Is specialization a viable strategy for Deep Space?


#61

That wasn’t me ant146 is Austin…


#62

That’s just semantics at this point, and arguably a misunderstanding of the word ‘specialist’. Specializing in one thing does not necessarily mean you outright abandon everything else, only that you prioritize it above everything else.

In fairness, the definition OP supplied suggests otherwise, but I don’t think it’s a good definition. If a team is so much more efficient at scoring cargo than hatch panels that it only makes sense to score a hatch panel during the sandstorm, when it also effectively scores a cargo, would that really not be a cargo specialist? Suppose another robot could pre-load and score a hatch panel, but had no means of intaking one during a match. Is that not a specialist either?


#63

that would be a rather interesting robot design, since the panels are scored and retrieved about the same way.

There are indeed different ways of looking at “specialization”. In this context, I think it’s fair to assume that a “cargo specialist” would not be able to handle hatch panels, and a “hatch panel specialist” wold not be able to handle cargo. A robot that can do both, might be a “low goal specialist”, or might be a generalist, able to score everywhere.


#64

Let’s say you are in a pick list meeting and you are looking at a specific robot that has the capability of scoring panels but it is slow and is much faster at scoring cargo. They are NOT a cargo specialist. You would describe them as a robot who is really good at scoring cargo with the ability to score hatch panels. They didn’t specialize, they just didn’t design a good hatch panel mechanism.

Specializing your robot design is deliberately designing your robot to only play part of the game so you can be much more efficient at it and be a strong contributor for your alliance.

To be a master of one task or a jack-of-all-trades, that is the discussion. For teams that can’t be great at every task, there needs to be a discussion about whether to specialize or not.


#65

Well if you’re putting it like that. Is a robot that can do both hatch panels and cargo but only in the lowest level count as a specialization? Especially since the lowest level is where most of the scoring positions are anyway.

So I guess what I’m asking is, are we considering scoring on specific levels specialization? Or just manipulating one game piece or the other?


#66

FTFY


#67

Yes! If you quoted all of my post, you will see that I said, “Specializing your robot design is deliberately designing your robot to only play part of the game so you can be much more efficient at it and be a strong contributor for your alliance.

They are deliberately designing their robot to not play all of the game to more efficient at the part that they do play.


#68

My mistake. Well it looks to me that they are definitely viable. With the nature of these game pieces the cargo are dependent upon the hatch panels. Because of this, if you have a specialist in hatch panels then they could lay the groundwork for cargo placement for other robots.

But for a Cargo specialist on the other hand, it will be tougher for them particularly in qualifiers. If they don’t get paired up with a robot that can do hatch panels they can’t do much. But in elims, they will have to be good at scoring the cargo. If they can keep up with their alliance partners that are scoring hatches things should turn out fine.

And if you’re just scoring on a specific level, that looks good as well. Since you can pair with a robot that does another level, or all the levels in some cases, and mesh well in an alliance. Especially if both are good on their particular levels. And also keep in mind, the lowest and easiest level is also where most of the scoring locations are. If you can score well down there, you will likely be good at the game in general.


#69

I believe that specialization may happen if they are strictly a ramp bot with a hatch collector, but its just my own personal thoughts


#70

Figured that out finally–he’s using a disguise! :yum:


#71

Teams that opt to build ‘specialist’ robots will typically perform well, because those are the teams who will have put a considerable amount of thought into their strategy for the game, rather than try and tackle everything like many teams try and do.

More specifically for this season, I will cut against the grain here and state that I don’t think station-loading hatch specialists will find all that much success in playoff alliances. Consider that the majority of playoff alliances will consist of two primary scorers and a third utility bot. In addition, the geometry of the loading stations will make it very difficult for three robots to effectively split time loading. Unless there’s a particularly extraordinary hatch specialist still on the field (i.e. one that can score more hatches than any other remaining bot can score total pieces), I would not spend a first-round pick (and essentially a human player station) on a robot that has limited scoring potential. These bots become even less desirable in the second round, as most alliances will have select two station-loading robots by this point, essentially restricting these teams to playing defense.

