My game Tepiiku did unexpectedly well when I released it last year, and I just had a comment on my blog yesterday from a fan who said he wished I’d make a post about making it. Meanwhile, I’ve just released my next game, Oliver & the Basilisks, and though it’s much more polished, it didn’t do as well as Tepiiku. Taking all that into account, it seems worth it to go back and look at the design elements that make Tepiiku work.
Like most of my games, Tepiiku was inspired by a desire to take certain aspects of other games I like, roll them into one, and throw in a bit of my own personal style. In the case of Tepiiku, the games I was looking to for inspiration were a popular dice game I know of as “Nada,” though it also goes by other names, and no-limit poker.
Nada, and other similar games, focus on an “improve-or-bust” mechanic, where the player can keep rerolling dice in an effort to improve his score, but at the risk of losing all progress on the current turn if the score does not improve. In Nada, the goal is just to score as many points as possible, rolling 1s (worth 100), 5s (worth 50), or three-of-a-kind (worth 100x the number, except triple 1s, which are 1000). There are more sophisticated games in the same family, such as Sid Sackson’s Can’t Stop. This mechanic provides a nice ramping-up of tension within each turn, along with progressively harder decisions.
What this reminded me of, in a strange way, was the escalating pot of a no-limit poker game, where a decision early in a hand might only involve two or three dollars, while one at the end of the hand might involve two or three hundred, depending on how the hand plays out. Again, the tension increases, as does the complexity of the decision-making as the hand plays out: As community cards are drawn (assuming this is Hold’Em we’re talking about), the number of possibilities increases, while the betting done earlier in the hand has given the players more information about each others’ hands.
The reason that the pot escalates so much in no-limit poker has to do with odds, and the so-called “dead money” already in the pot. A $10 bet may be more than enough to provoke a fold if there is only $5 in the pot at the time, but the same bet in a $100 pot will almost never do so against a reasonable opponent, because one needs to win the hand only 10% of the time for a call of $10 into $100 to be profitable. Thus, the rule of thumb is that most bets in no-limit poker should be between 1/3 and the full size of the current pot, usually around 2/3. This ensures that the decision for the opponent will be a significant one.
Given that bet sizes are proportional to the pot, the pot then grows exponentially with each bet, producing the same sort of ramping up of stakes and tension as the improve-or-bust mechanic in the Nada family of games.
Tepiiku was born from the idea of combining these two concepts into a single game, producing a sort of double-whammy of escalating stakes and double-or-nothing risk-taking. And indeed, once the pot has grown to a certain size, decisions to re-roll do become nail-bitingly difficult. To my knowledge, no one has ever tried playing Tepiiku for real money, but I can imagine that critical rolls in large pots would have the same emotional impact as all-in situations in no-limit poker, albeit with the strategy rooted more in statistics than psychology.
An interesting side-effect of Tepiiku’s rules is the fact that the intuitively worst position to be in – that of the Fish – is often the best, especially early in the round. Although the Fish is the only player who risks immediate loss, she is also the only player with any control over the round. She can get out cheaply, or she can try to improve to the Hook position… but most importantly, she can attempt to play the odds to control the pot size before doing so.
For example, if the current Hook has a hard-to-improve set of dice, including many Gems and/or non-negated Fires or Waters, the Fish may not wish to jump into the lead right away; if she shoots for the middle position, then the former middle will now be the Fish, and will probably re-roll, putting the original Fish at the bottom once more. If she then rerolls again and now becomes the Hook, the former Hook is likely to be at the bottom, while the pot has grown eight times as large. Even if she loses some small pots by rerolling more than necessary, she still turns a profit in the long run by ensuring that when she wins, she wins big.
The six faces of the dice were also not chosen arbitrarily, but rather to enhance these essential qualities of the game. Ingots and Gems, simply worth one and two, are obvious; the sort of baseline by which other rolls are judged. The Fire and Water mechanic is meant to enhance the contrast of weak-but-flexible vs. strong-but-committed; ending up with several Waters (and no Fires) in a big pot, as the Hook is good… but starting off with non-negated Fire or Water seriously impedes the player’s ability to re-roll, as having it negated means an almost-certain Bust.
The Skull, meanwhile, is the opposite. Though rolling one late in the round often spells doom, a starting roll with several Skulls is actually extremely powerful, as they can be re-rolled with relative safety – even one at a time, if the player wishes to build the pot a little before doubling it.
Finally, the People are there to reflect and enhance the exponential growth theme of the game. A starting roll with one or two People will tend to have a low score, but great potential for improvement. Each successive Person added increases the score exponentially, but also reduces the pool of re-rollable dice (as the player will never want to reroll his People), and thus lowers the chances of getting the next one. Getting five (or six!) People will produce an almost-unbeatable score, but getting there requires a lot of luck!
The trickiest rule in the game to get right was what to do when players tie. On the initial roll, it’s easy – the situation is symmetric, so both players just have to re-roll. But during the main part of the round, it’s different. There were many possibilities to consider: Do you allow the tie to stand, and if so, whose score is considered higher, the one who got it first, or the one who rolled the tie? If it doesn’t stand, who re-rolls, or do both players re-roll? Is it a “safe” re-roll, or can the player Bust? One die, all dice, or whatever dice the player wants? What’s the effect on the pot? Do you double it, halve it, increase it by one, or leave it alone?
Having both re-roll seems to upset the balance too much. Forcing the Fish to re-roll again, while busting him if he doesn’t improve seems unfairly punitive, while giving him a safe re-roll seems overly kind. The original rule was that the Fish was given a re-roll, and Busted only if his score came up less than his original score (before the tie). However, since this score was no longer shown on screen, players found it confusing.
In the end, it proved impossible to find a solution that satisfied all the playtesters, but I felt the rule that made it into the final game was the most fair to all players, and the most consistent with the rest of the rules. The other player (not the Fish) is forced to re-roll, but it’s a safe re-roll. This means it could be either good or bad, depending on the situation – anything situational adds to the tactical interest of the game, so that’s a good thing. Meanwhile, the pot doubles, because a change of player order is likely… but only on the first tie, not if the re-roll results in a new tie; this is to prevent two players from ganging up and re-rolling so as to try to repeatedly tie one another in order to produce an impossibly huge pot when the third player is in a bad situation or has a big chip lead. Also, to prevent it from being a way for a player to deliberately get into a pot much bigger than her remaining stack of chips.
Tepiiku certainly has its flaws; for instance, I’ve never been happy with the fact that Fire is almost never a good symbol to have showing, whereas the others are all more context-dependent. Also, a lot of players complain about the fact that you can’t re-roll when you don’t have enough chips to pay double; this is an inherent problem with gambling games when the players don’t have an infinite supply of money, however. Even the way no-limit poker deals with it, using “table stakes,” is a little clumsy; multi-way all-ins with different sizes of chip stacks and multiple side pots can get confusing, and break the elegance of the game.
Overall, though, I think it’s a charming little game, and the response from the Flash community was much better than I expected; I may even attempt another dice game at some point in the future.