CDPR, Pls explain why i have 56% of blue Coins via 1k games?

[Void_Singer]

[..]Who cares about math, [...]

You do, since you're complaining about it....

and while we're at it, since you got 56% that means logically someone else got 44% (or a bunch of people got slightly higher, but w/e)... you think they mind?

100-150 games to get perfect 50/50 for this game to be fair, that's why they have to implement fair coinflip.

the only way for there to ever be your definition of "fair" is for there to be no coinflip at all. and it doesn't address the real problem, which is that there is no real first mover advantage in this game. other games that use continuous draw make up for this by let the person that goes first, draw first... but those games don't have pass to preserve your hand mechanics.

Even if it is statistically not that impossible to get such a bad ratio, it shouldn't happen, because it massively effects the gameplay.
As important as coinflip is currently, a ratio as fair as possible for everyone should be included, which can be solved by pseudo random numbers.

it's not that it's a bad ratio... it's just a random one... and if first mover in round one wasn't so hamstung, you could get 100% of anything and it wouldn't matter. BTW I guarantee you they are using psuedo random numbers.... it's not a solution, because it assumes the same generator follows each person around... but then you'd just be competing with the other persons generator, and that would break the "evenness" of it too.

first mover vs reaction balance is what needs fixing.

 

[Mathspy]

BTW I guarantee you they are using psuedo random numbers.... it's not a solution, because it assumes the same generator follows each person around... but then you'd just be competing with the other persons generator, and that would break the "evenness" of it too.

You're technically correct since in computer science all number generators are pesudo anyway
But you're assuming how they work incorrectly at all; with pseudo RNG, everyone could at least be guaranteed 49%~51% coinflip

 

[Void_Singer]

no they couldn't because the cycle isn't short enough

 

[Dante2144]

What about rock scissors paper to see who goes first?

 

[partci]

5 out of 5 Blue Coins this morning. GG 50/50.

 

[NomanPeopled]

There's really two topics here that shouldn't be confused.

Item one is whether the generator actually works as expected. With the numbers at hand, it can't be shown that it does not. Standard deviation and all that.

Item two is whether there should instead be a generator that artificially decreases deviation to make sure that everybody actually gets the expected value consistently.

With a perfect 50/50 generator, it is inevitable that some players will have high deviations, particularly in small sample sizes.
That's fair in that everybody gets an equal chance every time, but it's unfair in that some people will be at an advantage or disadvantage while waiting for a regression to the mean.

I think it's a good idea to "correct for" deviation.

 

[Theo_FP]

You have a 50% to win the coin flip and you are complaining about 6%? It is perfectly fine. Be glad that you are that close to 50. If your win-rate is predicated on winning the coin flip, then you should just learn to play the game better. It is not the end of the world.

Also, I do agree that something has to be done about the coin flip.

 

[Mathspy]

Since I was feeling bored


Where total(Matches) and coinsLost are calculated based on this season's already played games. If the formula's result is ANY number above 50, player wins coinflip

How it'd work:
>This formula (or anything like it) would trigger on player attempting to find a ranked match
>The server will attempt its best to connect the closest two MMRs that got opposite results from the last step
>If the server fails or finds it'd approximately take more than X time, it abandons the plan and just connects to someone with a close enough MMR (even if it abandons the plan the match still contributes to total and coinsLost)

Showcase:
Here is a set of 1000 random numbers skewed linearly towards "losing" (less than or equal to 50) the coinflip at around 56% (to represent the OP)