I think there will be more room (most likely in later events) for cargo-specific specialist robots for the same reason. Unlike hatches, cargo is not restricted to the human player station, and will be in ready supply spread across the field in most matches. These robots would play a similar role to stack cappers from 2015, scavenging cargo from the field, and filling a bay after the primary scorers have placed the hatch. Thus, these types of robots can deliver considerably more scoring power to an alliance while the two primary scorers utilize the human player station to retrieve hatches.


#72

:joy::rofl::sweat_smile:


#73

I have to agree with ant146’s interpretation–a “specialist” does only that task or does the other tasks so poorly that they would not be assigned that task in a playoff alliance. Last year switchbots could not place on the scale. Some exchange bots couldn’t even place well in the switch. In 2017, 900 could not place gears and only shot fuel, meanwhile 180 could place only gears and had no fuel shooting capability. I expect this year that there may be panel specialists who have no way to handle cargo, but they will be viewed as being as valuable as switchbots were last year. Cargo specialists may be valuable this year, but they will be dependent on using null hatches and preplacement by other alliance mates. This will make them less valuable to higher performing alliances that may be able to score cargo quickly on their own. When we develop our strategy and pick lists, we will use “specialist” in this way.

Teams may “focus” on one aspect over another, just at 148 and 1538 did in 2017 by shooting fuel only during the auto period, and as many good exchange bots did over working the switch. But those are not specialists.


#74

While it might be a reason, id wager the prevailing reason is cost/time (and not always by choice of the team)

Id also wager that they will be as useful as switch bots last year, aka not very useful.


#75

What I meant is that they tend to be the teams aware enough to step back and realize that they don’t have the resources to adequately cover all of the game tasks, and instead focus on a few and narrow down their overall strategy. These teams will do better than those of equal ability who try and tackle everything and not have a clear idea of where they’re going.


#76

To be honest, I think that specializing in scoring in the second or third level of the rocket is unlikely to be effective because there will be so few robots with this specialty. The small number of goals at each of these levels, combined with there being no bonus scoring for one of level 2 or level 3 are the main reasons why. On the other hand, a team which decides to focus on both of the upper stages of the rocket while ignoring the ground level is going to be a great first choice for a low goal + L3 HAB specialist.


#77

For what it’s worth we are more or less following this as our strategy. In our prioritization we said scoring a preloaded Sandstorm hatch was a must, so it comes before anything else. Cargo scoring and floor pickup of cargo are next before the ability to score any more hatches. We have a hatch scoring mechanism but it plays second to ability to handle cargo efficiently so it is not located in an optimal position and we have spent much less time on it. We won’t be that efficient at acquiring hatch panels, which isn’t an issue for the preloaded piece.

Our design decisions are partially driven by the only event we will attend (AZ West). It is one of the weaker events for average team capability so we wanted to maximize points we could do if our partners can’t do anything. That means hatch panel Sandstorm + filling the 6 null hatch panel bays with cargo. We realistically won’t get a two hatch Sandstorm done and will probably have trouble getting beyond 6 teleop cycles so that optimizes our scoring potential. Plus we then win all the RP tiebreakers with our really high total cargo score.


#78

That is a well thought out rationale for your team’s strategy.


#79

Actually I totally agree with you here. Just 2nd level or just 3rd level probably won’t be viable. There just isn’t enough scoring locations on those levels. The 1st level has a lot of scoring locations though, with some variation with cargo when it comes to the Cargo Ship and Rockets. Otherwise yes, when it comes to levels of placement they will either have to do 1st level only, both 2nd and 3rd level only, or all levels.


#80

I don’t see anyone specializing only in 2nd and 3rd levels. It’s already more technical to score at those levels, so the teams able to do that can most likely do 1st level as well.