Cactus Game Design Message Boards

I'm glad there is so much support for allowing Mulligans in this thread, because to be frank, there isn't a single good reason not to allow them. They contribute to the fun of the game by drastically reducing the chances that a player will lose from the beginning because they drew a bad hand. Considering the fact that any kind of Best 2 out of 3 format isn't really viable in Redemption because of game lengths, I think implementing Mulligans is a great way to mitigate the possibility of really bad luck knocking someone out of a tournament. I genuinely don't believe there is a single good reason to oppose them, because they maximize player skill and reduce luck, while making the game less frustrating for everyone at every skill level.

That said, I think we're seriously over-complicating what a mulligan should look like. The less conditions or steps that have to be taken, the better. Thus, I propose the following:

The player who has drawn the most souls gets to choose whether to Mulligan first. A player can choose to mulligan one time per game. All souls drawn stay out. The entire hand is shuffled and the player draws 8, replacing souls as necessary.

It's silly to punish people for bad luck, but it still deters people from attempting to trade an average hand for a good one, by threatening additional soul draw. I don't think there's a single good reason that Lost Souls should be put back in the deck.

They increase luck, they just mitigate bad luck. There's a difference.

You're splitting hairs. It reduces the chances of one player losing because they had bad luck, which is what the game cares about.

Wow...ouch! I mean, you are right, but laying down the honesty hammer. With that being the case generally speaking, increasing the chance of souls being available for players who have superior decks to rescue and players having better hands to make bad choices with would be nice. 

As I said, on an opening draw of eight a player is more likely to draw 2+ LS then he/she is to draw 0 for any legal Redemption deck.

Obviously the probability of drawing 2 or more LSs is going to be higher than drawing none, since you are comparing 7 combined outcomes to 1. Since there can only be one outcome of the two trials, you would have to compare the probability of any one of the outcomes of 2 or more LSs to none, rather than all of them at once. If you do a proper comparison, then the probability of two versus none would be the closest. The probability of 3,4,5,6,7 or 8 LSs on the initial draw would certainly be lower individually than drawing none. I doubt you would argue that the probability of drawing 8 LSs is higher than the probability of drawing none. 

You base your categorical opposition for open mulligans on concerns about young players having fun. How do they feel about games where they draw three or more lost souls in their opening hands and have no way to prevent being overrun and losing in three turns?  Do they enjoy those games?

Yes, actually, they do!    At least they get to play their cards and have battles. Soul Drought causes them to discard a bunch of cards each turn in the Discard Phase.....  woo-hoo...

Actually, quite the opposite. This was Olijar-specific trolling made in direct response to his trolling of the thread.

Actually, I was trolling.... 

Something to consider: will allowing mulligans change deck-building strategy?

I think it would...instead of having a balanced deck, it would perhaps lead to decks that are skewed in some advantageous way, with its player mulliganing until they get that starting advantage. Good and proper deck-building should consider how to overcome a difficult start, how to survive and bounce back. Perhaps allowing a mulligan would actually have the opposite of the intended result and actually cause superior deck-builders & players to lose more.

Think about it...if you are playing the national champ, won't you mulligan if you get anything less than a fantastic starting hand? Especially if there's no cost? Is it fair to the national champ, who probably doesn't need a mulligan no matter what they draw? Perhaps this is why if we do allow a mulligan, there should be a cost after all...

First of all, no one is suggesting unlimited free mulligans (as far as I know, because that would just be silly) either a single "free" mulligan or having some kind of stacking detriment (usually losing cards in the starting hand).

Next I will suggest that mulligans will have little effect on deck building, since they likely won't be used in an attempt to turn an okay hand into a good one, but to turn a really bad hand into a workable one, by giving more opportunities to have a decent hand.

Quote from: EmJayBee83 on March 22, 2015, 11:10:49 PM

As I said, on an opening draw of eight a player is more likely to draw 2+ LS then he/she is to draw 0 for any legal Redemption deck.

Obviously the probability of drawing 2 or more LSs is going to be higher than drawing none, since you are comparing 7 combined outcomes to 1. Since there can only be one outcome of the two trials, you would have to compare the probability of any one of the outcomes of 2 or more LSs to none, rather than all of them at once. If you do a proper comparison, then the probability of two versus none would be the closest. The probability of 3,4,5,6,7 or 8 LSs on the initial draw would certainly be lower individually than drawing none. I doubt you would argue that the probability of drawing 8 LSs is higher than the probability of drawing none. 

Actually EmJayBee83's point makes sense (assuming his math is correct, which I'm trying to figure out, but am having some difficulties when I add it up sum the probabilities of drawing each number of lost souls from 0 to 7 and get 1.1. But all EmJayBee83 cares about is 0 compared to 2 or more, since 3 or 4 would also be just as (or more so unfavorable outcomes) so comparing 0 to 2+ makes perfect sense.

From what I understand, the point of a mulligan would be to speed up the game (no offence) or make the game more competitive and balanced (lost souls, but no defence). If there were NO penalties or restrictions, it would sometimes be used for those reasons, but what I think would happen, would be that it would be used whenever someone draws a lost soul at all, unless they had a really strong defence as well. This would neither speed up the game nor make it more competitive. It shouldn't be the norm, it should be a leader ditch attempt to get a defence if you have none but drew lost souls.