Spoiler

12, 85, 38, 55, 30, 65, 20, 38, 84, 53, 61, 7, 77, 57, 50, 80, 72, 14, 39, 12, 55, 7, 11, 31, 83, 22, 9, 70, 32, 67, 49, 30, 18, 37, 7, 59, 39, 24, 53, 21, 18, 53, 59, 57, 73, 16, 35, 24, 43, 8, 37, 10, 27, 82, 81, 42, 83, 19, 34, 23, 23, 71, 42, 83, 79, 4, 14, 28, 31, 76, 6, 54, 78, 36, 36, 33, 11, 12, 43, 56, 81, 49, 42, 20, 5, 45, 26, 87, 64, 87, 41, 70, 52, 72, 39, 61, 9, 80, 34, 7, 25, 17, 77, 68, 80, 41, 29, 52, 81, 23, 29, 31, 62, 25, 38, 26, 84, 10, 14, 4, 29, 21, 68, 8, 69, 12, 75, 64, 7, 44, 52, 34, 56, 85, 3, 39, 50, 48, 78, 84, 83, 19, 67, 67, 76, 85, 37, 59, 62, 31, 62, 7, 5, 6, 80, 79, 79, 73, 32, 39, 62, 34, 77, 77, 78, 4, 7, 51, 67, 65, 76, 49, 29, 58, 31, 17, 48, 11, 71, 33, 28, 64, 18, 72, 82, 1, 41, 82, 65, 57, 44, 35, 78, 23, 9, 2, 82, 55, 4, 57, 27, 11, 82, 5, 11, 64, 67, 13, 53, 12, 13, 61, 6, 83, 60, 74, 42, 73, 34, 24, 58, 45, 55, 64, 86, 34, 61, 78, 70, 65, 5, 37, 47, 10, 78, 88, 57, 37, 29, 30, 50, 23, 37, 56, 55, 23, 30, 88, 13, 55, 59, 29, 44, 48, 40, 70, 1, 55, 71, 40, 6, 81, 44, 28, 88, 36, 72, 53, 78, 72, 80, 58, 60, 88, 31, 71, 26, 27, 58, 2, 18, 66, 32, 34, 77, 17, 30, 56, 45, 35, 38, 10, 4, 38, 83, 45, 13, 20, 72, 17, 2, 1, 8, 72, 72, 62, 3, 70, 56, 20, 6, 53, 32, 21, 88, 74, 74, 20, 84, 32, 24, 80, 88, 60, 41, 44, 18, 23, 40, 45, 62, 54, 84, 66, 64, 22, 68, 44, 19, 22, 13, 14, 60, 55, 8, 26, 80, 49, 16, 84, 65, 48, 53, 74, 35, 25, 78, 16, 85, 78, 7, 74, 21, 62, 73, 85, 42, 71, 25, 45, 37, 37, 38, 72, 86, 75, 76, 32, 69, 62, 38, 32, 26, 16, 1, 6, 17, 2, 48, 42, 15, 4, 14, 83, 32, 25, 46, 3, 61, 70, 81, 57, 32, 63, 65, 17, 56, 55, 13, 33, 28, 86, 61, 22, 3, 16, 11, 19, 75, 25, 67, 49, 29, 77, 74, 75, 42, 40, 87, 47, 7, 76, 70, 44, 3, 60, 23, 11, 20, 79, 9, 52, 50, 55, 7, 36, 83, 38, 11, 9, 3, 30, 81, 18, 1, 68, 63, 4, 78, 39, 6, 14, 44, 75, 37, 44, 55, 14, 76, 71, 63, 63, 71, 1, 4, 73, 32, 5, 16, 83, 16, 41, 30, 35, 35, 76, 88, 75, 80, 35, 34, 71, 64, 25, 58, 20, 12, 58, 