Ladle 42 Voting Discussion

[INW]

The 3rd and 4th seeds don't necessarily mean that the 3rd is better than the 4th. The point in doing this like said above to ensure the 1st and 3rd or 2nd and 4th seeds don't meet again until the finals to prevent the same semi finals match-up as the previous ladle.

[Concord]

then call them both 4th seeds

[Mkay1]

Then call winner and runner up second seed.

If anything call it third, but that is illogical. Two three seeds.

[sinewav]

Fine. We'll just add a vote for a resolution on the 3rd and 4th place seeds.

Something like: 3rd-Place Match | Randomized | No Change

Note that "Randomized" is essentially going from a 4-seed system to a 2-seed one. Looking back over the seeding discussion that got us this far, almost no one was in favor of a 2-seed system. The general feeling was that having only 2 seeds wasn't worth the time and wouldn't impact the game enough. And yes, the 3rd place match was brought up then too.

There was a thorough, fruitful discussion on seeding that got us to this point. I'd hate to see it undone. But if you want to drive up the page count of this thread, more power too you.

relevant discussion.

[Concord]

no, this isn't something that changes the playing of the Ladle, it doesn't go to a vote.

randomized simply means you flip a coin for which of the two 4th seeds gets to face the 2nd place finisher in the next ladle and gets to call themselves a 3rd seed.

[INW]

I disagree.

It should be a vote.

[Concord]

The 3rd and 4th seeds don't necessarily mean that the 3rd is better than the 4th. The point in doing this like said above to ensure the 1st and 3rd or 2nd and 4th seeds don't meet again until the finals to prevent the same semi finals match-up as the previous ladle.

We have yet to have all four seeds advance to the semifinals. And in 41 Ladles, we have never have we had consecutive ladles with identical semifinalists, and this doesn't even take into account match ups. It doesn't happen in professional sports either. Major League Baseball is an apt comparison, each league sets 16ish teams competing for four spots. The type of occurrence you describe has happened once in the American League Divisional Series (1998&1999) and never in the NLDS. That's 1 occasion out of 30 potential opportunities for it to happen.

Now, under the current system, we punish potentially the 3rd best team by putting them in the same bracket with the best team. Both the 3rd and 4th seeds lose to both the 2nd and 1st seeds, meaning that who they lose to does reflect any difference between the two teams in question. It follows that the better of the two may be given a worse seed, for no reason other than the system. I have proposed two solutions to this problem. A third place game ensures that the third best team isn't punished. The other alternative is to decide randomly which of the two losing semifinalists is the 3 seed and which is the 4 seed. The first suggestion is fair in it's ensuring the better team is not punished. The second one is unfair in that it gives the better team and the worse team equal chances at getting punished. In some sense, random assignment is similar to the current system.

The current system reinforces it's incorrectness with each increasing ladle, however. It's no accident that brackets have become increasing uneven since we adopted seeding. And I can explain how if you are interested.

[Mkay1]

I get it now, nice post.

Plz do concord .

[INW]

I do agree with the 3rd place (consolation) match to determine the last 2 seeds and I would participate in it. I am just not sure if the 2 teams would. If they both opt out of it, we should then determine the 3rd and 4th seeds like we currently do and if 1 team wants to do the match and the other doesn't, the forfeiting team agrees to take the lesser 4th seed. This "consolation" match could occur during or after the finals?

Just a thought.

Btw, I don't think the word "punished" is the correct term to use but rather an unfortunate paring of teams in the same "half bracket".

[Mkay1]

A third place match would suck. An extra hour on intense action. Nty.

[sinewav]

And I can explain how if you are interested.

Yeah, go for it. As I look at the L-37 to present, I'm seeing an awful lot of the same people getting to the finals as before L-37 (except uNk is the new TX). I think you could just as easily say seeding has had almost no effect on Ladle, except we don't get those crazy lopsided ones like L-34.

If I were to choose, I would have a 3rd place match. But most people are either exhausted or ragequit to even have another match. And how tedious is it if your 3rd place match is still going on after Ladle ends and everyone is posting about the winners? Maybe while you're at it, you can think of an incentive for people to play it.


Word: If you're reading this, you can copy all of Concord's contribution to this thread and edit it for zomgholers. Then it's like PlayFort never ended, haha.

