Round Robin on Einstein

[Joe Ross]

I think it would be great if the tournament structure on Einstein was changed to incorporate a round robin portion. I think the following structure would be an improvement on Einstein:

Have the first round be round robin. Each team plays every other team once. The two teams with the best record then play best 2 of 3 for the championship. This would require 6 matches to get through the round robin part and an additional 2 or 3 matches to crown the champion.

The biggest advantage of this system is everyone is guaranteed 3 matches (up from 2 under the current system). Since these games often play out like rock-paper-scissors, it allows each of the teams a chance to show off all their strategies. The alliances that have made it to Einstein have worked hard and deserve the chance to show everything they have, in front of the entire audience. It can be very disappointing to go out there and lose two quick matches and be done.

Another advantage is that this might make scheduling easier for FIRST. Under the current system, there are a minimum of 6 matches to play and a maximum of 9. With the round robin plan, there are a minimum of 8 matches, and a maximum of 9, which should make things more consistent year to year and easier to plan.

Here's how I think it would play out this past year.

Newton beats Galileo (like they did in the first match)
Curie beats Archimedes (like they did in the first match)
Newton beats Curie (guess)
Galileo beats Archimedes (guess)
Newton beats Archimedes (guess)
Galileo beats Curie (like they did in the finals)

Since there are awards between the matches, you wouldn't have to worry about turnaround time.

At the end, Newton has a 3-0 record and Galileo has a 2-1 record, and then play best 2 out of three. Both alliances have now seen a lot more of the other alliance and can better plan, plus you have the best two meeting for the championship, rather then in the semifinals.

About the only disadvantage that I can think of is the GDC would have to figure out a way to break ties. This could happen if there were 3 2-1 teams and a 0-3 team. An easy way would be the most points scored followed by the least points allowed, similar to the way the do it in soccer.

[DonRotolo]

I like it, for the reasons stated.
Of course, you mean "alliances" and not "teams", but I understand the point.

[Graham Donaldson]

I like it as well. For example, if we look at this year, alliances facing Galileo had very little chance of beating them 2 of 3. Suppose that the field Galileo faces in the semi's (I forget which) is better than either of the remaining two, yet does not get to play in the finals because they were forced to play Galileo.

A minor point, to be sure, but just another reason to change the system. If it works smoothly enough, maybe FIRST would considering expanding it to field playoffs, and even regionals, however unlikely in the near future.

[R.C.]

I like this also, maybe 1114, 217, and 148 might of lost this season to 233, 968, 60 in the finals???

[Cory]

I agree with Joe. The last two years, the two best alliances faced each other in the Einstein semi finals. The finals were almost anticlimactic, in some ways. I would greatly prefer this method. It would produce a more deserving champion, rather than having one alliance make it to the finals by playing a weaker division.

[Dan Petrovic]

Yeah, I really like this idea.

Plus, who cares if the matches on Einstein go on longer? They're awesome matches to watch!

[tennispro9911]

For the sake of playing devil's advocake.

First of all, if it isn't broken, don't fix it.

I do agree that sometimes certain fields produce better winning alliances than others, but all 4 alliances on Einstein are still good enough to beat any of the other ones in a given match. It may seem to be an assured victory, but we know that nothing is ever impossible, especially from an alliance that has already won on a field.

Robots will break. I know that teams need to build robust robots, but things happen. Those robots have already played 2 or so matches in qualifying, 3 rounds to win thier field, and now you're asking them to play another 4 rounds on Einstein. 23 matches in a day is a lot for any team and any robot.

Ties. Matchups matter a lot. Matchups mean everything especially because of defense. A beats B which beats C which beats A. With close matches, this will happen a fair amount. I'd bet at least every other year you are going to have a tie, which means you need to be able to break the tie.

Can anyone think of a truely fair tiebreaker? Head-to-Head is great, but doesn't work for a 3 way tie. Points scored? ... doesn't account for defensive powerhouses. Point differential? ... much better, and this would work well in a game that is scored like this game, but in a game that is scored exponentially like 2007 and others, that can be hugely skewed by one match that an alliance scores say 128 points. Even point differential can be hugely misleading.

Also, I know no alliance would ever tank a match or anything like that, but you also can create the situation where one alliance is up 2-0 and guarenteed a spot in the finals, and another is down 0-2, and already eliminated. You make it so that a match (or two) could be played that doesn't have any chance of affecting the standings.

That said, I love the idea. I think it would be great. I'm mainly pointing out possible flaws/weaknesses so that the idea can be strengthened.

[Akash Rastogi]

Originally Posted by tennispro9911

For the sake of playing devil's advocake.

First of all, if it isn't broken, don't fix it.

If we stuck by that saying, then we might not have that thing called evolution and innovation

I like the idea.

[GaryVoshol]

Originally Posted by tennispro9911

Can anyone think of a truely fair tiebreaker?

That would be a concern.

Also, I know no alliance would ever tank a match or anything like that, but you also can create the situation where one alliance is up 2-0 and guarenteed a spot in the finals, and another is down 0-2, and already eliminated. You make it so that a match (or two) could be played that doesn't have any chance of affecting the standings.