47, 71, 4, 63, 60, 17, 83, 84, 65, 42, 73, 54, 49, 76, 80, 20, 1, 72, 43, 31, 51, 66, 24, 83, 52, 45, 87, 14, 26, 76, 15, 69, 24, 78, 12, 81, 71, 61, 8, 9, 12, 15, 26, 51, 5, 51, 11, 26, 34, 11, 56, 85, 22, 50, 20, 14, 54, 80, 60, 52, 86, 20, 62, 51, 30, 29, 48, 5, 68, 17, 62, 18, 2, 46, 71, 45, 53, 27, 68, 19, 25, 38, 29, 37, 13, 5, 54, 21, 29, 5, 62, 77, 37, 35, 9, 17, 32, 84, 45, 69, 44, 4, 24, 75, 52, 78, 72, 30, 24, 19, 26, 55, 64, 83, 26, 6, 42, 56, 46, 64, 24, 39, 80, 70, 86, 32, 72, 56, 65, 45, 80, 38, 43, 23, 63, 52, 59, 70, 5, 13, 17, 43, 9, 55, 69, 37, 8, 81, 22, 65, 32, 50, 75, 70, 3, 31, 82, 55, 77, 9, 55, 69, 33, 50, 54, 87, 1, 72, 50, 81, 71, 46, 37, 19, 62, 85, 72, 73, 8, 40, 46, 32, 55, 55, 39, 29, 19, 65, 84, 15, 84, 37, 21, 61, 41, 48, 10, 58, 1, 11, 85, 80, 67, 74, 76, 40, 34, 31, 77, 65, 68, 56, 51, 14, 63, 36, 49, 64, 16, 30, 1, 5, 21, 21, 56, 8, 1, 35, 25, 47, 75, 77, 40, 22, 78, 41, 48, 17, 37, 7, 81, 25, 48, 76, 64, 30, 6, 76, 36, 75, 86, 37, 34, 4, 65, 63, 9, 77, 87, 49, 49, 51, 52, 18, 57, 20, 80, 69, 77, 22, 15, 66, 23, 68, 13, 18, 19, 6, 10, 71, 1, 53, 31, 58, 30, 26, 23, 75, 70, 4, 1, 42, 71, 26, 78, 16, 62, 19, 10, 2, 55, 71, 6, 7, 19, 39, 68, 43, 55, 13, 36, 22, 8, 72, 31, 4, 75, 56, 75, 70, 32, 52, 41, 73, 40, 43, 22, 12, 55, 47, 51, 84, 86, 29, 29, 38, 46, 53, 44, 49, 17, 74, 70, 72, 10, 11, 62, 5, 28, 59, 76, 15, 17, 46, 53, 28, 56, 48, 7, 61, 5, 18, 46, 86, 21, 50, 86, 70, 50, 17, 72, 2, 16, 6, 6, 61, 78, 5, 72, 73, 46, 50, 3, 43, 78, 18, 72, 34, 77, 75, 30, 33, 43, 74, 20, 27, 11, 1, 70, 66, 73, 48, 53, 1, 15, 44, 51, 72, 52, 46, 64, 28, 79, 6, 54, 25, 5, 53, 61, 47, 81, 37, 74, 70, 33, 1, 32, 35, 71, 72, 74, 20, 79, 61, 79, 2, 39, 45, 8, 86, 15, 72, 72, 77, 62, 22, 41, 69, 29, 59, 40, 29, 47, 27, 47, 16, 2, 16, 83, 42, 24, 8, 40, 69, 80, 84, 15, 43, 53, 44, 7, 44, 81, 47, 49, 8, 55, 82, 81, 69, 14, 15, 3, 64, 67, 56, 9, 16, 17, 1, 22, 9, 25, 59, 9, 53, 83, 25, 5, 75, 67, 84

