News Shop
Events Forums

Ranking system for the individual specs


#1

If we are going to give the individual specs a rating or ranking, what would be the best way of doing so? We obviously can’t do it on a 1 vs. 1 basis as the different specs other than the neutrals weren’t designed to function that way. What I propose is giving the neutral specs a score of 1.0 each. They will serve as the baseline. As the other specs were more or less designed to function in a group of 3, I propose that in an effort to score the relative strength of the other specs they should be teamed up with the 2 neutral specs. I realize that this isn’t entirely fair as some specs might have more synergy with the neutral specs than others, but what better way is there?

Matchups should then be made between the different specs with their matching starters teamed up with the two neutral specs against one another. I imagine with 18 specs to test and rank this way that it would take a very, very long time. I doubt this will ever get done, but it would definitely be cool to see how the different specs fare against one another. As each spec would be matched up with 17 other specs (neutrals excluded) theoretical scores would range from 17-0 to 0-17, which would translate to 2.7 to -0.7.

These scores could be established by playing a best out of 3, with the winner of the 3 getting the win and the loser getting the loss for the matchup. Each matchup victory would count as +0.1, and loss as a -0.1. These victories and losses would then be totaled with the base score of 1.0 thus resulting in a score ranging fro -0.7 to 2.7.

In the case of a tied score at the end, the higher ranking would then go to the spec that won the head to head matchup between the two specs. In the instance where there are multiple ties and no clear cut winner can be established, they would all share a tied ranking.

I’m fairly confident that this will never get done and is merely a smoke dream. Nevertheless, I’m still wondering what else could theoretically be done to rank the specs. I’d like to read what others think, please feel free to comment.


Codex data thread
#2

You can’t. It is like trying to give a ranking to Rock, Paper and Scissors.


#3

I don’t think it is like RSP. With RSP each thing is more or less identical except for the fact that it loses to one, wins against another and ties with itself. With the specs, there are a lot of differences that should make comparisons possible and interesting to analyze. I think it is more comparable to thinking about the different specs as if they were different teams in some league.


#4

I’d argue it would be less a matter of ranking them on a “best to worst” scale and more a spectrum with X and Y axes as “generalist vs. specialist” and “straightforward to play vs. complex to play” given that they’re not really isolatable in a meaningful sense.

It’d be like ranking everyone’s c.MP in Street Fighter 5 or something.


#5

Or if it was 1 vs 1, maybe you could do a weak against/strong against chart.

List all specs on both x/y axis and then map out comparison between them…

Like Pokemon types! Blood type is super effective against Growth type, but weak against Strength type…


#6

Or original premise is flawed because Finesse is one of the top 10 specs, and Bashing is the worst, so you can’t give them the same rating.


#7

I think it’s important to understand what the purpose of “rating” a thing even is. You don’t play Codex with a single spec, so it’s not particularly valuable to give single specs an “overall” score. It would be pretty valuable to do a write up on specs, which discusses different things that you would consider when building a Codex with them: what’s the Hero and their spells good at, what Tech I options does it give you, what strategy does their Tech II options allow/force you to pursue, what well does their Tech III support that/how important is their Tech III.


#8

Thank you for pointing that out the two neutral specs aren’t equal. I’m confident that @EricF has a much better read of the game and that what he say is most likely true; however, even if the neutral specs aren’t equal, I think the ranking/rating system could still work for the other 18 specs because of the fact that all the specs would be paired up with them.

I imagine once all the spec ratings have been determined then you could take the two specs that are closest to 1.0 and substituting them as the new “neutrals” repeating the process with two original neutral specs against the remaining specs and one another.


#9

This would just give you a rating about the specs that work best with both neutral. These specs might worth 0 with the 2 neutral but be the most OP shit with anything else.


#10

@Wildhorn is spot on with his assessment that the context of the ratings would indeed be based relative to the neutral specs and that the ratings could be way off when the specs are combined with other specs, but nonetheless by having the ratings system wouldn’t there now exist a baseline for future comparisons.

Even still, perhaps ranking the individual specs is not as valuable as rating different codices as the base units. I suppose if this were to be done, this seems like it would be a more feasible thing to try to measure as there would only be 6 different codices to test to establish a baseline.


#11

This would be interesting, but how would you determine such things? Would it be purely subjective, one person’s opinion?


#12

I would say the best way to do what you want to do is simply to aggregate tournament results, as EricF has helpfully done for the tournaments he has run. Looking at the overall average win rate of each spec against the field will give you an idea of how well each spec does in general. There are provisos you need to take into account, e.g. Bashing did extremely well (even though it is generally regarded as one of the weaker specs) in the last tournament, but that was because only one person was playing it, and that person is an extremely good player. However, these issues should be overcome as we collect more data, as long as all specs see a reasonable amount of play.


#13

Yeah, I’m thinking something like one of those tier list graph makers. Entirely subjective.


#14

I agree that such a ranking approach is probably not going to lead to anything useful. There are just too many interconnected variables at play that separating them is likely impossible. Probably we could analyze mono-color matchups and maybe some of the more popular multicolors as a whole, but even that task seems daunting.

I would love to see something like mysticjuicer’s suggestion:

For example, some specs have such good Tech I units they alone nearly make an argument for including the spec.


#15

I prefer to rank out all the individual cards / groups. Then you can puck your specs based on which cards you want access to, or pick 2 specs for some plan, then look at their weak points, and pick a third spec that complements it.

Giving a rating for a spec as a whole is only useful at a very rough level, since it’s very common to use a spec extensively, and never use the Tech 2/3 from it (looking at you, Demonology). Or, conversely, to never use the Tech 1 cards, since your other specs have better (see: Future).

The raw data are still too sparse to indicate anything, but in anotger year, maybe we can draw some conclusions.

Mono Green does really poorly, though, even accounting for player skill.


#16

That’s another thing. Players skill vary, so a good player with a bad deck could outperform a bad player with a good deck.

Also, I don’t see the point of trying to rate the specs. Play something you like and master it like nobody.

Like people says Bashing is the weakest, yet I perform damn well with it.


#17

I don’t get why you say monogreen perform poorly. When I play it I perform very well.

Create a wall of creatures with Growth and Argagarg then Circle of Life with Midori to bring out the good stuff or just Stampede after Stampede since Argagarg seems to never want to die.


#18

Eric is probably specifically referring to his experience from the tournaments/games run on the site.


#19

rabid’s rage-o-meter just counts the cards that make me angry:

Balance: 6 rage inducing cards out of 13
Feral: 6 rage inducing cards out of 13
Growth: 6 rage inducing cards out of 13

Anarchy: 3 rage inducing cards out of 13
Blood: 5 rage inducing cards out of 13
Fire: 8 rage inducing cards out of 13

Bashing: 9 rage inducing cards out of 13
Finesse: 6 rage inducting cards out of 13

Demonology: 8 rage inducing cards out of 13
Disease: 6 rage inducing cards out of 13
Necromancy: 5 rage inducing cards out of 13

Law: 8 rage inducing cards out of 13
Peace: 5 rage inducing cards out of 13
Truth: 6 rage inducing cards out of 13

Discipline: 8 rage inducing cards out of 13
Ninjitsu: 7 rage inducing cards out of 13
Strength: 8 rage inducing cards out of 13

Past: 11 rage inducing cards out of 13
Present: 6 rage inducing cards out of 13
Future: 6 rage inducing cards out of 13

of course, yinz probably want to rank things by some other metric


#20

[Past]/Bashing/(any of:Strength/Discipline/Law/Demonology/Fire) hereby named Rabies