As such, as long as at least one lost soul stays out, I would be fine with it. I don't think anything else would be required, since that would solve the main problem of it being used to induce drought. But otherwise I would be strongly opposed.

Quote from: kram1138 on March 23, 2015, 09:14:55 AM

...it [a no restriction mulligan--mjb] would be used whenever someone draws a lost soul at all,...

If you had one LS out, and you took a free mulligan, your chances are better to have a greater number of LS out afterwards than they are to have fewer LS out. If you are really trying to cause LS soul drought this would be a sub-optimal decision.

Quote from: YourMathTeacher on March 22, 2015, 11:24:50 PM

Quote from: EmJayBee83 on March 22, 2015, 11:10:49 PM

As I said, on an opening draw of eight a player is more likely to draw 2+ LS then he/she is to draw 0 for any legal Redemption deck.

Obviously the probability of drawing 2 or more LSs is going to be higher than drawing none, since you are comparing 7 combined outcomes to 1.

As a direct counter-example... In a legal redemption deck you are about 30% more likely to draw 1 Lost Soul than you are to draw 2 or more. These results have much more to do the relative probability of the draw distribution as opposed to the number of outcomes.

To put numbers to this discussion, with a 50 card deck with 7 lost souls, the probability of drawing 1 lost soul is 42%. The probability of drawing 2 or more is 31%. The probability of drawing no lost souls is 27%. I'm not sure where you're getting that 30% more likely to draw 1 than 2+. 42 - 31 does not equal 30%. You are 11% more likely to draw 1 than to draw 2 or more. So, since the chances of getting 2+ lost souls on a redraw is not much higher than getting none, the benefit of getting none would arguably be worth the chance, compared to the cost of getting 2+ when you had 1 before. But even still, if I draw 2+ ls, there is still a 70% that I decrease the number of lost souls I drew if I mulligan.

Once you consider a 56 card deck with 7 lost souls, the probabilities are 1: 42%, 2+: 26%, 0: 32%. In this case you're 6% more likely to draw none than 2+, and probability would say to mulligan every time you draw a lost soul, unless you had a really strong defense.

And the numbers are essentially the same for T2, except that for 105 cards, the chances of drawing 0 are only 2-3% higher than 2+.

So, I seem to have missed this thread... what about taking a page from Hearthstone's book?

You draw 8 as usual (souls going into play and cards draw to replace them), then at that point, you can decide what cards you want to replace. You shuffle those cards back in and re-draw to replace them (souls go into play and you draw to replace them)

Being able to keep some cards while replacing others will give both players a chance at having a strong opening hand. You can't use it to force a soul drought, and both players end up with 8 cards.

Quote from: kram1138 on March 23, 2015, 10:39:38 AM

Quote from: EmJayBee83 on March 22, 2015, 11:10:49 PM

As a direct counter-example... In a legal redemption deck you are about 30% more likely to draw 1 Lost Soul than you are to draw 2 or more.

To put numbers to this discussion, with a 50 card deck with 7 lost souls, the probability of drawing 1 lost soul is 42%. The probability of drawing 2 or more is 31%. The probability of drawing no lost souls is 27%.

Thank you for providing the number.  I wanted someone else to post them as a cross check of what I did.

Rather than going back and actually having to remember how to calculate probabilities, I wrote a quick program that simulated drawing over 1,000,000 trials or so. Those numbers might not be exact, but with a million trials, its close enough that no one cares.

Quote

I'm not sure where you're getting that 30% more likely to draw 1 than 2+. 42 - 31 does not equal 30%. You are 11% more likely to draw 1 than to draw 2 or more.

You are not correctly calculating the comparison of the the likelihood of the two specific events. If I flipped two coins and asked you you how much more likely you were to get one head and one tail (50%) than you were to get two tails (25%), the correct answer--well not correct per se, but the standard answer--is twice as likely (50%/25%) , not 25% more likely (50% - 25%).

So, 42% / 31% = 1.35--which means you are 35% more likely to see 1 LS than you are to see 2+ LS. (This is actually higher than my stated value, because I mis-remembered the exact numbers.)

Ah. That's what you meant by it. Then yes, you are correct. But again, this is dependent upon the number of cards and lost souls in your deck.

And to restate my original position, I would prefer no mulligan, but if it was going to be done, keeping a ls would be preferable.

The percentage of games actually lost to "bad luck" is extremely small. While most players don't actually realize it, most games are lost to inferior deck design and/or less than optimal choices during the game.

I would argue that the existence of Son of God and New Jerusalem (along with dominants in general) makes that less true for Redemption than it is for other CCGs. Quite a few games come down to who draws (or searches) one or both of them first.

So, 42% / 31% = 1.35--which means you are 35% more likely to see 1 LS than you are to see 2+ LS. (This is actually higher than my stated value, because I mis-remembered the exact numbers.)

But again, as kram pointed out, you are using a 50-card deck for your argument. A 56-card deck would not have that same result, since you are more likely to draw none than to draw 2+. A 50 card deck in Type 1 is more of an exception because there are 7 LSs per 50 instead of 7 LSs per 56. This exception would occur each time a player added one LS to reach the next threshold (i.e. 8 LSs in a 57 card deck). I would argue that those people want to draw LSs for their strategy, therefore it is not a proper example for this discussion. Since I am arguing on behalf of new/young players, they are far more likely to have a 56-card deck than a 50-card deck, unless someone instructed them to make their deck smaller.