This simple JavaScript code can be used to test them without the formula:

Spoiler

let randoms = [12, 85, 38, 55, 30, 65, 20, 38, 84, 53, 61, 7, 77, 57, 50, 80, 72, 14, 39, 12, 55, 7, 11, 31, 83, 22, 9, 70, 32, 67, 49, 30, 18, 37, 7, 59, 39, 24, 53, 21, 18, 53, 59, 57, 73, 16, 35, 24, 43, 8, 37, 10, 27, 82, 81, 42, 83, 19, 34, 23, 23, 71, 42, 83, 79, 4, 14, 28, 31, 76, 6, 54, 78, 36, 36, 33, 11, 12, 43, 56, 81, 49, 42, 20, 5, 45, 26, 87, 64, 87, 41, 70, 52, 72, 39, 61, 9, 80, 34, 7, 25, 17, 77, 68, 80, 41, 29, 52, 81, 23, 29, 31, 62, 25, 38, 26, 84, 10, 14, 4, 29, 21, 68, 8, 69, 12, 75, 64, 7, 44, 52, 34, 56, 85, 3, 39, 50, 48, 78, 84, 83, 19, 67, 67, 76, 85, 37, 59, 62, 31, 62, 7, 5, 6, 80, 79, 79, 73, 32, 39, 62, 34, 77, 77, 78, 4, 7, 51, 67, 65, 76, 49, 29, 58, 31, 17, 48, 11, 71, 33, 28, 64, 18, 72, 82, 1, 41, 82, 65, 57, 44, 35, 78, 23, 9, 2, 82, 55, 4, 57, 27, 11, 82, 5, 11, 64, 67, 13, 53, 12, 13, 61, 6, 83, 60, 74, 42, 73, 34, 24, 58, 45, 55, 64, 86, 34, 61, 78, 70, 65, 5, 37, 47, 10, 78, 88, 57, 37, 29, 30, 50, 23, 37, 56, 55, 23, 30, 88, 13, 55, 59, 29, 44, 48, 40, 70, 1, 55, 71, 40, 6, 81, 44, 28, 88, 36, 72, 53, 78, 72, 80, 58, 60, 88, 31, 71, 26, 27, 58, 2, 18, 66, 32, 34, 77, 17, 30, 56, 45, 35, 38, 10, 4, 38, 83, 45, 13, 20, 72, 17, 2, 1, 8, 72, 72, 62, 3, 70, 56, 20, 6, 53, 32, 21, 88, 74, 74, 20, 84, 32, 24, 80, 88, 60, 41, 44, 18, 23, 40, 45, 62, 54, 84, 66, 64, 22, 68, 44, 19, 22, 13, 14, 60, 55, 8, 26, 80, 49, 16, 84, 65, 48, 53, 74, 35, 25, 78, 16, 85, 78, 7, 74, 21, 62, 73, 85, 42, 71, 25, 45, 37, 37, 38, 72, 86, 75, 76, 32, 69, 62, 38, 32, 26, 16, 1, 6, 17, 2, 48, 42, 15, 4, 14, 83, 32, 25, 46, 3, 61, 70, 81, 57, 32, 63, 65, 17, 56, 55, 13, 33, 28, 86, 61, 22, 3, 16, 11, 19, 75, 25, 67, 49, 29, 77, 74, 75, 42, 40, 87, 47, 7, 76, 70, 44, 3, 60, 23, 11, 20, 79, 9, 52, 50, 55, 7, 36, 83, 38, 11, 9, 3, 30, 81, 18, 1, 68, 63, 4, 78, 39, 6, 14, 44, 75, 37, 44, 55, 14, 76, 71, 63, 63, 71, 1, 4, 73, 32, 5, 16, 83, 16, 41, 30, 35, 35, 76, 88, 