rush finals (i <3 gremlin).

  4. shooters will still exist, as a necessity for skills. there’s no way any t300 teams are going to take off their shooters just for tournament (even though china’s doing it). teams are going to keep shooters for skills, so we CANNOT forget how to defend shooters, or a shooting team will win worlds

  5. effective barrier crossing is very, very critical. IFT and 9123 drives as if the barrier isn’t even there (as in crossing is very easy, not as in not crossing). this is why their match scores are so high. being able to barrier cross and intercept your opponents bowling wins you matches.

  6. autons will win matches (hence why we had a 6ball rush way back in august XD). the 16 point swing, as well as scoring all the triballs on the field already, will be what determines the match flow in driver control. i suspect the late season meta will be a lot like TiP, with auton rushes, then very defensive gameplay.

18 Likes

Sorry… what is bowling in Over Under…?