The Underdog Bonus (aka Unpopularity Bonus)

[4RM3D]

In the eternal fight against netdecks, I cometh with another suggestion to combat it: the Underdog Bonus.

Every deck (in the deck builder) gets a new stat called underdog status. This stat is a percentage based on the amount of unpopular cards you have in your deck. The more unpopular cards you have, the higher the percentage is and vice versa. When you play a netdeck the percentage is (close to) 0. This percentage is the bonus you get when playing games. This bonus applies to both your experience and your MMR (rating). Let's say that you have an underdog status of 25%. This means you get 25% more experience and MMR.

The exact formula needs to be worked out. But it goes something like this for every card:

Spoiler: Mathematical stuff

CARDS: (55% - P) * (1 + (1 * (100% - P))) = bonus %
LEADER: (35% - P) * (1 + (1 * (100% - P))) = bonus %

where P = percentage of players using this card.

Underdog bonus is the average of all bonus percentages from the cards and leader.

NOTE: the bonus cannot drop below zero!

The idea behind this formula is that you'll get an exponentially bigger bonus for unpopular cards. For example:
A card with a popularity of 10% gives a 85.5% bonus.
A card with a popularity of 40% gives a 24.0% bonus.
A card with a popularity of 50% gives a 7.5% bonus.

All cards above 55% play rate give a penalty. For example:
A card with a popularity of 60% gives a 7% penalty.
A card with a popularity of 70% gives a 19.5% penalty.
A card with a popularity of 90% gives a 38.5% penalty.

However, the penalty for popular cards is still less that the bonus for unpopular cards. As shown in the example above, you get a 85.5% bonus for 10% played, while only getting a 38.5% penalty for 90% played. This means that you can include both and still get a bonus on average.

 

[FG15-ISH7EG]

My first question is, is that something that is really desirable?
Getting to a higher rank due to the underdog bonus still doesn't mean that you will win more often, just that your wins are more valueable. That would mean that in the end you will loose a lot more playing an underdog deck, because you are higher ranked than the actual strength of the deck. And that can be really frustrating. Also with more "bad" decks around at higher ranks, strong netdecks will be even more likely to win, which will make those more appealing to play.

Besides that problem, I like the idea. Though, what does "P = percentage of players using this card" exactly mean? Is the overall playerbase counted or only decks of the same faction, or even leader? And what does using mean? Is it enough if I created the deck, or do I have to play at least 1 ranked match with it, or do the number of ranked matches count too?

Furthermore, those values are pretty high. Back in Beta when we had regular Gwentup Metareports, having a leader at over 25% percent was extremly rare (Brouver Elf Swarm) and strong leaders were those that had more than 15% popularity at high MMR.
The same with cards. Even silver spies, which were the most autoinclude cards in the game only had an inclusion rate of 88% at the end of Open Beta.

 

[4RM3D]

Also with more "bad" decks around at higher ranks, strong netdecks will be even more likely to win, which will make those more appealing to play.

Good point. However, there is still a difference between throwing in all the "junk" cards and netdecking 101. There is a middle road and having a small motivator in the shape of a bonus might be enough to switch around some cards.

Though, what does "P = percentage of players using this card" exactly mean?

That also needs to be determined. I guess it would be the play rate of the card, within the faction, measured over a period of a week, where, at the start of each new week, the popularity is reevaluated. The play rate should be based on an average of all matches. So, playing a card once, doesn't count as being played. It needs a more intricate formula, but that's beyond the scope of the suggestion. Though, I will mention that it's possible for a neutral card to have a different underdog rating based on the faction its being played in.

Furthermore, those values are pretty high.

Like I have said, the values might need to be tweaked a bit, but the principal remains.