75, 80, 35, 34, 71, 64, 25, 58, 20, 12, 58, 47, 71, 4, 63, 60, 17, 83, 84, 65, 42, 73, 54, 49, 76, 80, 20, 1, 72, 43, 31, 51, 66, 24, 83, 52, 45, 87, 14, 26, 76, 15, 69, 24, 78, 12, 81, 71, 61, 8, 9, 12, 15, 26, 51, 5, 51, 11, 26, 34, 11, 56, 85, 22, 50, 20, 14, 54, 80, 60, 52, 86, 20, 62, 51, 30, 29, 48, 5, 68, 17, 62, 18, 2, 46, 71, 45, 53, 27, 68, 19, 25, 38, 29, 37, 13, 5, 54, 21, 29, 5, 62, 77, 37, 35, 9, 17, 32, 84, 45, 69, 44, 4, 24, 75, 52, 78, 72, 30, 24, 19, 26, 55, 64, 83, 26, 6, 42, 56, 46, 64, 24, 39, 80, 70, 86, 32, 72, 56, 65, 45, 80, 38, 43, 23, 63, 52, 59, 70, 5, 13, 17, 43, 9, 55, 69, 37, 8, 81, 22, 65, 32, 50, 75, 70, 3, 31, 82, 55, 77, 9, 55, 69, 33, 50, 54, 87, 1, 72, 50, 81, 71, 46, 37, 19, 62, 85, 72, 73, 8, 40, 46, 32, 55, 55, 39, 29, 19, 65, 84, 15, 84, 37, 21, 61, 41, 48, 10, 58, 1, 11, 85, 80, 67, 74, 76, 40, 34, 31, 77, 65, 68, 56, 51, 14, 63, 36, 49, 64, 16, 30, 1, 5, 21, 21, 56, 8, 1, 35, 25, 47, 75, 77, 40, 22, 78, 41, 48, 17, 37, 7, 81, 25, 48, 76, 64, 30, 6, 76, 36, 75, 86, 37, 34, 4, 65, 63, 9, 77, 87, 49, 49, 51, 52, 18, 57, 20, 80, 69, 77, 22, 15, 66, 23, 68, 13, 18, 19, 6, 10, 71, 1, 53, 31, 58, 30, 26, 23, 75, 70, 4, 1, 42, 71, 26, 78, 16, 62, 19, 10, 2, 55, 71, 6, 7, 19, 39, 68, 43, 55, 13, 36, 22, 8, 72, 31, 4, 75, 56, 75, 70, 32, 52, 41, 73, 40, 43, 22, 12, 55, 47, 51, 84, 86, 29, 29, 38, 46, 53, 44, 49, 17, 74, 70, 72, 10, 11, 62, 5, 28, 59, 76, 15, 17, 46, 53, 28, 56, 48, 7, 61, 5, 18, 46, 86, 21, 50, 86, 70, 50, 17, 72, 2, 16, 6, 6, 61, 78, 5, 72, 73, 46, 50, 3, 43, 78, 18, 72, 34, 77, 75, 30, 33, 43, 74, 20, 27, 11, 1, 70, 66, 73, 48, 53, 1, 15, 44, 51, 72, 52, 46, 64, 28, 79, 6, 54, 25, 5, 53, 61, 47, 81, 37, 74, 70, 33, 1, 32, 35, 71, 72, 74, 20, 79, 61, 79, 2, 39, 45, 8, 86, 15, 72, 72, 77, 62, 22, 41, 69, 29, 59, 40, 29, 47, 27, 47, 16, 2, 16, 83, 42, 24, 8, 40, 69, 80, 84, 15, 43, 53, 44, 7, 44, 81, 47, 49, 8, 55, 82, 81, 69, 14, 15, 3, 64, 67, 56, 9, 16, 17, 1, 22, 9, 25, 59, 9, 53, 83, 25, 5, 75, 67, 84]

