The Fall of the House of Tank Drive

I missed the reply, whoops.

I’m sure it’s well-known but my team (4143) also ran Swerve that year. Only difference was we put a tank tread on the module instead of a wheel.

We even named it Sir Twerve’alot


Greyt products won’t raise the floor until the floor knows they exist. In their current state, they raise the mid tier into the upper mid tier (just look at all the climber kits this year… holy moly). Most teams at the floor only know about AndyMark and maybe maybe Vex.

Swerve is definitely the most expensive and advantageous of off the shelf mechanisms. At least with brushless motors, they don’t wear out and the costs can be recouped in about a year. Heck, REV gave out TWO FREE SPARK MAXES this year! That’s insane value! 6619 reused the 2020 robot Falcons for drivetrain and the two free spark maxes for our Everybot arm.

I don’t see a realistic way to stop swerve drive from becoming meta without making every game have a large step to drive on, which would also make the KOP chassis much worse without pneumatic wheels. Or changing the rules to ban prebuilt mechanisms. I’m not looking forward to telling the teams I mentor to spend $8k on swerve modules and tread.

How do you figure? A CIM is $35. The cheapest motor controller is $50 (Victor) but most teams want Talons ($90). A Spark Max and a NEO is $125, which is pretty close to the same cost overall, and they can be refused if treated decently well. Brushed motors die frequently by comparison.


This but unironically, FRC teams push FRC on rookies way way too much and villainize swapping programs to rightsize. Teams should be encouraged to participate in the program that fits them best, not to struggle in FRC because “its the best program! You’re missing out if you don’t do FRC!”

FRC is a fundamentally bad program for low resources teams, it leaves them feeling like they suck and aren’t deserving of participating when their biggest problem is a lack of tribal knowledge and funding.

We need to stop acting like FRC is god’s gift to the world and start telling people to play in the competition where they’ll be the most sucessful.


Interesting fact: some people actually use the FIRST forum. There are like 5 of them, but they’re there.
Every time I remember that fact it’s a reminder of how screwed over some of the rookies are.


Ain’t gonna win any Chairman’s awards with that attitude!


I’m not sure how to write this without sounding like a crotchety old-timer, but I have some reservations about the developments.

I think the root cause isn’t game design or even COTS implementations. It’s the electronics. Both from the side of the motors and the power distribution system, there is so much more flexibility to blow eight high-powered motors on a drivetrain than there was even four years ago. With the REV PDH, you’ll even have 12 motor slots left over.

The explosion of swerve is coupled to the downfall of shifting. In the past, adding the complexity of swerve often meant sacrificing either pushing power or top speed, or stacking even more complexity to add the functionality of shifting back in. 4 Falcons provide about 45% more power at 40 A than 4 CIMs, so nobody bothers with that anymore.

On both fronts, the outcome is that some teams are finding it easier to “have it all” in terms of functionality. I’m not against raising the floor. I’m not against the fact that robots are getting cooler or more up-to-date with modern technology. I don’t want to be accused of coming at this from that angle. But I wonder—I just wonder—whether the loss of the need to face any real tradeoffs in the motor and construction rules is the right way to go.

The risks I’m seeing are twofold. First, as Kara mentioned, it’s hard for many teams to keep up when the “middle class” standard of play accelerates so quickly. Second, are we creating the best environment possible for students to learn about practical engineering concepts when the calculus always boils down to “just use the most powerful one for everything”? I don’t know. FIRST says right in their description of FRC that it is designed to be contested “[u]nder strict rules and limited time and resources.” That is, to me, the three-legged stool of FRC—each limitation should be balanced in proportion to the other two. Relaxing the bag rules, for instance, made a lot of sense because time was arguably the strictest limit of the three up until then. But I think we would do well to consider whether we are letting the rules get too lax for the purpose we want to achieve.

I’m not asking to save teams from themselves. I love seeing what everyone is doing, so I wish I were more confident in my opinions here, but I can’t shake the feeling that there’s something amiss.

Edit: wow I make a lot of typos when I’m tired.


To sum up your point, it isnt about fairness anymore. That went out the window when it went from a 6-week challenge to an unlimited arms race.
With that being said, we learned a lot this first full year of a No Bag season.
Time to use whatever resources we have and can get our hands on, to not fall too far behind.


