History of Combine

Discussion in 'Announcements' started by skeller, Nov 20, 2012.

  1. skeller

    skeller Well-Known Member

    By Vivafringe

    Curious how some of your favorite Puzzle Strike chips came to be? In Hidden Gems, we explore the development history of various chips. This time we take a look at Combine.

    [​IMG]

    Sirlin has already written about the Combine change. As a result, I’ll focus less on the elegant and wonderful solution and more on the blood and sweat it took to get there.

    During development, we had consistently underrated mono-purple strategies (buying nothing but Combines, money and the occasional Crash Gem). Still, we had made strides. Whenever a mono-purple strategy cropped up, we’d figure out a way to force the character to buy puzzle chips.

    Valerie was the S-tier monopurple character that we missed in version 1 development (the one we caught will have to wait for next week). Her original chips were:

    [​IMG][​IMG][​IMG]

    The optimal strategy, as it turned out, was to use Burst of Speed as soon as possible to get it out of your deck. Then, you would buy as many Combines as you could muster. Unsurprisingly, the degenerate Burst of Speed got the axe in the version 2 upgrade pack. As mentioned in my Jaina article, one of our favorite tricks was adding blue, red, and brown arrows, which gave big incentives to go for puzzle chips. Burst of Speed was replaced with:

    [​IMG]

    In addition, more and more emphasis was put on Puzzle Chips that punished you for playing Combines. The base set already had Stolen Purples. But plenty of other “purple-hate” chips were added as well, including these older versions of One True Style and Combinatorics. Again, note the magic combination of brown, blue and red.

    One True Style (brown, 4 cost)
    +:psbrown: +:psred: +:psblue: +:pschip:
    Reaction: If an opponent plays 2 or more combines in his turn, you may discard this chip to trash all of them.

    Combinatorics (brown, 4 cost)
    +:psbrown:
    Ongoing: Whenever an opponent plays or buys a combine, +:pschip:.

    Unfortunately, even after fixing the most egregious offenders and introducing “hate” puzzle chips, the threat of mono-purple continued to oppress development. When testing a new character, experienced testers went mono-purple by default, only moving to other strategies after being convinced it wasn’t viable. As a result, character designs were often auto-rejected for being “too mono-purple.” Puzzle chips had a similar fate. If they were never better than Combine, why bother?

    About 80% into development, Sirlin announced he was changing Combine. Many were shocked! Puzzle Strike had evolved into a dizzyingly complicated set of interlocking parts, and Combine was a central cog in the complex machinery. Changing Combine now felt like deciding that your foundation needed work after you had already built most of the house! The fact that Sirlin knew this and went ahead anyway showed a lot of dedication, and perhaps a bit of insanity.

    We tried the simplest thing we could think of first: “you can’t play more than 1 combine in a turn.” Unfortunately, this didn’t work for a couple of reasons. First, when going mono-purple, you weren't punished enough. A lot of the time, you could get away with three Combines in your deck by not drawing any two at the same time. For many characters, the reward of being able to Combine nearly every turn outweighed the risks.

    More damagingly, strategies built around card draw (e.g. It’s Combo Time) were powerful because they could eventually Combine a lot in one turn. Adding this rule hurt these decks more than it hurt boring mono-purple!

    A more lenient iteration was proposed that limited the amount of times you could combine based on pile height, but this eventually failed for the same basic reasons: too little actual damage to mono-purple, too much splash damage to legitimate strategies.

    Our next idea was motivated by the insight that we were okay with people playing a lot of Combines in a turn, as long as they were playing something else. It was proposed that Combine’s +:psblack: be changed to a “non-purple” arrow. There were two immediate objections to this. First, characters with washer chips (e.g. Geiger) became unreal good and would all have to be tweaked. Second, we wanted to preserve being able to play a Crash Gem off of a Combine. This led us to replacing the +:psblack: with “You may play a puzzle chip or a Crash Gem.”

    [​IMG]

    Amusingly, this was the only time I ever got to play multiple Master Puzzlers in the same turn on fantasystrike.com. By using a Master Puzzler, then gaining and playing a Combine, I was able to nest Master Puzzlers like Russian Matryoshka dolls. In case you’re curious, yeah I won those games.

    However, a deeper issue was how this new Combine changed the basic gameplay. Players were frustrated that you couldn’t play your character chips off of it, and disliked how crucial washer puzzle chips became for your deck to function in the endgame. The entire solution felt inelegant.

    After that, we finally stumbled on adding :ps$-1:. You can read about how well this solved our problems in Sirlin’s article. It was the kind of change that seemed obvious in retrospect, making us wonder why it took so long to find.