let total = 0;
let coinsLost = 0;

for (let i = 0; i < randoms.length; i++) {
    let roll = randoms[i];
    total++;
    if (roll <= 50) {
        coinsLost++;
    }
}

console.log(`The average at the end is: ${coinsLost}/${total} = ${coinsLost/total*100}%`)

Which results in a 56.599999999999994% coinflips lost

By changing one line:

let roll = (((coinsLost*2+1)/(total+1))**2) * randoms[i];

and rerunning the code, we end up with only 51.9% coinflips lost

And no, it won't be a one-liner in real Gwent because they will need to modify how matching players is processed

no they couldn't because the cycle isn't short enough

I am not really sure what exactly do you mean by "cycle" but the desire isn't a guaranteed 50%, if that was the case easiest way would be to make it so that you lose coinflip a game, then win it the next, which I am not the biggest fan of
Pseudo-randomness like the one I made above are meant to "stir" and "guide" the randomness towards the right direction without taking the randomness from it
You can throw any set of random numbers (even an "unlucky" one as shown) at this and it will make it stir closer by a decent amount towards 50%

 

[Void_Singer]

for near idealized odds you have to either track through on a single player so that the run averages over all the runs of that player (this is actually impossible*)...
or your cycle of numbers has to be short enough that every run represents all values (so two) in every minimum set (also 2).... this would require flip flop logic rather than RNG..

now why is the first one not possible? because you can't take only one player tracked, because the odds are oppositional... you are also up against the percentages of every other player... guess what, that is going to give you a normal distribution, not a flat one across all players. and that is why your formula is guaranteed to fail... it's not oppositional.

so what about that second option? well you could bin all players into two groups, red/blue, and simply match them up with the other bin every new match, and swap the bin they're in after every match right? every other match should end up with a blue coin for every player... except that assumes that all players play the same number of matches, at the same frequency, at the same times... eventually that system will end up with an iimbalance that will become progressively worse resulting in lots of searching for opponent problems.

Even if you could solve the coinflip balance to be near perfect for all players, none of that actually solves the core problem of the wide imbalance between going first or going second. When you can guess the outcome of a match simply based on who goes first, with a high degree of certainty, it should be clear that the problem isn't who goes first, it's that currently there is a first mover disadvantage, both in CA, and because Gwent is largely reaction based. either first mover has to be given some goal they can reasonably achieve faster to couteract that, or there need to be more proactive plays available (such as the new Morenn FC). it's also why dry pass protection is so powerful (and thus hated by some).

 

[Mathspy]

for near idealized odds you have to either track through on a single player so that the run averages over all the runs of that player (this is actually impossible*)... or your cycle of numbers has to be short enough that every run represents all values (so two) in every minimum set (also 2).... this would require flip flop logic rather than RNG..

If I understood what you said correctly, then I believe you misunderstood how this idea would work

How it'd work:
>This formula (or anything like it) would trigger on player attempting to find a ranked match
>The server will attempt its best to connect the closest two MMRs that got opposite results from the last step
>If the server fails or finds it'd approximately take more than X time, it abandons the plan and just connects to someone with a close enough MMR (even if it abandons the plan the match still contributes to total and coinsLost)

There is no need to run it on a single player because we run it even before we match the players together. It gives us a flag, red or blue, which we connect the players according to as well as according to their MMR of course
>Mathspy got blue, MMR at 1000, look for nearest player to 1000 with red
>Whoever got red, MMR at 1500, look for nearest player to 1500 with blue
And of course like I mention, if it finds this to be infeasible within the next X amount of time, it gives up and just connects two to a match like right now, purely based on MMR

And the flip flop logic you mentioned, that's actually what I mentioned here: (although you definitely explained it better)

I am not really sure what exactly do you mean by "cycle" but the desire isn't a guaranteed 50%, if that was the case easiest way would be to make it so that you lose coinflip a game, then win it the next, which I am not the biggest fan of

And like I said, I wasn't a fan of it in first place anyway, though now I see it wouldn't have worked out mathematically well, too, for the health of the server

 

[xBleedingSoulx]

Has anyone compared the amount of red and blue coins of people who have spent real money on the game and people who haven't? It's not unheard of for companies to do things like favour people who spend money.

 

[partci]

Has anyone compared the amount of red and blue coins of people who have spent real money on the game and people who haven't?...

I've spent a good amount on Gwent and I can tell you - I am getting s#1t canned all the time. Something like 65/35. I can say I am dumbfound every time I win a CF. Don't know what to do.

 

[GenLiu]

I've spent a good amount on Gwent and I can tell you - I am getting s#1t canned all the time. Something like 65/35. I can say I am dumbfound every time I win a CF. Don't know what to do.

Good to know I'm not alone.
I don't think people with a "normal" luck can understand how annoying it is to see this bloody blue coin for the 15th, 2Oth time when you're trying to rank up...

 

[Hellsmoke77]

Good to know I'm not alone.
I don't think people with a "normal" luck can understand how annoying it is to see this bloody blue coin for the 15th, 2Oth time when you're trying to rank up...

Yeah it seems like I have "go first" written on my forhead as well. Many days I just give up on ranked because I'm just tired of it.

