It's a well known fact in the marketing world that perception and reality are often different.

In my studies years ago in college, we learned that the perceived wait time (waiting in line at the bank, waiting at the bus stop, etc) averages higher than the actual wait time. Perceived wait time can be estimated with p = β0 + β1a + ε, where a is the actual wait time, p is the perceived wait time, β0 is the intercept of the regression, β1is the slope of the regression line, and ε is the error variable with a mean of 0.

In other words, we multiple actual wait time by a number, add a number to this result, and that gives us the perceived wait time.

In this case, perception is worse than reality.

Human beings remember the bad things much easier than the good. When we are confronted with an acquaintance that does something mean to us, we remember them for the mean. One good act doesn't seem to erase the mean one, whereas one mean act can erase the good one.

When confronted with several losses in a row in a game such as TFW, we associate those with the perceived OP cards. We don't remember our wins as much as our losses. (For example, it's EASY to think of a game that I've lost because my opponent drew two Nether Plasmas in the first five rounds, but hard to recall one that I've won when I drew two NP's in the first five rounds, though both have happened several times).

This begs the question-> are OP cards OP? Or do they seem that way because we remember them leading to losses more often than not?

So let's settle this with code.

Jed, can you write a script that compares the use of an OP card with wins vs losses?

I'd like to see a study that compares the use of Nether Plasma/Ix in the first five rounds vs games won. Is there a strong correlation between this card being played early in the game and winning?

Is there a strong correlation between a player having 3 of an "OP" card in his/her deck and that players win percentage. Does a player win 90% of the time when they play Blind Assassin?

Of course, we'd have to take player ratings into consideration, so perhaps limit the studies to similar player rankings?

It would also be interesting to see the win ratios of each domain vs each other domain against similar players.

_________________

Altren wrote:

I agree with Keyser, and that's what I'm planning to implement actually.

|| The Rise of the DCC | Plasmatium Netherious | Tastes Like Chicken | The Astridian Conspiracy ||

Guild -> | Platypus Rising|

I hope you take no offense to this...

Are you just trying to play devil's advocate to Np/Ix combo? Everyone (even players who use the combo) have stated that they believe it is OP'd. If there was a vote taken by the player base and players voted what they truely believed, nerfing would win in a tremendous landslide.

_________________

I dont think that any data we get from such a program will show us if a card is OP or not. If many good players chose not to play Ix and lots of not so good players chose to play it, it will show us that Ix is a worthless card since decks using it almost allways lose. I am convinced that Ix is the most powerfull card in the game right now, but I still havent been playing with it very much lately. I think we just have to use common sense on this one.

No, Themist, I am not referring to just IX/NP.

I'm referring to any and every card that is considered OP.

If you compare games among similarly ranked players and find, for example, that 90% of the time, if a player plays a construction site and then Dispel Flux, they win, then you can conclude that Dispel is OP.

If you find that 87% of the time, a player playing an NP within the first 5 rounds leads to a win against similarly ranked players, you'll assume it is OP. If it leads to a win 50% of the time, you should conclude that its not OP.

In other words, I'm wanting to see data that shows, or at least backs up, the premise that a card is OP.

This is not a post FOR or AGAINST the nerfing of any particular card.

It's a post asking to have actual statisstical game data back up the nerfing of a card.

_________________

Altren wrote:

I agree with Keyser, and that's what I'm planning to implement actually.

|| The Rise of the DCC | Plasmatium Netherious | Tastes Like Chicken | The Astridian Conspiracy ||

Guild -> | Platypus Rising|

While nerfing based on data is better in theory, I think collecting the data would mean a lot of extra work and due to the small data set and other variables i think there would be too much variance for the data to be very useful. Suppose that I've played 50% of the games that involved Yalrinian Fiend and i won 80% of them. then Yalrinian Fiend would have a win percentage of around 65%. If you take into account that the other players with the Fiend probably have more experience and better cards the win percentage could be more like 70%. A card like altar of ix which is widely available and widely played will probably have a win percentage much closer to 50%.

_________________
ȁ̎̉̈̂̎͋̈́̍n͂͑̿ͥͣ́̓ġͤ͑ͯe͌̓̊ͩ͗̅l̽̓ͬͬͥ̊͗aͣ̔̃ͧ̓̾̾t͊̑̆hͤ͆̓̉̌ͣ̚e̋ͩ͒̈́ĭ̌̒̎ͧ̓̋ͪs̋ͤͧ̅̇̋̎ͫ̂̾ͨ̽̄ͫͯt͒͐̌ͭ̂̃ͪ̽ͣ͗̒ͣ

As a research methods enthusiast, I would also question the validity and return on investment of the data you are proposing to gather.

I don't know that win vs. loss would be a meaningful metric. Frequency of card use might be interesting, but would at best only correlate with OP-ness, as you would expect players to play "good" cards more frequently, but not all good cards are OP.

In general, I think there are so many variables that I wouldn't feel confident making decisions based on this data, even if it did show some strong correlations.

So don't make the decision ONLY based such data. Rather, take it into consideration BEFORE nerfing a card but after that card has been brought up by players as being OP.

For example, if everyone says the The Auryn is OP, then you look at the data. If you find that 100% games in which the card is in Atreyu's deck, that player wins, then you can be relatively confident the card is OP. If, however, you find that Atreyu only wins 12% of the games in which his deck contains The Auryn, you could assume it's not really OP.

_________________

Altren wrote:

I agree with Keyser, and that's what I'm planning to implement actually.

|| The Rise of the DCC | Plasmatium Netherious | Tastes Like Chicken | The Astridian Conspiracy ||

Guild -> | Platypus Rising|

Point well made. Can I have that card?