Pokémon TCG: Sword and Shield—Brilliant Stars

Poke-metrics: Seeing the numbers behind the pretty cards.

Hi Pokegym,

I've developed a formula, and plugged it into excel to give a mostly objective value rating to all Pokemon cards. I say mostly objective, because some criteria, attack text, or pokepower/ability text for example needs to be read and interpreted subjectively, as it is not numerical and objective, like all other aspects of the cards.

The criteria that goes into the whole formula at different weights is as follows...

-HP
- Weakness, and which type
- Resistance, and which type
- Retreat Cost
- Stage of Evolution
- Attack Cost
- Attack text, including being able to inflict special condition(s).
- And pokepower/ability text

As you will see the type of the Pokemon itself is not listed, because I think it is subjective, or arbitrary to say one is "best, and similarly one is "worst". On a related note, the weakness and resistance formulas were altered by looking at the "what won battle road's posts", and making it so that the pokemon with resistance to the more successful pokemon get weighted slightly higher, and likewise with weakness and slightly lower,


I don't pretend that this going to change the game for anyone, or that it will even affect anyone's view of a single card. What it can be really useful for is when just looking at two cards and wondering if you should play X vs. Y. This formula can tell you that the difference in value of the cards may not be as close as the eye shows.

If anyone has any questions, or comments about this I would be ecstatic to hear them.

Here are three sample Poke-metric scores.

- Zekrom BW= 22.867

- Bearctic EP= 19.348

- Togekiss UD= 14.604


And before you ask, yes, of course I will plug any card anybody wants into the formula.

Thanks for reading.

THIS POST HAS RECEIVED GURU POINTS! ( click here to discover why ) - GuRu SQuaD
 
Last edited by a moderator:
Very cool. Could you possibly attach the excel file under password protection? (Alt + RPS or Alt + RPW)

I'm curious as to see how your algorithm rates similar cards with slight differences, especially the babies. Could you post the ratings for Cleffa HS, Elekid HS, Tyrogue HS, Magby TM, and Smoochum HS?

Additionally, how would your algorithm rate Suicune-Entei Legend, Raikou-Entei Legend, Raikou-Suicune Legend, Kyogre-Groudon Legend, as well as Palkia-Dialga Legend?

I ask this because rating Zekrom, Beartic, and Togekiss gives too much variation in the quality of the card. That's like ranking a lemon, apple, and watermelon on sweetness. To really test whether a system can pick up the sweetness of fruit, you would have to rank the sweetness of 3 fruits with similar sweetness (i.e.: watermelon, mango, and honeydew).


EDIT: I see that this is your first post on the 'Gym. Welcome!
 
Last edited:
haha... Sabre-Metrics FTW (and super confusing stuff sometimes). Can't wait to see where this goes.
 
Though this is a really cool feature, I don't think it will do much for the game in it's current state. Humans are unpredictable (or rather, their data is too hard to collect) and things like metagame will have to be considered when rating a specific card. It might be useful in some way if you develop it further, but because of the unpredictability of humans its uses will be limited. Still, I have to say you've done a good job doing this, and I would love to see it further enhanced so that it in some way can be used in competitive gameplay.
 
Way cool. I'd like to see how the program ranks cards which have a single overwhelming feature that they are used for. Like how would is rate Uxie (LA), Azelf (LA) which were played almost exclusively for their powers. How about Emboar Ability vs Badboar.
How about Ninetales and how it compares with Uxie?
 
Very cool. Could you possibly attach the excel file under password protection? (Alt + RPS or Alt + RPW)

I'm curious as to see how your algorithm rates similar cards with slight differences, especially the babies. Could you post the ratings for Cleffa HS, Elekid HS, Tyrogue HS, Magby TM, and Smoochum HS?

Additionally, how would your algorithm rate Suicune-Entei Legend, Raikou-Entei Legend, Raikou-Suicune Legend, Kyogre-Groudon Legend, as well as Palkia-Dialga Legend?

I ask this because rating Zekrom, Beartic, and Togekiss gives too much variation in the quality of the card. That's like ranking a lemon, apple, and watermelon on sweetness. To really test whether a system can pick up the sweetness of fruit, you would have to rank the sweetness of 3 fruits with similar sweetness (i.e.: watermelon, mango, and honeydew).


EDIT: I see that this is your first post on the 'Gym. Welcome!


I'll be the first to admit, that this is least effective with babies just because they are all essentially identical in the eyes of the formula besides the attack, which is just subjective in fitting your deck's needs anyways.

Legends will be much more diverse, so, here we go...

Suicune-Entei: 20.695

Raikou-Entei: 22.832

Raikou-Suicune: 22.352

Kyogre-Groudon: 21.798

Palkia & Dialga Legend: 19.034


will edit this post with other's requests.


WOw, pretty cool. Can you plug Reshiram and Revenge Bouffalant into it please?

Reshiram= 21.982

Bouffalant= 15.982 Bouffalant is a card that is getting a lot of hype that my formula would actually say is not deserving of such hype.



Should rate Donphan.

Donphan= 21.752



should do reshiram and rdl emboar attack

Resh is already done...


Emboar= 20.629


Way cool. I'd like to see how the program ranks cards which have a single overwhelming feature that they are used for. Like how would is rate Uxie (LA), Azelf (LA) which were played almost exclusively for their powers. How about Emboar Ability vs Badboar.
How about Ninetales and how it compares with Uxie?


Nice post...


Uxie: 19.145

Azelf: 16.949

Ninetales: 16.878
 
Last edited:
This is interesting. But, from a data analysis standpoint, I don't really read the relevance of the values.

There's no interpretive information given. It's just... numbers at this point.
 
^
I guess that's fair, like what is "the max score" what cards from the past have scored the highest, maybe that'd be cool. although i will admit i like numbers....
 
So far all you've done is list decent-to-playable cards. What about something severely underwhelming like Dialga CoL or some random binder common?
 
Could you send me the Excel spreadsheet please?

And also, I'd like to see Yanmega Prime.

I'm building a new rogue deck using nothing but this formula for my battle roads this weekend, if it goes up in flames, i'll release it then.

Yanmega Prime: 18.992


Magnezone Prime?
Kingdra Prime?
Pachirisu CoL?
Shaymin UL?

Magnezone Prime: 23.087

Kingdra Prime: 20.526

Pachirisu COL: 14.865

Shaymin UL: 18.348


So far all you've done is list decent-to-playable cards. What about something severely underwhelming like Dialga CoL or some random binder common?

Dialga COL: 13.285 (an incredibly low rating)

Raticate UD: 15.258


Cinccino! BW, do the wave one
Zoroark (Foul Play)

Just for fun: Basculin (suicide fish)

Cinccino: 16.748

Will not do Zoroark, his value is directly based on what your opponent is playing.

Basculin: 9.294 (Horrible...)

Cubone: 12.350
 
Last edited:
What would be really amazing is if you some how modified the formula to put an entire deck list in
Posted with Mobile style...
 
What would be really amazing is if you some how modified the formula to put an entire deck list in
Posted with Mobile style...

I agree, however, I just don't think Microsoft Excel is capable of being programmed with the requisite amount of AI needed to see and understand the synergies between each and every card.

I'm sure a program could be made to do such a thing, but it's not Excel, and I certainly am no program designer.
 
Back
Top