    Anyway, even after finding the :ps$-1: change, our work wasn’t over. We had changed a central cog in a complex interlocking system, and now needed to deal with the inevitable ripple effects. The main ripple effect on puzzle chips was somewhat obvious; many of the “purple-hate” chips we had made seemed sort of silly now that mono-purple was mostly solved. One True Style and Combinatorics were eventually changed for this reason. Ouch! and Stolen Purples were still very powerful, so they remained the same.

    [​IMG] [​IMG]

    The ripple effect on character chips, however, was wildly counterintuitive. At the time, the two characters that went mono-purple the most were Rook and Argagarg. Now, with this change, one would naturally expect the characters to be weaker. In actuality, they were far, far stronger! The problem was that the best strategy against both characters was to mono-purple yourself. After the change, when both players nuked their economy by buying a lot of :ps$-1: Combines, the Rook and Argagarg players won the war of attrition.

    As a result, both characters got nerfed substantially. Stone Wall, which originally reflected gems back at the sender, was changed to the version we have today. Argagarg’s changes were more complex, so I won’t go into them here.

    At the time, I admit to having trepidation about messing around with Puzzle Strike’s basic mechanics. However, looking back now, I’m glad we took the trouble to find a global solution to the mono-purple problem. The :ps$-1: change, while small-looking, is easily the biggest improvement we made in version 3.

    Next Article: History of Lum in Puzzle Strike
    Previous Article: History of Zane in Puzzle Strike

    Attached Files:

    • slide.jpg
      slide.jpg
      File size:
      131.5 KB
      Views:
      23,381
  2. Star Slayer

    Star Slayer Well-Known Member

    I joined development of 3rd edition and Shadows expansion just about when Sirlin announced: Enough is enough! Combine has to change! Crazy, page long discussions ensued. While I wasn't immediately sure if the :ps$-1: Combine was the solution to the mono purple problem, I really liked that it - onlike some of the previously suggested versions - had no text on it, only symbols. I love the elegance of all the basic bank chips having no next (the reminder text on the wounds doesn't count... it does nothing, after all).

    We immediately tried this new Combine out in our local playgroup and it really worked. Players who previously bought almost exclusively Combines now only bought one or two until the late game.

    A second important factor that led to the reduction of mono purple was pushing puzzle chip alternatives. With more powerful puzzle chips at a cost of 4 and 5, Combine became less desirable in comparison. One aspect of this was turning puzzle chips that punished people for buying Combines (OTS and Combinatorics) into ones that combined themselves and could be better than yet another Combine under certain conditions. I really like the new Mix-Master for this reason, too.
    vivafringe likes this.
  3. ApolloAndy

    ApolloAndy Well-Known Member

    I remember this long saga and I will say that I think the final version is elegant and interesting and does everything it should. Sadly, because it came so late in Shadows development, I think there are a few tweaks that could still be made (for instance, Stunlock) but overall, I was surprised at a) how well it solved the problem and b) how little else (relatively speaking) needed to be changed for it to work. I'm surprised the +crash or +puzzle version was even considered. It was super inelegant and drastically changed the valuation of a number of chips and characters.
  4. zem

    zem Super Moderator Staff Member

    I just noticed that the images at the top have “old” and “new” written under them, in black.
  5. Star Slayer

    Star Slayer Well-Known Member

    It was quickly dismissed for these reasons. It didn't even survive two days, if I remember correctly. Brainstorming for new Combine ideas was important at that point, even if it led to a few stupid suggestions. Something had to be done and having to deal with nine stupid suggestions was ok as long as it lead to the tenth being the perfect solution.

    They come from the Sirlin article mentioned above. The black words are clearly visible on the light background of his blog. That's why they are black, even if this makes them very hard to read here. Another reason not to use light text on dark background in an online forum.
  6. Inkstud

    Inkstud Well-Known Member

    Nitpicky, but didn't the original brown-banner OTS cost (4)? Anyway, another great article!
  7. Turbo164

    Turbo164 Well-Known Member

    It also had :pschip: for a while. Then that got removed...from the text but not the actual code, which was fixed the following day iirc :oops:
  8. JohnnySmash

    JohnnySmash Active Member

    I once saw a puzzle strike review on youtube where every time they used Combine they combined four 1-gems into a 4-gem. >.> Also they used Combos Are Hard to get two Double Crash Gems........
  9. Star Slayer

    Star Slayer Well-Known Member

    The new Combine is not only less attractive and less powerful than the old one (making you buy fewer of them), but it is also no longer a 100% play. You have to really think about your possible buys and decide if you want to combine or not. Having to make more relevant decisions is of course great in a tactical game like Puzzle Strike.
  10. Atma

    Atma Active Member

    Guilty here as well.
  11. vivafringe

    vivafringe Moderator Staff Member

    Fixed! Thanks.

Share This Page