Competitive meter

TappedOut forum

Posted on Jan. 3, 2024, 3:37 p.m. by FatFreddiesCat

It seems to me that every one of my decks shows up on the competitive meter as 100% Casual and 0% Competitive. While a number of my decks are certainly for fun, it seems that some should be more highly rated than zero. Do I have some setting wrong in my account setup?

sergiodelrio says... #2

That metric is disfunctional alltogether and should be ignored

January 3, 2024 4:04 p.m.

Crow_Umbra says... #3

There have been a handful of similar forum questions regarding the competitive meter over the past few months, to maybe a year. You haven't done anything wrong, as this seems to be a bug that's affecting most decks on the site.

January 3, 2024 4:15 p.m.

FatFreddiesCat says... #4

Thanks, I guess. I liked the function, even though I realized it could never be 100% accurate.

January 4, 2024 8:10 a.m.

Crow_Umbra says... #5

If you liked that feature FatFreddiesCat, I'd recommend checking out commandersalt.com, if you haven't already. I forget their username, but one of the site developers posted on here sometime in the past year asking for feedback on the site.

Their FAQ goes a bit more in depth about the various factors to "score" decks, but a few factors include:

  • "Saltiness" of cards in deck, based on Salt scores from EDHREC.

  • Relative quantity of Ramp, Removal/Interaction, Combos, and the overall speed/efficiency of these factors.

  • Overall "synergy" of a deck, based off reading of Scryfall/oracle text.

The site gives each deck various scores, including an estimated 1-10 Power level rating based off the various factors listed above.

It's not perfect by any means, but has improved since some of the initial feedback gathering last year. The scoring isn't really consistent across deck-building sites that it pulls its data from either. A deck pulled from TappedOut could be rated a 6.7/10, for example, and the exact same deck list pulled from another deck building site could be rated slightly higher or lower. Card swaps can also have weirdly different effects on the overall deck rating. A card swap could improve the overall score for a deck from Moxfield or Archidekt, but dramatically drop the score for the exact same list pulled from TappedOut, or vice versa.

Whether it's the TappedOut competitive meter or the CommanderSalt scoring metrics, take it all with a huge grain of salt. It can be fun to try to get a rough quantified estimate of how your deck stacks up, but a lot of this is still fairly subjective.

January 4, 2024 11:51 a.m.

sergiodelrio says... #6

Soon we'll plug our decks into the Quantum AI Overlord which will then rate it in 2 seconds after simulating millions of matches vs all other decks

January 4, 2024 6:20 p.m. Edited.

Crow_Umbra says... #7

Honestly sergiodelrio, it's only a matter of time before someone tries to add AI to one of these deck rating tools lol. With how much more widely AI was made publicly accessible last year, I'm a little surprised someone hasn't tried applying it to one of these rating tools. Then again, web/app development is very far out of my professional wheelhouse lol.

"Your deck was determined to be a 4.7/10. Perish."

January 4, 2024 6:41 p.m.

grumbledore says... #8

I thought about using machine learning (ai) for commandersalt but I wanted repeatable reliable results. Plus its just not my strength lol.

Crow_Umbra - I got your note about the import issues... Sorry for not replying sooner! I hope to tackle it soon.

January 4, 2024 6:46 p.m.

Crow_Umbra says... #9

Hey grumbledore, there you are! No worries, totally understandable.

My last job was in study facilitation for healthcare research pertaining to the use of patient-facing apps. The main study I was on utilized machine learning in its app. The machine learning focused meetings with the app developer were really technical at times. It def seems complicated.

January 4, 2024 7:06 p.m.

grumbledore says... #10

for sure. cost prohibitive as well at scale. im at about $130 - $200 a month now out of pocket for this lol. ai would pump it up a bit

January 4, 2024 7:34 p.m.

yeaGO says... #11

We will have a new model out soon and will review your decks. It could just be we didn't have much data about them in the last model

January 5, 2024 6:04 p.m.

yeaGO says... #12

Please save and review your scores and let me know if things seem improved or not

January 9, 2024 2:48 a.m.

Crow_Umbra says... #13

yeaGO, I've tried saving and reloading a couple of my decks, but haven't really noticed any changes. I recently brewed and published a Lord Windgrace deck, and that's the only one of my decks that currently doesn't have a meter that's experiencing the issues that have been described here.

January 9, 2024 11:38 a.m.

yeaGO says... #14

Can you link decks that you want me to review

January 9, 2024 2:34 p.m.

Crow_Umbra says... #15

Here are a couple:

At first, I thought it was just the decks that were visible on my profile page that were affected, but some of my older stuff that I haven't edited in months to years also seem to be impacted. No need to look at all of that. I think just these 2 that I linked should be fine.

January 9, 2024 2:45 p.m.

sergiodelrio says... #16

Pretty sure that happens when the deck is actually more than 100% competitive, so the graphic is off the screen...

I will see myself out now :D

January 9, 2024 7:09 p.m.

yeaGO says... #17

The meter should by fixed. Sorry for the delay

January 29, 2024 1:59 a.m.

Please login to comment