Another Look at 2v2 Super Titan

[GreatGonzales]

I know, I know, the dreaded forum-crashing topic. As well-worn as it is, now that Alex has taken some action to adjust the promotion threshold for 2v2 ST, and some time has past to let the new normal set in, I think this is worth another look.

A little backstory:

Before, there were only very few 2v2 STs (only 2), and this was the result of a more strict promotion threshold than the respective threshold for 1v1 ST. So, Alex took some action to allow more 2v2 teams into the highest league of Super Titan. Many feared that he would "over-correct" and allow too many teams in. That may be the case.

The numbers:

Rawkhawk and I were just promoted to 2v2 ST, bringing the current total to 33 (controlling for duplicates). As a percent of the total player base for 2v2, that's approximately 2%. For comparison, there are currently 116 1v1 STs, or .5% of the 1v1 player population. Please look at the OSN statistics page for some handy visuals of these numbers, keeping in mind that Codepenguin does not control for duplicate values.

Very likely, there are still some 2v2 ST teams who have not been promoted. I compared the existing 2v2 ST to their most recent overall rank, and there are some teams who will almost certainly be promoted when they complete their next league game (most notably: mihmo & Dharma (15) and EnterNicknameHere & Irwinss (27)), so the total number of 2v2 STs will likewise increase beyond 33, perhaps ultimately reaching about 40. So, IF we assume that the 1v1 ST league is the ideal standard, then there's something amiss.

However, it's worth looking at this from a couple different angles. 1v1 and 2v2 are quite different, both in playstyle and in the player population, so it may be prudent to have differing promotion thresholds. For example, the 2v2 player base is more "top heavy" than 1v1; Taking the Masters + ST players as a percentage of their respective player base, 1v1 is only 5% Masters and greater, whereas 2v2 is 23% Masters and greater.

With that in mind, we should compare the number of 1v1 ST players to the number of 1v1 Masters players, and likewise for 2v2. We get: for 1v1, 10.76% of ST + Masters players are ST; for 2v2, 8.35% of ST + Masters players are ST. So by this metric, 2v2 ST seems about right.

The non numbers:

Even so, I can't help this nagging feeling that the promotion threshold is now too loose. The 2v2 ST having the lowest true rank according to 3/15 rankings is timdoke & Prosspy, having rank 69th. I have played many of the lower ranked ST teams, and I'm not sure they're worthy of the highest league in 2v2.

I'll speak for myself - I also have a team with ACFD Lee, and we are currently Masters league and ranked 45th as of 3/15. We just defeated a ST team, alanthefast & raymundo8, suggesting that we may be promoted to ST if we win a few more. But I tell you, I don't think ACFD Lee and I deserve to be ST. We both make a lot of mistakes, him because he is a weaker player, and me because I use these games to relax and so I don't put a lot of thought into my moves.

Anyway, what do you all think about the new 2v2 ST league? Is it now too loose? Seems about right? I'm not sold on any particular argument; would like to know what the community thinks.

[CombatEX]

(03-21-2013 02:57 AM)GreatGonzales Wrote:  The numbers:
As a percent of the total player base for 2v2, that's approximately 2%.

0.9% actually. This is a little higher than 1v1, but not enough that I see it as a problem. Especially because I agree with you that 2v2 is more top heavy than 1v1.

I imagine you probably only counted 2v2 arranged. Since arranged and randoms play each other and the distribution is based on all 2v2 you really should include both random and arranged. Should 2v2 arranged play randoms in the first place? Now that's a good question. But as long as they do, you should really consider them together.

2v2% (1v1%)
0.9% (0.5%) ST 43
11.4% (4.5%) M 542
15.6% (26.3%) G 742
29.8% (31.6%) C 1418
42.3% (37%) F 2014
TOTAL 4759

I think my biggest gripe with the 2v2 distribution involves Fluffy and Gifted, not Super-Titan.

[baopu]

GG, you are a ST in modesty my friend, ha ha.

[GreatGonzales]

(03-21-2013 03:01 AM)CombatEX Wrote:  

(03-21-2013 02:57 AM)GreatGonzales Wrote:  The numbers:
As a percent of the total player base for 2v2, that's approximately 2%.

0.9% actually. This is a little higher than 1v1, but not enough that I see it as a problem. Especially because I agree with you that 2v2 is more top heavy than 1v1.

I imagine you probably only counted 2v2 arranged. Since arranged and randoms play each other and the distribution is based on all 2v2 you really should include both random and arranged.

FYI I went ahead and used 33 in place of 39, because the 39 from OSN is yesterday's number and also includes 6 duplicates. There are other duplicates I realize but I figured I'd at least control for 2v2 ST duplicates since this is the most important number.

Hmm. You think we should include 2v2 random players in the distribution? I'm not so sure... Though 2v2 arranged and 2v2 random teams may play together, they will never be included in the same league, right? When we discuss 2v2 arranged Masters league, for example, this is comprised 100% of arranged teams, no random teams, right? Plus, it's kind of weird to combine arranged teams (each 2 people) with random teams (each 1 person), seems like a category error. So I don't think it makes sense to include 2v2 random players in this analysis.

[CombatEX]

(03-21-2013 03:11 AM)GreatGonzales Wrote:  

(03-21-2013 03:01 AM)CombatEX Wrote:  

(03-21-2013 02:57 AM)GreatGonzales Wrote:  The numbers:
As a percent of the total player base for 2v2, that's approximately 2%.