For that reason, the last two group matches are played simultaneously in the World Cup. This came about several cycles ago when the last match scheduled didn't mean anything if the teams tied, but if one of them won, a team watching from the sidelines would advance. Both teams playing that last match were content to bunker down for the tie. While we might say that GP would never allow this to happen, why tempt it?

If those two problems could be addressed, I'd be interested.

Another alternative would be a double-elimination tournament:
Round 1
- A beats C
- G beats N
Round 2 ("loser's bracket")
- N beats C
Round 3 ("winner's bracket")
- G beats A; G advances to finals
Round 4
- N beats A
Finals
- N plays G. N must win twice to take the Championship; G would only be required to win once.

No team is eliminated until they lose twice. The only variable is if the finals go to one match or two.

Edit: A double-elimination scheme would work for regionals and the division finals as well.

[smurfgirl]

Originally Posted by tennispro9911

Ties. Matchups matter a lot. Matchups mean everything especially because of defense. A beats B which beats C which beats A. With close matches, this will happen a fair amount. I'd bet at least every other year you are going to have a tie, which means you need to be able to break the tie.

Can anyone think of a truely fair tiebreaker?

I have to agree with this. I love the idea- I think it would be more fun to play in a tournament structured like that, and more fun to watch it as well- but I was wondering what exactly would have to be done about the ties. As tennispro mentioned, certain methods seem more appropriate for certain games- maybe there would be a different tie-breaker method each year that the GDC thought would work best with the game?

[Elgin Clock]

Originally Posted by smurfgirl

I have to agree with this. I love the idea- I think it would be more fun to play in a tournament structured like that, and more fun to watch it as well- but I was wondering what exactly would have to be done about the ties. As tennispro mentioned, certain methods seem more appropriate for certain games- maybe there would be a different tie-breaker method each year that the GDC thought would work best with the game?

That's an interesting proposal. As if I didn't already have a love/hate relationship towards the GDC & the rules after all the crazy rules in the 8 years I've been involved, what if they decided that in the event of a tie the rules were changed up a bit for that tie-breaker?

For example, in the event of a tie this year, we could have had a rule that said the tie breaker game would have added an additional trackball into the mix for each team (ala' Beantown Blitz this year).
(Just as an example).

The change for the tie-breaker would have to be something that would swing the game in favor of an alliance who understood the game completely, because with a tie you've already proven you are (so far) equally matched playing it "regular style". (Ever get bored playing a competitive racing video game in the normal way, so you challenged an equally matched friend to race the course with you backwards for example & see who could handle the change?)

In this idea, you would throw in some leverage for the alliance who knows the game inside & out & would see who the best is by seeing which alliance could prove themselves & play it (even more) strategically than the first 2 matches.

[EricH]

Originally Posted by InfernoX14

Plus, who cares if the matches on Einstein go on longer? They're awesome matches to watch!

Actually, I think it's to reduce the number of matches that might need to be played. And yes, we care if they run long! Team Social (or whatever it's called) is calling, and we want our share. But we can't get it unless we leave early and miss matches. Also, NASA TV broadcasts the finals--but ends coverage before the matches end. Why? Schedule! If it's more consistent, then it'll be easier to figure out when you can switch to something else--if you're early, you can end early.

I like the idea too.

As for "if it isn't broken, don't fix it", the engineer has two counters for that. One is "If it isn't broken, it doesn't have enough features" and the other is "If it isn't broken, fix it 'till it is".

[Cory]

Originally Posted by EricH

Actually, I think it's to reduce the number of matches that might need to be played. And yes, we care if they run long! Team Social (or whatever it's called) is calling, and we want our share. But we can't get it unless we leave early and miss matches. Also, NASA TV broadcasts the finals--but ends coverage before the matches end. Why? Schedule! If it's more consistent, then it'll be easier to figure out when you can switch to something else--if you're early, you can end early.

I like the idea too.

As for "if it isn't broken, don't fix it", the engineer has two counters for that. One is "If it isn't broken, it doesn't have enough features" and the other is "If it isn't broken, fix it 'till it is".

For the record, NASA knows that the event will never end on time. FIRST's refusal to modify the schedule to accurately reflect the end time is the reason why coverage is cut short.

[Koko Ed]

Originally Posted by Cory

For the record, NASA knows that the event will never end on time. FIRST's refusal to modify the schedule to accurately reflect the end time is the reason why coverage is cut short.

It's not the fault of the competition itself. It's the endless parade of people giving speeches, which kills me because people have been complaining about too many speeches going on for too long for years! So what's FIRSTs response? Get more speakers! Way to know your audience guys!

[JVN]

Originally Posted by Koko Ed

It's not the fault of the competition itself. It's the endless parade of people giving speeches, which kills me because people have been complaining about too many speeches going on for too long for years! So what's FIRSTs response? Get more speakers! Way to know your audience guys!

Do you really believe you are the target audience? I have a feeling I'm not.