A Statistical Overview of the "Ban ADP" Debate

Christian Chase 


Editor's note:
The conclusions in this article aren't representative of everyone on the PokéStats staff, but whether you agree or not, Christian has done some very solid statistical analysis and given good thought to his conclusions. I agree with his finding that ADP does not affect metagame diversity to a greater degree than other decks in previous formats (think Buzzwole-GX or Yveltal-EX/Garbodor), but “gatekeeping” or affecting the meta has never been my problem with ADP; I dislike ADP because I feel it significantly decreases the skill cap of any format it exists in. Regardless, the analysis here is worth reading and is useful for metagame analysis beyond simply the ADP debate. For those of you who don’t know Christian, I think he is one of the most underrated newer players in North America and he has a great mind for the game; I’m excited to work with him on more PokéStats content in the near future! (He’s been doing a great job keeping our OnToPP rankings running!) -Tate


Hey guys! My name is Christian Chase, and I have been playing competitive Pokémon TCG for about two years now, finishing the abbreviated 2019-20 season with 554 CP. I have been a part of the PokéStats team for about six months now; as a math major and statistics minor at the University of Florida, I'm very interested in the statistical and data-oriented side of the game we all love, and since I'm not the biggest fan of online play, this is where I've been focusing much of my Pokémon attention recently. 

Currently, one of the more controversial topics within the PTCG community is whether or not Arceus & Dialga & Palkia-GX ("ADP") should be banned. I decided to dive a little deeper into whether or not an ADP ban is is warranted. The algorithms I utilized for this are pretty baseline and intended for a wide audience to be able to understand the results. 


Overview:
  • A statistical analysis of why ADP is not the BDIF
  • Is a meta with ADP healthy?
  • Is the meta without ADP healthy/healthier?
  • A possible solution
  • Statistics of the solution
  • A look at the current Standard metagame through nine top decks


Decks

Eternatus

ADPZ

Blowns

Mew3

Centi

Pika

Luca w/ Zong

Inteleon

DeciGoons

Average

Eternatus

0.5

0.5

0.45

0.6

0.45

0.55

0.4

0.55

0.3

0.4777777778

ADPZ

0.5

0.5

0.5

0.4

0.4

0.45

0.6

0.55

0.4

0.4777777778

Blowns

0.55

0.5

0.5

0.55

0.55

0.55

0.6

0.4

0.35

0.5055555556

Mew 3

0.4

0.6

0.45

0.5

0.5

0.5

0.5

0.5

0.5

0.4944444444

Centi

0.55

0.6

0.45

0.5

0.5

0.5

0.35

0.35

0.35

0.4611111111

Pika

0.45

0.55

0.45

0.5

0.5

0.5

0.45

0.8

0.35

0.5055555556

Luca w/ Zong

0.6

0.4

0.4

0.5

0.5

0.55

0.5

0.6

0.35

0.4888888889

Inteleon

0.45

0.45

0.45

0.5

0.5

0.2

0.4

0.5

0.4

0.4277777778

DeciGoons

0.7

0.6

0.6

0.65

0.65

0.65

0.65

0.6

0.5

0.6222222222

Table 1: Analysis of the current Standard metagame through nine "top" decks.

This is a very rudimentary look at the current Standard metagame, including nine of the top decks and their matchup spreads. The average matchup spread has some relevant external factors that I will address when looking at each deck specifically. (Shoutout to Justin Lambert and Justin Kulas for their help with creating the matchup spreads.)


Let’s get into it. First, two disclaimers: 

1. DeciGoons
The matchup spread I present here points towards Decidueye/Galarian Obstagoon being a top-tier deck because of its wide range of positive matchups. But these matchups are based on opposing lists that don't have techs for DeciGoons. DeciGoons is a deck that a) lacks in consistency compared to the rest of the format and b) and is extremely dependent on the techs that its opponents play. The deck can win, and has found itself in top cut in multiple online tournaments (most notably winning a large Hegster TCG event); however, as its meta share grows, decks will tech for it. With techs put into consideration, DeciGoons would most likely find itself at the bottom of the matchup spread chart. These external factors are too significant to ignore, and because of this I will treat DeciGoons as an outlier and not include it in the final analysis.

2. PikaRom vs. Inteleon
This matchup skews the matchup spreads of both decks significantly, because of how favored PikaRom is over Inteleon. Inteleon (43%) looks a lot worse, and PikaRom (51%) looks a lot better. This is actually one of the few very lopsided matchups in this format (something that will be explored more ahead), and taking this matchup out of both decks’ matchup spreads leaves Inteleon at 46% and PikaRom at 47%, which are both better representations of how the decks actually fare in the current meta.


Can we say ADPZ is the BDIF?
With multiple decks so close in these matchup spread rankings, it is hard to definitively choose a BDIF based on matchups alone, so this choice has to come down to deck consistency--put another way, how often the deck’s engine allows it to achieve that ideal matchup spread. We can identify this by identifying what deck “takes the least” to achieve its core strategy. Obviously, ADPZ doesn’t need much to pull off its overall strategy, but when going second, it needs a hefty amount of cards to pull off the turn 1 Altered Creation GX. Because this isn’t as likely as Eternatus’s core strategy of simply attaching an Energy to an Eternatus V on turn 1 (whether going first or second), I believe ADPZ cannot be considered the BDIF, considering that its matchup spread is not significantly better than other decks. What makes both ADPZ and Eternatus “S tier” is their lack of reliance on external factors. They both have almost identical 50% matchup spread and their game outcomes are based primarily on sequencing skill and drawing luck. Other decks in the format rely more on external factors such as specific matchups, bricking off Marnie, etc. ADPZ and Eternatus remain mostly unphased by those factors due to their very strong draw power and linear strategies.

