Tuesday, July 2, 2013

Points Granularity

Warmahordes is a game of resources. These resources differ at different levels:
  • At the meta level are matchups and players. You are thinking about what list you can put together to counter metabenders like eLich, eHaley, eMorvahna, eLylyth, or generally any epics that don't suck and you expect to see since the local nerd king loves playing OP crap. This could involve playing different factions, bringing in mercs to your drastically one-dimensional Khador forces, or switching from Warmachine to Hordes.
  • At the macro level, you have points. You want to get the best value for your points. This is where you think about making 'the best lists'.
  • At the meso level, you have focus/fury. You are thinking about how your turn will go, and how you want to use those resources. Focus is a bit like getting pocket money, whereas fury is like owning a bank (especially apt now that there is so much fury management that you basically can just invent money and the government bails you out if you screw up).
  • And at the micro level, you have models - stats, abilities, bodies. You're thinking about how each model can be used most optimally at its current location in order to achieve victory, whether that includes the high skill tactic of jamming Iron Flesh super tough Kayazy with Countermeasures in your opponent's face, or something a little more delicate like moving a Warpwolf Stalker 3 times out of activation and picking his targets for maximum Warpath+berserk potential before sprinting away.

This article is about the macro level, the points system. For you young ones out there, Warmachine Mk1 used a points system that was roughly 12 time more inflated. The Behemoth was something like 150 points, Karchev was ~100 pts (WJ points didn't exist, and nobody really played jacks except certain characters and Cryx arc nodes spam), a Journeyman was ~32pts and the Choir was a measly 18/22pts. 500pts was the rough equivalent of 35pt today, 750 = 50 and 1000= 75.

In Mk2, they radically compacted the points system. The rationale was that they wanted people to chose between different options based primarily on what they bring to the table, rather than an 11 point difference.

I actually really liked this change because I didn't have to carry a calculator in my bag anymore. But my gaming group, as well as large sections of the PP forum community, did have one concern with this new points system - it wasn't granular enough, especially in the 1-4 pts bracket.

Pareto Efficiency


Here is an article by Tmage over at Muse on Minis on Pareto Optimum Efficiency being PP's business model and how they aim to attribute points costs to models (as opposed to the GW model which tries to expand the curve with each expansion, a.k.a. power creep).

This was one of the charts used to make his point:

Cryx - it's a hard life.
Some may disagree with the finer points, but the general idea is that '90% of the time, this is the best thing you can get for X points in this faction, all things considered'. Now, it may take some time for players to figure this out and come to an agreement, and the points on the curve may shift a little bit due to meta-changes (for example, anti-tough tech becomes more valuable as every single model suddenly develops a tough exoskeleton), but essentially there will be an optimum, or combinations thereof.

This made me think about points efficiency. Not only are some models just ball-bustingly strong for their points cost, but also blow competing options out of the water 95% of the time. Let's take a look at a few examples:

  • Gorman vs..... almost anything for 2pts.
  • eEyriss vs Yuri the Axe.
  • War Dog. 1pt of holy crap bananas.
  • Reckoners.
  • Tartarus. This guy was 3pts in the original field test, then in the final version magically became 4pts.... along with Ghostly, more ARM and generally being unkillable before having the opportunity to bend you over a table.
There are a few ways to change this around and keep the meta/macro-game exciting and vibrant:

  • You can change the models. This is usually done by errata, since PP doesn't like changing models in fundamental ways. For example, the errata on the Satyxis Witch UA to remove 'no transfers' on its anti-tough/healing aura changed the Witches + UA from being the best choice 100% of the time to the best choice 90% of the time. Cryx - it's a hard life.
  • You can change other models which interact with the models. For example, you can errata Bulldoze to not work outside activation, thereby eliminating the Zerkova/eIrusk dick bubble lists and pushing Khador down the hill (and it hasn't stopped rolling since).
  • Release new models as band-aids- usually this is a UA, like the Black Dragon IFP UA, or the Errant UA. It could also be a caster with specific buffs, like eMorvahna giving concealment to Reeves.
  • Change the metagame - as mentioned, giving every single sh*tty infantry model tough makes anti-healing tech more valuable, and Colossals make anti-ARM lists more prevalent. Releasing powerful new casters increases the odds of seeing those new casters on the table. Essentially this is a change at the meta level that impacts the macro level.
  • You can change points costs. This should be the most efficient and direct method by addressing the issue directly at the source, but has yet to happen outside the Mk1 -> Mk2 change, and will probably not happen until Mk3. Nobody likes to have their 'perfect' lists invalidated by technicalities.

