Introducing... Havilering

A couple of months ago I waffled on for a while about a loose concept for a card game. The core mechanic of this was a trick-taking card game based around a non-transitive trump structure, but I also wanted to incorporate some kind of context-dependent point-trick thing.

I fiddled about with it for a while, and towards the end of last month I posted a page on the completed game, which I am now calling Havilering. That page has the relevant information on how the game works — do go there and check the rules out!

Here however I want to document (lightly) the process of getting to that game. It's not perfect, but I think it is interesting enough to spend a bit of time with. I didn't want it to be a project that hung around too long and just died on the vine, so I deliberately tried to keep the number of iterations relatively low and just get something that works fairly well. There is always room to spin off to other variants if there is a better game lurking somewhere in a nearby space.

Version 1 - 1/8/24

I started out with a plain trick game, just to try and get a feel for the trick-taking mechanics. I toyed with an alternate trick-winning rule, which I labelled as 'progressive'. However I decided the difference between this rule and the one described in the blog post were minimal, so probably not worth looking into unless there seemed to be a real problem with the existing rule. I made a big table showing all the trick-suit possibilities, and only a couple of rows differ. I also now can't quite recall why that rule seemed like it might be simpler.

So to recap, the trick-taking rules were as in the blog post linked above. So with a cycle of suits:

• one suit played means no trump suit
• two adjacent suits played means higher is trump suit
• two non-adjacent suits played means no trump suit
• three adjacent suits played means highest of three (in unbroken sequence) is trump suit (i.e. the one below the absent suit)
• all four suits played means no trump suit

It's not really important to know, but the suit cycle I used was drawn arbitrarily from a second deck:

... < Clubs < Spades < Diamonds < Hearts < Clubs < Spades < ...

The game was just a simple plain-trick game, scoring a point a trick. In partnership.

Play notes

From a couple of plays. Quite a few of these relate to getting used to the tactics for the trick-winning rules

• Initially there was a tendency to hang on to high cards too long, and then they were wasted, as trumping takes over once voids start appearing.
• Note that once someone has trumped, it is usually better to overtrump their suit (play the next suit up) rather than beat their card in the same suit.
• Thought it perhaps might be nice if somehow hierachy could be 'inverted', to avoid situations where winning a trick is impossible regardless of card played. NB: later I decided that this would not only be needlessly complicated, but also would diminish the game by reducing tactical scope.
• Aspects of the play seemed similar to Quinto, but without any suit being overall 'good' or 'bad'.
• This plain-trick version seemed it would probably work better with players playing for themselves, rather than in partnership.
• First half of hand less interesting, until trumping comes into play. But the tactic of trying to generate voids to be able to trump was nice, and felt a bit different to other games, as the void you generate is related to the suit you must use to trump it.

Overall it seemed interesting enough to pursue something shaped pretty similarly.

Version 2 - 2/8/24

I wondered, in regards to it being less interesting until people get trumping, if a shortened deck might help, as more extreme distributions become more common.

This is the same as version 1, but played with a Piquet deck (i.e. a shortened 32-card deck with no cards 2-6).

Play notes

• Seemed less interesting than the full version. A bit too short for much interesting to happen. Game probably benefits from a richer deck in this instance.

Didn't pursue this further, but instead decided to move onto another idea...

Version 3 - 2/8/24

Back to 52-card deck, but now changing the scoring from a plain-trick game to a new idea.

The points were based around the cards won in each trick. These were based roughly on Poker hands, with scores roughly aligned with the number of ways to make that combination (don't take these numbers as gospel - haven't double-checked these are correct!). Flush of course is devalued, as it naturally occurs frequently due to the trickplay rules.

Play notes

• Flush probably overvalued, given its frequency (/unavoidability)
• Limited options for getting more interesting scoring options, especially as so many tricks require following suit

Version 4 - 4/8/24

Similar concept to version 3, except that scoring combinations are more Cribbage-based, and they can stack (i.e. you may score for more than one combination in a trick).

Still as in final version of game, prials don't score as pairs, etc.

I also forgot to include thirty-ones or runs of four, for some reason. And I think I forgot double-runs exist, rather than that being a choice to only allow one.

Play notes

• Base score of one per trick just seems like a complication, and doesn't add much
• This version where 'cricketing' tactic became apparent, which is an interesting feature
• A few interesting moments where score of trick would be the same between a couple of options of cards to play, so choice is decided by cards that would be useful / burdensome later in the hand
• Quite fun, but maybe too busy with all of the scoring combinations and trickplay rules?