It’s never been about fairness, really, but there were at least vague nods toward parity. (Remember when you could only have four CIMs on your robot, and after that you had to resort to other motors? Remember when you had to put your robot in a bag at the end of build season and there was no withholding allowance?)

I agree with @Ekcrbe – and our last two robots are almost all Falcons (and, honestly, we should have put falcons on the intakes, too). There’s no trade-off, there’s no decision-making to be had; if you can afford the high-power brushless then you use the high-power brushless. Only the teams who can’t have to make do.

In the time I’ve been involved in FRC, the resources necessary to really be in the hunt at the regional level have skyrocketed – more money, more time, more space – and we’re lucky to have two of those three (thanks in part to “neighboring” teams who let us use their practice field, only seventy miles away!) The removal of bag-and-tag made things a little cheaper for approximately a minute, before “swerve or lose” stepped up to the counter and said, “hello”.

Just this past Saturday, I was speaking to a former mentor from a local(ish) team that folded recently, and he said (paraphrased), “It’s just too much. I spent every minute after school trying to get money, recruit more mentors, to design meaningful off-season projects that would help us on the competition field, chasing down storage space, finding new sponsors, etc, etc. With three full-time mentors–paid–we could give it our all just to show up to competition and have no chance of being in the top eight.”

…and there are many replies one might give this gentleman, but useful replies start with acknowledging that he’s got a point. The effort, energy, and resources you have to put into an FRC team just to be middle-of-the-road are astounding, and as the program matures all of that gets worse.

We have a pretty good varsity soccer team here in Naples, NY, for our division, with a couple of undefeated seasons in recent years. We also have an Olympic athlete who lives in town (multiple gold medals for rowing) who trained, and still trains, like an Olympic athlete. The level of effort necessary to run an even okay FRC team is more like the latter and less like the former, and getting moreso with every passing year, and like Erik, I have serious misgivings about it.

If the meta forces a “swerve or lose” on the floor, then what’s next? And what’s after that, and after that, and after that? Is it really beneficial toward the goal of transforming our culture to force kids to put in Olympian effort in order to be basically competitive at the high school level?

I, too, wish I were more confident in my opinions here, and I, too, can’t shake the feeling that there’s something amiss.


If we are talking about making Einstein, then a kit chassis has an almost zero chance of on-field success. But any concept of “swerve or lose” is still constrained to the highest levels of competition where every team’s above-chassis mechanisms are expected to be perfect, and even then, teams like 973 and 6328 have shown they can keep up.

If my team’s goal is to make DCMP every year and get picked for playoffs, I’m not even sure swerve is objectively the better choice. There were many swerve drives at Lehigh, but none of them ended up on the winning top-seed alliance. Next year will be different as more of the top teams make the switch, but I can still see an opportunity for tank drive robots being competitive for at least one more year.

Swerve is a powerful tool, but it’s expensive, and most importantly at my team’s level, buying swerve doesn’t guarantee that the rest of our mechanisms are going to be effective. If we can focus on making our other mechanisms as reliable as possible, that will easily overcome the shortcomings of tank drive.


Also, any drivetrain that doesn’t destroy itself every 6 matches will be better than the one my team used this year. One inspector I met at Champs said he was flabbergasted that we performed well enough to qualify, given the known unreliability of this product. I joined the team basically after kickoff and could not prevent the team from using it.

Most teams are low resource. For those teams, I’m not sure anything will change during the swerve revolution until the kit chassis itself is swerve. The 24th best robot at most district events will most likely be using tank drive, as no non-championship event outside of California has enough money distributed between the teams that 24 functioning swerves will show up. My point here is that the impact of the swerve revolution on these low resource teams may be less than this thread makes it out to be. Of course, teams with means should continue to help them learn and have fun.

As you may know @LukeB as a 2976 alum, there was a period of time where lower resource PNW teams made Houston Einstein en masse. Those glory days may be behind us with 1champs making things more competitive. But before I turn this into a post for that other thread, I should probably stop typing and get back to work…


I’m talking about “be in the running as alliance captain at a typical event.” That’s getting harder and harder to do.


Looking at some of the other tank teams, I think of 179 this year. They had everything a swerve team needed (auto turret, large intake), but they used tank drive. We saw how good they were at events this year. I think they will also be able to keep up well with the swerve teams.