However, there are problems with the points adjustment method as well. The issue I feel is that the points system is not granular enough to make points changes. There's a lot of things that feel like '.5's - not quite good enough to go up a full point, but not quite good enough at its current value. And this is especially visible in the 1-4pts bracket. Think of it like this:
  • 1pt to 2pts = 100% increase.
  • 2pts to 3pts = 50% increase.
  • 3pts to 4pts = 33% increase.
  • 4pts to 5pts = 25% increase.
  • 5pts to 6pts = 20% increase.
  • 6pts to 7pts = 16% increase.
 And so on - the % increase slowly gets smaller and smaller, but starts off pretty significant. If War Dogs are too good for 1pt, the only place to go is 2pts, and that's TWICE the cost. That's a big increase. Gorman @ a hypothetical 3pts is a pretty big increase for a guy that dies to lucky AoE deviations (as long as that AoE isn't fire or corrosion #trollgorman).
On the flipside, comparing a Warpwolf Stalker to a Feral Warpwolf seems like a no-brainer. For 1 more point, you get prowl, pathfinder, reach, berserk, a really good animus, more base ARM, and that extra point of damage your faction needs like a kiwi farmer needs to expose his pasty white thighs to the world. All for an 11% point increase - the 1pt doesn't seem like a big deal in this case. The Stalker seems like a straight-up better deal 90% of the time.

Now imagine the Feral costs 5pts - 50% less than the Stalker. I think you'd be seeing a lot more Ferals. That's kind of the step up we're seeing a low points levels. The key difference between large % point changes in the 1-4pt bracket and changes in the 5-10pt bracket is the "absolute" value of a single point: 2-3pts is only about 5% of your 50pt army building budget, whereas 10pts is about 20% of your total points. It seems less significant at that stage - a single point is relatively more valuable at lower point brackets. This is quite important when you want to optimize list building and get the most for your resources.

Doubling the points system would effectively maintain the absolute value of a point (relative to your total expenditure), while also decreasing the relative value of a point (relative to the opportunity cost). It would also lead to more points along the Pareto curve - and more points along the curve means more options means a more varied and interesting macrogame. It would also open up a bit more design space, as currently the only way to displace an "optimum" option is to release something that does more or less the same thing at the same cost but better, or does something equally powerful but drastically different (very difficult since 'power' is a very abstract concept).


Conclusion

After several years of Mk2, I think the concerns regarding the lack of granularity in the points system proved to be legitimate, especially in the lower points bracket as suspected. I hope they double the points system in Mk3. I don't see any downside to doing it - I can still make armies without a calculator, and the numbers will still be small enough so that the difference between a point and the next will still be meaningful.

Gatorman Posse would be 9.5s and Wrastlers would be 8.5s btw.



4 comments:

  1. What are your views on specialists? (Sorry if you stated them, I skim read the post)

    ReplyDelete
    Replies
    1. I didn't talk about specialists at all (or anything related), but it seems like a decent idea to help with some hard counters. Would have to actually play a few tournaments with it to see how it works out in actuality.

      Delete
    2. What it means is that you have the chance to decide between choices like vassal, madeline, rhupert, reclaimer etc... for your two point slot and not just gorman. It helps playstyle preference breakaway from convention a little bit because there are further listbuilding decisions to make. I also like this because a lot of the time at tournaments I find myself in games where the are so many 2 or 3 point options that would be more useful than gorman or eiryss. If you are fighting something like eButcher's tier, would you rather have gorman or Kell Bailoch? If you are up against eKreuger, would you rather have eEiryss or Yuri? It's something.

      Delete
    3. Ah I see. Good point - specialists increase the value of 10% situational models by allowing you to take them only when the situation comes up, and leave the 90% stuff in your core list.... 90% of the time.

      What I'd be interested to find out is whether 1 list + specialists (like Magic) is "better" than list-chicken.

      Delete