Version 5 - 4/8/24

As version 4, but I remembered to add thirty-one (still forgot 4-run!), and ditched the constant 1 point per trick.

All point rates the same.

Play notes

• Having tricks of no points felt a bit weird - especially when this came from an opponent 'ruining' it. But maybe that's a feature?
• In this version the power of running a suit where the one above it is exhausted became apparent. Wondered whether that would be too powerful

Overall this felt pretty good. Wanted to see if alternate trickplay rules might work better, but felt like this was broadly converging on something nice.

Version 6 - 5/8/24

Same scoring as V5, but a change to the trickplay rules, relaxing the requirements into something inspired by 'All Fours' rules (in which you may follow suit or trump).

In this version if you can follow suit you can either do so (as before), or play a card of suit one higher than that led. You may still play any card if you are void in the suit led.

Play notes

• Felt flat - lots of negative play. And many zero-point tricks. Didn't feel satisfying
• More chaotic - could even less try and control suits, and harder to keep track of what the status of each suit was
• Fewer scoring combinations arose in practice

Felt like this was probably a step in the wrong direction, but before returning to whisty rules I wanted to try a variant of this one to see if that worked any better.

Version 7 - 6/8/24

Still the same scoring schedule, but a couple of changes to the trickplay.

In this version if you can follow suit you can either do so (as before), or play a card of suit one higher than the current best card. You may still play any card if you are void in the suit led. This allows to 'overtrump' even when you hold cards in the suit led.

There was also a tweak to the trick-winning rules: when all four suits are present, the highest card of any suit wins (i.e. they are all trumps). This change was to allow other avenues for promoting cards / winning in fourth seat.

Play notes

• Worked better than V6, but not as interesting as V5
• Still excessively chaotic, particularly as no effective way to draw suits
• Standard tactic emerged while everyone has all suits - second hand follows low, as they know partner can overtrump low and win

Version 8 - 11/8/24

Back to the original trickplay rules (but keeping the trick-winner rule where all four suits = all trumps).

Some changes to the scoring schedule however. Aside from the changes in values, two new categories:

• Run 4, as I finally realised I had neglected to include it
• Jack Havel. Had been toying with a Jack score in Cribbage style, but wasn't until the seasonal idea occured that I felt happy with a way to have a preferential suit without having a permanent preference. I guess drawing a card before the deal could have worked, but that didn't really click as appealing.

The other differences:

• Thirty-one - this felt a bit flat before / easily ignored, so made it a little more valuable
• Prial/double-prial & Running flush of three/four - slight boost to three-card combos, and dampen the four-card combos

The reason for the latter set of changes was more theoretical than being suggested from play-throughs (as the four-card combos are relatively rare). For example I'll discuss prials/double-prials, but had similar reasoning for the running flushes.

The only way a double-prial can happen is if a prial is set up on the third card of the trick. This will almost never be forced (only at the very end of the hand), as the player won't be following suit. So this player will only play the prial, if the value of it is enough when compared against the value to opponents of a possible double-prial (weighted by the likelihood of the fourth player being able to complete one). I felt that the previous scoring regime meant this was almost never going to work out favourably - the reward for a prial was not worth the risk of the opponents could make a double-prial. My hope with this revised scoring table is that this now becomes more attractive as a prospect, which in turn should mean that double-prials go from 'never' to just 'very rare'.

The situation is not dissimilar for running flushes of three/four, and so I made an analagous adustment here. It's not as extreme in this instance - running flushes of four may still come about anyway as:

• players may be forced to play cards to complete it, if they have few options of the led suit available
• there are a couple of options of cards that turn a length-three into a length-four
• a four-length run doesn't need to come via a three-length (e.g. 2453)

Nevertheless I still thought an adjustment of this type would prove useful for allowing these higher-scoring combinations to occur slightly more often.

Another required change to the trickplay rules became apparent via double-prial considerations — if they are ever going to occur, then it would almost certainly need the final player to win the trick. I reasoned this could be easily acheived if tied best cards went to the last played. This is a pretty minor tweak, as ties are only possible in the case where all four suits are played.

Additionally settled on the final suit cycle. Handily noticed the multitude of well-known games that have some kind of suit hierachy which are cyclic permutations of one another, so felt like a no-brainer to align with to reduce friction.