Now, I am not saying Eternatus is definitively the BDIF, but rather that it does what ADPZ does, just a little bit more consistently. Because of this ADPZ, in most cases, will be slightly behind Eternatus in the expected metagame analysis for most tournaments currently.

The standard deviation of the matchup spread I presented above is about 2.5%. This extremely low diversity in matchup spreads conversely implies a pretty diverse and balanced metagame. In terms of metagame diversity, this format seems to be a pretty good one.

What does this say about the format? Pick up any deck and you have about the same chance of winning as with any other deck? Not necessarily, because these matchup spreads do not consider factors such as player skill, personal list engine choices, or techs. There is still a lot of variance to take into account when looking at this format; however, this is probably the most diverse metagame we have seen in a while. (Editor’s note: format quality can be determined by average skill level of top decks, or by parity of matchup spreads for top decks, not simply by diversity of metagame. As Christian notes, this format does display fairly strong parity of matchup spreads, although I would not say the skill cap to pilot most decks is particularly high. -Tate)


Proving that this meta is healthy
Most Pokémon players I have talked to throughout my (relatively short) career in the competitive scene claim to prefer a metagame that is diverse. I would agree--but what does “diverse” mean? In order to test whether the current Standard meta is diverse, we can construct a simple ANOVA table in order to conclude whether there is a significant difference between the matchup spreads of the current decks. This will allow us to see if any of the averages mentioned above are significantly different from each other, which would point to a unbalanced meta in favor of the deck with the higher average. My work is below (gotta prove I did it!). Feel free to skim and read my conclusion afterwards. I even whipped out my old statistics notes for this one. (Note: as mentioned earlier, DeciGoons is an outlier, and will not be used.)


What does this mean?

As stated above, most Pokémon players prefer a meta in which a variety of decks have about the same average chance to win any randomly assigned matchup. By using the simple ANOVA test above, the current meta is shown to have no significant differences among the matchup spreads of top decks. Considering the context of ADP in the format, statistically, we can see that we still have a “healthy” meta of five or more decks that all can succeed consistently. Note that it took until the ninth Hegster TCG tournament this season for one deck to win for a second time (Eternatus).

Now, we must test if a format without ADP would be better statistically. If it isn’t, we show that a ban is unnecessary.


Is a meta without ADP healthier?
We actually don’t have to completely improvise results from a format with ADP banned, because of the fairly large no-ADP tournament recently held that Nick Robinson won with Sandaconda V/Coalossal. After talking with multiple players, I came to the conclusion that because of Nick’s win and the deck’s theoretically solid matchup spread, Sandaconda could replace ADPZ in the “top tier” decks in this no-ADP format. The updated matchup spread table replacing ADPZ with Sandaconda is below:


Decks

Eternatus

Sandaconda

Blowns

Mew3

Centi

Pika

Luca w/ Zong

Inteleon

Eternatus

0.5

0.3

0.45

0.6

0.45

0.55

0.4

0.55

Sandaconda

0.7

0.5

0.35

0.35

0.5

0.7

0.35

0.6

Blowns

0.55

0.65

50.5

0.55

0.55

0.55

0.6

0.4

Mew 3

0.4

0.65

0.45

0.5

0.5

0.5

0.5

0.5

Centi

0.55

0.5

0.45

0.5

0.5

0.5

0.35

0.35

Pika

0.45

0.3

0.45

0.5

0.5

0.5

0.45

0.8

Luca w/ Zong

0.6

0.65

0.4

0.5

0.5

0.55

0.5

0.6

Inteleon

0.45

0.4

0.45

0.5

0.5

0.2

0.4

0.5

Table 2: Analysis of a Standard metagame with ADP banned (replacing ADPZ with Sandaconda V/Coalossal.)



What does this mean?

This table shows that the meta without ADP has just as much matchup parity as the meta with ADP, implying that the banning of ADP has no significant effect on the meta. The meta still gives eight “top” decks the opportunity to succeed, and any one of them could reasonably win an event. ADP itself is not the issue.

The issue with ADP is the linear strategy that it inherently brings to games where it is used. This strategy allows players (even unskilled ones) to achieve those ideal matchup win percentages with relative ease. Some other decks in this format, such as Eternatus and Blacephalon, share this trait. The metagame appears to be diverse with or without ADP, but the argument can be made that the quality of the game itself is slowly deteriorating as the strategies of more and more top-tier decks require less and less skill to achieve consistently.


A possible solution and final words
I think my analysis makes it clear that ADP does not hurt the diversity of the metagame at this time. Sure, it gatekeeps certain one-Prize decks from succeeding within the format, but on the other hand, we have a diverse set of “top” decks that can succeed at any event, which is extremely appealing.

Although I believe nothing needs to be banned at the moment, if even more metagame diversity is desired (such as allowing fringe decks such as Sandaconda to see more success), my other plausible solution would be simply issuing an errata for ADP. Making Altered Creation cost three Energy could be one these plausible fixes, allowing ADP to still be played, but essentially slowing it down by one turn, giving one-Prize decks a better chance to compete.

All in all, ADP isn’t toxic for the metagame, it just shows the game's progression towards higher-HP Pokémon over smaller Pokémon. If you decide to play in an online tournament right now, you should expect a fairly diverse meta and you can reasonably win the tournament by selecting any of the eight “top” decks from my tables above. I think that’s all we can ask for right now, especially after past formats where it seemed like only two or three decks could ever actually win.

Thanks for reading, and let me know if you have any questions or comments!

Christian

No comments:

Post a Comment