But isn’t it more fair now, because good teams already built 2 robots to get around the bag rules to begin with?

Also, I don’t think it’s “swerve or lose”. But I do think that when swerve makes it 3x easier to become an alliance captain or win, it becomes a lot harder to justify tank. Personally, I don’t think 6328 and 973 are guaranteed to go tank next year. All the other reasonably good WCD teams I talked to this year wanted to try swerve in the offseason - and I think they’ll like it. The cost benefit ratio is just so high for teams with a lot of money.


Being in the running as alliance captain at a typical event has almost nothing to do with the drivetrain.

My team used the worst tank drivetrain possible. It was way worse than the kit chassis in terms of reliability - see the addendum in my previous post for an explanation. At our district championship, we lost multiple matches and RPs because of drivetrain issues. And yet, we were a semifinals alliance captain. Not to mention, the winning alliance of the event didn’t even have a swerve drive on it. With 5895 (no swerve, no turret, no climb higher than mid) putting up 60+ OPR, I find it unbelievable that this board thinks that any advanced feature is required to be successful.


It always has been. Compare the capabilities that teams that were #1 or #2 from the mid 2000s had to now. As the number of teams grow and the number of alumni mentors* grow, it will be more and more difficult to keep up with the joneses. This isn’t necessarily bad. It gets harder to do this because everyone is getting better.

*I have found that a student that participates on lower-to-mid level team and goes off to college and stays in the program has a huge impact on the teams they mentor. They spend 4 years learning all the problems, 4 years learning the solutions and then a few years learning how to mentor and all of a sudden, a team that has had no on-field success starts consistently being an alliance captain.


Don’t get me wrong–we were alliance captains at FLR on the #6 alliance despite being absolutely terrible this year. Luck definitely threw us into that spot; we didn’t earn it, we shouldn’t have had it, and we got vaporized by 870 and friends in the quarterfinals.

That kind of ridiculous “luck” doesn’t really have any bearing on my point.


Amen. I think the #1 difficulty of sustaining my program is that, if we do our jobs right, we never see those kids again (figuratively, and sometimes literally). Our town is beautiful, and is great for skiiers and hikers and vintners, and has absolutely nothing to offer an engineer that somewhere else can’t offer better. One of my alums became the lead engineer on the Falcon 9 rocket (up through the first successful double-landing of booster stages), one designs robotic prosthetics for a living, one helps put people out of work–that is to say, increases efficiency–via automation for one of the largest venture capital firms in the world, one works on Blue Horizons, etc., etc., etc. Those are all awesome things, and none of those opportunities exist here.

Some of our local teams that are extremely strong see a lot of alumni coming back to mentor after (and sometimes during) their college years. The institutional memory and drive that 340 and 3015 have gained over the past decade from returning alumni is astounding, and it shows – they’re both well on their way to being world-class teams, if they aren’t already. Some of the teams that visit our regional have the same benefit.

For teams that can manage it, it’s an amazing asset.


I agree with what you are saying to a degree.
But my response to Luke, was about removing limitations and design trade offs that he referenced.


What is your point? I don’t think most tank drive event winners were just lucky. And there were many of them this year. My example of 5895 still stands. Or you can take a look at any of the following repeat winners I highlighted from the Insights page on TBA:

If your point is that fielding a competitive robot has been getting harder since the beginning of time, I’m not sure that’s true. Capabilities have increased, sure, but I would say many things have gotten easier. For instance, vision tracking has become easier since the Limelight came out. Collaborating on CAD has become easier since Onshape came out and started offering the use of their product for free to the teams. And these are only the things that came out recently!

In the 90s, you basically needed an industrial machine shop to build a robot. You needed mentors who knew exactly what vendors to get every part from. My understanding is that there were no common FRC standards. They weren’t all using half inch hex shafts and 1.125-inch bearing holes. Programming was so low level that many teams found it difficult or impossible to teach students how to code. Comparing that to the current ecosystem with vendor cross-compatibility, WPILib, the Kit of Parts, and all of the awesome team-created resources, it’s no wonder we’re able to build such awesome bots nowadays.


My point was that if the OP’s premise is correct and this is the Fall of the House of Tank, that’s going to be yet another significant hurdle teams will have to clear in order to be basically competitive.