Play notes

• Run of three feels a bit rubbish in comparison to a pair, given how much harder it is to win (or comparatively easy to guard against).
• Still pretty chaotic, but think in a good way

Version 9 - 14/8/24

Small tweak from version 8. Mainly geared towards making runs more prevalent in practice.

• Boosting scores for plain (i.e. non-flush) runs
• Runs wraparound (which feels like it fits the theme also)
• Last trick scores 2 (to balance the fact that later tricks can end up lower-scoring).

Play notes

• Scoring feels about right - occasional big scores, but main contribution is accumulation of smaller scores. Pointless tricks rare.
• Wraparound runs worked nicely - harder to tiptoe around scoring contributions

Version 10 - 16/8/24

Only difference is dealer winning Jack Havel now being four, just to increase the jeopardy around what happens with this card.

Play notes

• Felt fairly balanced, and although chaotic, was about the vibe I was after

Happy enough with this, at least for a working version. This aligns with the version on the Havilering page. Nothing to stop future revisions spin/offs.

The name

What about the name? Names are always tricky. They are both unimportant (as they don't affect the game in any way), and yet also massively important (as they can make a big difference to whether or not someone is intrigued, or offput). There are plenty of good games with not-so-great names, and also some less good games with good names. It's not a make-or-break thing, but once you have a name, it is hard to shift momentum to change it.

I find with these sorts of things that once you start using a name in your head for too long it ends up kind of sticking, so I was keen to get a name I was fairly happy with early on, before I lazily stuck with one I was less enamoured with.

I also like to imagine a history to a game. Having an evolution and stories associated with it is more interesting to me than games just being wrenched into existence, even if this history is a work of fiction. It just gives things a little more flavour. I always keep a sketch of this imagined version of history in mind when thinking up games, and when thinking up names. I'm not generally striving for strong 'this could almost have happened', but at least a hearty whiff of near-plausibility. For this game the core of my thinking was it being a Sussex game rooted in the late eighteenth century/early nineteenth, inspired perhaps by Chanctonbury ring, and egalitarianism.

I spent a lot of time thinking about cycles, and rings, and things of that nature for inspiration. I also got quite into the four-valued nature of the cycle, especially after introducing the seasonal suit idea. When the Cribbage-inspired scoring came into things I also thought about things from that angle.

I started thinking about names in terms of 'themes'. I wasn't too bothered with a wholly descriptive name, but wanted at least some sort of link to the nature of the game itself, even if that was via some perversely tortuous route (as I ended up with). The themes I toyed around with (to greater or lesser levels of seriousness) were:

• Cycles
  • Rings
  • Circles
• Flatness (as in the the fact there is no overall hierarchy to the suits)
• Colours (I'm not really sure what I meant with this one)
• Cribbage/Crib
  • Brag (as 'running flushes' came from here. Not really seriously)
• Seasons
  • Humours
  • Classical elements
• Crabs (corruption of 'Crib', or to do with their sideways motion suggesting something in the direction of flatness)

I settled on a hybrid of 'haviler-ring' (haviler being an old Sussex word for a crab), which for brevity became 'Havilering'. Jack Havel is similarly an old Sussex name for a male crab (vs. Jenny Havel for a female), which fit nicely with me wishing to find a scoring opportunity for Jacks. It may be somewhat difficult to remember (if you have no context for it), but is at least distinctive (rather than some games that have very similar, or even the same, name as other games, or concepts, that make searching for information difficult).

Some other names I wrote down, similarly with different degrees of seriousness:

• Trickcycle (really just the WIP name)
• Heaverring (Heaver also a crab in Sussex - same concept)
• Chanctonbury (for the famous ring)
• Crib-catch (yuk)
• Seasonal whist (something nice about this, but not sure the world needs another 'XXX whist' game)
• Rings of the farisees (i.e. a fairy ring. Whimsical but sounds like a board game - has something of the 'Cones of Dunshire' to it)
• Humours
  • Bilious

Thanks for reading this far

So that's it. That's the journey of how I got to the game. It's probably not over yet - further revisions may prove necessary. But time will tell. If you give it a try, do let me know, as I am keen to hear any opinions on it. And I'm sure at some point I will get my 'variants' hat on, and tinker with some related games.

But for now, happy Havileringing!