[Concord]

Seeding began after Ladle 37, and so the semifinalists of Ladle 37 determined the first seeds. I will remind us now that the brackets for Ladle 37 were completely random and to become a semifinalist was potentially quite easy. Fortunately, the brackets were somewhat even, and the two easy quadrants happened to have finalists in them. The seeds initially were quite fair. 1. Crazy Tronners 2. Speeders 3. Rogue Tronners 4. Team Unknown. We know these were fair since three of these teams returned to the semifinals. They happened to be the top three seeds, Team Unknown was dispatched in the first round by Drug Addicts.
This is where we started to get in trouble, the very first round of the very first seeded Ladle. We have a new team, evidently quite good, beating a seed. The combination of seeding and the fact that the player pool allows for only 1-2 good new teams to be spontaneously created each Ladle almost ensured Drug Addicts would get to a semifinal. Seeding ensured no previously proven team would challenge them in the quarterfinals, and as a consequence of the player pool, no new team would either. So Drug Addicts gets to the semifinals at which point they are assured a seed. They lose to Speeders who lose to Rogue Tronners and thus are assigned a four seed. The former four seed, Team Unknown, is unseeded. Rogue Tronners get a 1 seed for beating the at least one of the three best teams (Speeders). When Rogue Tronners beat Crazy Tronners in the Ladle 38 semifinal we do not know whether they beat the third best team or the fourth. However, Crazy Tronners are said to be the third best and therefore are in the same bracket as Speeders in the next Ladle.

Ladle 39 now. Let's start with the problem area again. Drug Addicts and Team Unknown are again slated to meet before the semifinals, partly because they did the last Ladle. This time Team Unknown wins, and reaches the seed round. They lose to the eventual 2 seed, Rogue Tronners, and become the 4 seed again. This is important. Summarizing team Unknown's Ladle, they beat potentially the third best team in the previous Ladle and lost to the best team in the previous Ladle. They are a 4 seed. Oracle, a new team, beats the 2nd best team from the previous Ladle and then loses to potentially the third best team from the previous Ladle. How does our system going about whether Team Unknown or Oracle deserves a better seed? It plays a match between the two teams that beat them, Rogue Tronners and Crazy Tronners. As it turns out Team Unknown is declared the 4 seed and Oracle the 3 seed.

Ladle 40. Oracle is a 3 seed and to reach the final beats Speeders (who they beat the prior Ladle as well). For beating the Speeders in a quarterfinal in Ladle 39, they eventually got a three seed. For doing the same thing in a semifinal they get a two seed. Now, Speeders had beat the 2 seed Rogue Tronners in Ladle 40 when they hadn't in Ladle 39. This doesn't make Speeders any better a team necessarily, they didn't play Rogue Tronners in Ladle 39 at all. Why not? Because they were seeded completely opposite them. So Oracle, despite turning in an essentially identical performance, gets a 2 seed simply because of how their opponent did. Had Rogue Tronners won Ladle 39 and been the first seed, things might have turned out identically. Oracle would have then been the four seed, in the same bracket as Rogue Tronners. Team Unknown would have been the 3 seed and would have been in the same bracket as Crazy Tronners. The Ladle 40 brackets were independent of the Ladle 39 final. The seed pairings would have been identical regardless of the final outcome. This is also the case for every Ladle. This is profoundly troubling.

Oracle creates a 2 seed out of thin air, and then proceeds the fall apart. The Ladle 41 2 seed goes and loses in the opening round. This has consequences, it means Team Unknown, who has not beat a Ladle finalist in the concerned period (Crazy Tronners, Speeders, Oracle, Rogue Tronners) gets a free pass to the final and to a 2 seed. It also means that Speeders (that team that lost to Oracle in both Ladle 39 and Ladle 40) loses to a team of which half is made up of Spin, Lizmatic, Eckz, the same half that beat them the previous Ladle. Now, shouldn't Crazy Tronners, who in the prior two Ladles beat Rogue Tronners, Oracle (2x), Team Unknown (three of the semifinalists in Ladle 41) have been waiting to once again defeat this half-Oracle? Well, because the Crazy Tronners faced a winner of a play-in game, they had a greater chance of facing one of these new teams that spurt up than the non-seeds. Twixted Xats isn't a new team, but the lineup that faced CT was.