 

[Barracuda88]

I feel like I'm missing out on some secret, but I just can't make myself care about the damned coinflip. Do people really like auto-lose when they have to go first? Or are we all being a bit too dramatic?

 

[4RM3D]

I feel like I'm missing out on some secret, but I just can't make myself care about the damned coinflip. Do people really like auto-lose when they have to go first? Or are we all being a bit too dramatic?

The coin flip is probably Gwent's biggest problem right now. So, no, people aren't being over-dramatic. The actual impact depends on your deck and the opponent's deck. Certain cards and mechanics, like CA spies, carry-over and Wardancer, can potentially close the gap, but ironically, they also make the gap even bigger when you don't have the cards needed and your opponent does have them.

The problem when going first, is that you need to keep up with the opponent's tempo or go down one card (CA). This is made worse when the opponent plays a CA spy and you don't have one yourself because that means winning with minus 2 cards or losing on even. Neither options are great. In theory, you could dry pass the first round, which is a valid tactic in certain cases. However, against certain decks, that would be suicide.

 

[Esmer]

I feel like I'm missing out on some secret, but I just can't make myself care about the damned coinflip. Do people really like auto-lose when they have to go first? Or are we all being a bit too dramatic?

Regardless of cards and mechanics, a player which goes second controls the flow of the match. The best outcome for the second player is winning on equal cards or losing with the opponent 2 cards down. The best outcome for the first player is to win with -1 card or to lose with opponent 1 card down.

This means that no matter what you do on a blue coin, your opponent can always choose to be 1 card ahead. And if he has a carry over, you can't even dry pass to regain CA.

Sure, spies can revert the coin flip, but you give your opponent 13 points to do that (and he can play his own spy regardless).

The last season I checked my stats over 30+ ranked games with the same SK deck. My winrate on a blue coin was 60% and my winrate on a red coin was 85%. I had almost equal distribution of blue/red coins and my winrate was around 70%. Still, as you can see, the difference between going first and going second is tremendous.

 

[RockPassage]

CDPR: Forget Arena, fix the coin flip.

Less RNG = Good.
More RNG = Bad.

 

[Barracuda88]

The coin flip is probably Gwent's biggest problem right now. So, no, people aren't being over-dramatic. The actual impact depends on your deck and the opponent's deck. Certain cards and mechanics, like CA spies, carry-over and Wardancer, can potentially close the gap, but ironically, they also make the gap even bigger when you don't have the cards needed and your opponent does have them.

The problem when going first, is that you need to keep up with the opponent's tempo or go down one card (CA). This is made worse when the opponent plays a CA spy and you don't have one yourself because that means winning with minus 2 cards or losing on even. Neither options are great. In theory, you could dry pass the first round, which is a valid tactic in certain cases. However, against certain decks, that would be suicide.

Maybe you're right that certain decks are less susceptible to this, and maybe I run a lot of "those" kinds of decks, but I feel like it's fairly easy to avoid going down two cards if I'm paying attention, and going down one card to win the first round is acceptable to me. I feel like it will happen against certain decks regardless of if I go first or second, and sometimes I will lose with a card up regardless of going first or second, idk. If I win with a card down and can't drypass, I will usually try to bleed out all the good cards or near enough, and I will try to out tempo that 3-point wardancer difference until I can even out the CA. Again, sometimes it will work and sometimes it won't. Sometimes I run spies, sometimes I run carryover and lose, sometimes I don't run anything like and still win. I can say I've NEVER drypassed round 1, much less forfeited a match just because I drew the short straw.

I DO feel there's something wrong with OP's 56% rate on a 1000-match sample. That does seem out of whack, and I hope there isn't some weird algorithm at work here, but if it's "fixed" to be a more consistent 50%, I'd probably be ok with it.

I mean, it might be better to do away from CA all together and start every round on even cards, but then you'd have to make sure both players play equal number of cards per round, too, which will probably break more than it will fix.

 

[partci]

10 games - 8 Blue Coins.

BRAVO!