0.9% actually. This is a little higher than 1v1, but not enough that I see it as a problem. Especially because I agree with you that 2v2 is more top heavy than 1v1.

I imagine you probably only counted 2v2 arranged. Since arranged and randoms play each other and the distribution is based on all 2v2 you really should include both random and arranged.

FYI I went ahead and used 33 in place of 39, because the 39 from OSN is yesterday's number and also includes 6 duplicates. There are other duplicates I realize but I figured I'd at least control for 2v2 ST duplicates since this is the most important number.

Hmm. You think we should include 2v2 random players in the distribution? I'm not so sure... Though 2v2 arranged and 2v2 random teams may play together, they will never be included in the same league, right? When we discuss 2v2 arranged Masters league, for example, this is comprised 100% of arranged teams, no random teams, right? Plus, it's kind of weird to combine arranged teams (each 2 people) with random teams (each 1 person), seems like a category error. So I don't think it makes sense to include 2v2 random players in this analysis.

I don't think they should play each other, but as long as they do, we need to consider them together. I imagine that a Super-Titan random player is rated because they have a similar hidden skill as a 2v2 random team (pretty amazing all things considered). As such I don't see why we would not look at the stats with both in mind. You're right that in general 2v2 arranged are at a higher tier than 2v2 random, but that is already reflected in the stats (the fact that the higher leagues are larger in arranged than random).

Also, sorry for my ninja edit. I posted this too:

CombatEX Wrote:Should 2v2 arranged play randoms in the first place? Now that's a good question. But as long as they do, you should really consider them together.

2v2% (1v1%)
0.9% (0.5%) ST 43
11.4% (4.5%) M 542
15.6% (26.3%) G 742
29.8% (31.6%) C 1418
42.3% (37%) F 2014
TOTAL 4759

I think my biggest gripe with the 2v2 distribution involves Fluffy and Gifted, not Super-Titan.

If we use your 33 ST number then ST is even more reasonable. I do think random and arranged shouldn't even play each other, but as long as that's the case, let's count them together.

[GreatGonzales]

Yeah I still don't think we should be grouping 2v2 arranged and 2v2 random and counting them as though they are as one. True, they may play together, but:

1. 2v2 arranged and 2v2 random have distinct, exclusive leagues, and we are specifically discussing leagues here.
2. Using OSN numbers, a 2v2 arranged team is 2 individuals; a 2v2 random "team" is 1 individual. So, I think at the very least we should divide the 2v2 random totals by 2.

[CombatEX]

Sorry, I should have elaborated before posting. Here is why I think it is necessary to consider them together. It did in fact occur to me that random is 1 player while arranged is 2 but, as I said above in terms of perceived skill a Master random player has reached the same skill as an arranged Master team.

Let us consider an arranged team and a random player with the same hidden skill. The arranged team's players have an advantage from knowing each others' play styles and basic strategies they want to execute on each map. The random player does not. What does this mean? Simply that the random player has reached the same skill hidden rating despite not having the same luxuries as the arranged team. It DOES NOT mean the random player is at a worse hidden rating (by the very nature of the problem statement - both have the same hidden rating). This is why I feel we should consider random players and arranged teams together as they are all open to facing the same opponents and hence if they are of the same hidden rating they should be treated equally in terms of distributions.

The fact that leagues are exclusive means nothing. It's a decision made purely for organizational purposes (it would look weird to have a list where some entries have 2 players and some have 1). What is important is that both random and arranged teams face the same competition.

I agree however that there is a question of how to count random players and arranged teams since it is 1:2.

[GreatGonzales]

But the purpose of this thread is to discuss the promotion threshold for 2v2 arranged, not 2v2 random. I'm not so much concerned with whether a 2v2 random masters player is greater than, less than, or equal to the skill level of a 2v2 arranged masters player; I want to discuss whether there are too many 2v2 arranged ST teams. And since no 2v2 random player(s) cannot possibly be counted among a 2v2 arranged ST, I don't think it makes sense to include them in a breakdown of the league distribution. In other words, this is really a discussion of arranged leagues, specifically; not 2v2 leagues in general.

[CombatEX]

I respect that, but in that case this discussion does not concern me. What I would like to discuss is separating 2v2 random from arranged, but until that is done I feel any conversation regarding 2v2 arranged distribution is overshadowed by the large number of random players skewing the results.

I believe Eijolend and I have encountered random players in only 1 game and I imagine this is similar for STs in general. As such it doesn't really have an impact on us as I imagine is the case with your ST team too, but as a whole 2v2 random teams have a huge impact on 2v2 arranged distributions since 2v2 arranged teams play 2v2 randoms whether you want to ignore it or not.

[TheGreatAnt]

I am not very large into statistics, but the similarity I have found between the top lists in 1v1 and 2v2 is that Super Titians 1v1 gravitate to the top of 2v2 list. I understand that their level of play will be higher, but I also question if 1v1 rank effects placement in 2v2. I don't know if AFCD Lee is 1v1 ST. But I see the competition of ranking as follows. Get highest in division -> Get to the highest division -> Get on the top 200 list -> Climb the top 200 list -> gloat over everyone else. So, I don't see how there can be "too many" STs. I just see it as a progress to getting to the best.

And as for you and AFCD Lee casually playing and making mistakes, well. The real matter is if you are winning or not. If you keep making mistakes, but still winning the game, yes you are going to get into the ST rankings.