If you still follow, congratulations. For those who didn't I will summarize the major consequences of these seeds.
1.) There are more than four good fortress teams. This means a seed will almost certainly fail to return to the semifinals. Which of the seeds this will be is random, but that seed then becomes the fifth team that in the next ladle often deseeds someone.

2.) Watch what happened to Team Unknown. Ladle 39, they lose to a loser, become the 4th seed. Ladle 40, they lose to a winner, an opportunity they had only because they lost to a worse team the ladle before, and then become the 3rd seed again.
Over the same time span Oracle beats Speeders in Ladle 39, displacing Speeders seed. They then beat Speeders in Ladle 40, this time it is in a semifinal because they beat in the quarters before and the randomization set them just so. By simply beating Speeders in consecutive Ladles, Oracle becomes a 2nd seed.

3.) These two events coincide to set up Team Unknown getting a bye to the finals. Now becoming the two seed, they are poised to do it again unless randomization interferes, but wasn't the whole point of seeding to avoid this. Should we be at a point where we depend on a certain randomization to make the brackets even?

Now, here's the problem, the bottom bracket is at risk of a seed being won by upstart teams. This happens because the 3 seed only has to lose to the winner and the two seed only has to beat the 3 seed. The bottom bracket becomes systematically weaker. Upstart teams are more likely to disband or weaken rapidly, which further aggravates the problem. The problem is that the new four seed, which was the old crappy three seed, is eventually recycled back into the weak bracket. Teams from the top bracket have won three of the four seeded Ladles, Ladle 39 being the exception. This was a good final as was Ladle 38, mostly because the seeding had yet to take hold. This might seem insignificant, however, the 1 seed has won just 1 of the seeded Ladles. This indicates how much stronger the top bracket is becoming and how weak the bottom is.

How might this have been solved by a third place game? Well, had a third place game been played in each of these Ladles, the bottom bracket now must contain the 2nd and 3rd best teams. This does not run aground after five ladles, like the current system has. The current system where we're not even assured of the 2 seed being the 2nd best, since all they have to do to become the 2 seed is beat their semifinal opponent, who has become continuously underskilled.

[Concord]

The Ladle 40 brackets were independent of the Ladle 39 final. The seed pairings would have been identical regardless of the final outcome. This is profoundly troubling.

Just to elaborate on this. This is so troubling because it suggests that, despite seeding, each Ladle's outcome is completely independent of the previous Ladle's. Seeding was supposed to link them together to help even out the brackets. If you don't buy that our particular seeding is having the opposite effect, then at least recognize that this then means our seeding is meaningless.

Before the final is played, the winner of semifinal A will be paired with the loser of semifinal B and vice versa. The final determines top or bottom bracket, that is all.


Here's an illustration, and this is the case for every and any pair of Ladles.

Code: Select all

Ladle X

If
Team A - future Seed 1
def.
Team B - future Seed 3

Team C - future seed 2
def
Team D - future Seed 4

Then Ladle XI

Top Bracket
Team A
Team D

Bottom Bracket
Team B
Team C

Code: Select all

Or equally


Ladle X

If
Team A - future Seed 2
def.
Team B - future Seed 4

Team C - future seed 1
def
Team D - future Seed 3

Then Ladle XI

Top Bracket
Team B
Team C

Bottom Bracket
Team A
Team D

[sinewav]

Well done. But I'm still not convinced we have a problem. Even if the system is faulty, it does two things it was supposed to do. The first is to stop the two winners from the previous Ladle from meeting in the opening round of play. That's an important one and it does it well. The second is that it gives people a good feeling to get a seed, even if it is void of real meaning.

You should direct your energy in a different way. I think the best solution is a 3rd place match. How do you get people excited to play it?

[INW]

Very good job Concord and see your points. One thing I did notice was this:

2.)...By simply beating Speeders in consecutive Ladles, Oracle becomes a 2nd seed...

Oracle beat SP twice, CTb, TX, and Drug Addicts to become a 2nd seed

But I do see your "logic" in this. Your point about there are more than 4 good teams in fortress goes back to the vote on how many seeds (2/4/8/all).