Thursday, March 23, 2017

Margin of Error

Let me start off by saying I have a lot of respect for anyone that runs a Kings of War tournament - I know how much work goes into it.  I also have a lot of respect for the rules committee for Kings of War - an group of volunteers taking on a job to keep the game we all love balanced (as much as possible) and fun.  I also greatly respect everyone at Mantic, for coming out with a great game that is as balanced as any miniature war game I have yet seen.

This is not, and should not, be construed as negative towards any of these people (or company), nor towards the game itself.  This is my opinion as a player and a tournament organizer, and does not represent any group, company or other person.

Ok - is that enough disclaimers?  This is my thoughts, not belonging to or representing anyone else. They are not happy thoughts, about puppies and ice cream and rainbow unicorns.

Image result for rainbow unicorns
Do you know how hard it is to find a picture when I'm ranting about a rule/idea?
If all you are a special snowflake that needs to be in a safe space, then don't read this post, as there is nothing safe nor special here for you.  You have been warned.

So where do I start?  With the recent release of the new Clash of Kings tournament supplement for Kings of War, one of the things you will notice is the complete removal of the Kill or Kill and Pillage scenarios from the suggested tournament scenario lists.  These are not recommended for balanced competitive play, in face, pg 37 of the CoK book quotes:
So we can all agree that kill is bad.  Don't do kill.  Kill is like drugs - and don't do drugs, because drugs are bad, mkay.

A lot of tournaments (including the Masters, Adepticon Clash of Kings, Gencon and others) use a Margin of Victory table.  There are even two variants in the new Clash of Kings book.

Lets back up a bit here though, and look at tournament scoring.  Now I had gone to WFB tournaments for about ten years (because I'm a glutton for punishment I guess), and I have watched the scoring there evolve as a participant.  Then I started running Kings of War tournaments, and started to incorporate the things I liked and drop the things I didn't.

The base score is a simple Win / Tie / Loss.  In the CoK it uses a 3 / 2 / 1 score for this type of setup. This has the problem of clustering the scores resulting in lots of ties.  A person with two wins and a loss has the same score as a person with one win and two ties.  This can result in needing tiebreakers, which are never what you want.

A second scoring system is the 20-0 system - where you earn battle points based on the difference in victory points in the game.  If  you table your opponent and don't lose any units yourself you get 20 points and they get 0.  If you are tied you both get 10, and he points scale between them.  This can work for a completely kill point system - but with Kings of War moving away from that it doesn't fit. It is really good for spreading out the scores so you are more likely to have a single, clear winner. Even if two players have the same win/loss/tie scoring, they will most likely NOT have the same scores on a 20-0 system.

An adjustment to this system is to use the margin of battle.  This uses a straight 15 / 10 / 5 point system for Win / Tie / Loss, then modifies the scores up (and down ) based on the Margin of Victory table - turning it then into a 20 - 0 system.  The problem is, like the original 20-0 system, it becomes all about kill points, because the winner needs the extra +5 for utterly crushing the opponent if they want to be near the top.

Unfortunately this system can also punish people in scenario based games, which KoW is pushing now.  If you play to win the scenario, your score can actually suffer because you didn't go for the kill. An example is Loot!.  Let's say you grab all three loot tokens with cheap units and head back toward your deployment zone with them, leaving a wall of your toughest units between them and your enemy.  Your opponent throws everything at your tough units, eventually killing them, but they did their job protecting your units that score, and you win the game with 3 tokens to none.  You get 15 points, your opponent gets 5.  In this case however, you may lose several points (and your opponent gain them) because they killed units that, in this case, were inconsequential to winning the game.  So instead of getting a good score for a solid victory, you get a poor score (say 12 - 3), dropping you way below the fellow who won a game holding only one token, but crushing most of his opponents units, and went 19-1 after the adjustment.

Because of the scoring here, ALL games become "scenario X"-KILL.  25% of the score is kill points. All the work to remove the kill scenarios to help balance out the game has been thrown out the window, because the Margin of Victory table turns every game back into a KILL scenario.

I don't like the straight Win / Draw / Lose scoring because they are often not reflective of what happened.  Winning a Loot! scenario with all three tokens is not work any more than winning with only one, and again the scores cluster.

What I do then, is define every game by Major Win / Minor Win / Draw / Minor Loss / Major Loss. The book defines a win (in a kill scenario) as routing 10% more point value of your opponents than they did of yours.  (So in a 2000 pt game, you have to have 200 more points than your opponent to win).  I would define that as a minor win, and then if you double that (i.e. beat them by 400 points) you get a major win.  Point values for these would be  20 / 15 / 10 / 5 / 0.  In a Loot scenario a major win is having all three loot tokens, a minor win would be having 1 or 2 more than your opponent, and a draw is having the same amount as your opponent.

This ends with a little less clustering than the straight win / draw / loss system, and like that system, does not require you to base things off of kill points.

In order to break up the clustering and help us have clear winners, I then like to add objective points to each game.  These are not based around the victory conditions at all, but are objectives for you to achieve to get up to five additional battle points.   These may be things like claiming terrain or table quarters at the end of the game, killing certain units (highest point cost / lowest point cost), keeping your own units alive, getting a rear charge with a non-flying, non-individual unit, etc.  The bonus here is that they can be achievable by both sides - so both scores go up, and you are never 'penalized' for actually playing the scenario.

When setting objectives, I like forcing the players to make hard decisions.  A common one I like is in an Invade! scenario (where you score points for units you have on your opponents side of the table at the end of the game) is to give you objective points for controlling terrain on YOUR side of the table - thus forcing you to choose - does this unit score to win the game, or hang back to get that extra point?

I limit each game to 5 objective points (either all unique, or some  unique and others that can be earned a limited number of time (like controlling scenery)).  These really help to spread out the scores as well as challenge the players, giving a spread of 25 - 0 possible points per game.  (I actually adjust all my points slightly upward, so that soft scores (painting, sportsmanship) have the appropriate weight).

So if Kill! is bad, we should not be encouraging it with the awful Margin of Victory table - instead offering other ways to score games and spread out the scores to avoid ties.  So far of the 24 tournaments I have run for Kings of War, I have NEVER had to use a tie-breaker for the overall champion.  

Because it is all fun and games . . .


  1. This is my first post here although I read your blog regularly from the Fields of Blood link. As an introduction I have been playing WHFB and now KoW tournaments for 20 years and running tournaments for about 10 years.
    While I agree that Kill! is bad as an tournament scenario, as a modifier to scenario scoring I don't think it is bad. This is a wargame after all and you are trying to outkill your opponent.
    The Loot example you gave shows a Pyrrhic victory. The victory has been bought at great cost to your army and surely the victory points should reflect this? If this was a campaign and your battle here was linked to a future battle then I'm sure you would be wondering if you had truly "won".
    The kill points also give the spread that is required to determine a tournament winner.
    I like your ideas of major/minor victories for different margins of loot/objective counters.
    In summary, I would argue that a Kill! in moderation is fine and reflects the type of game we are playing.

  2. @Neil...I agree with you in principle, but we have to ask how we define a "win"? If the goal is to kill Hitler in 1938, stall the Persians at Thermopylae, or get the One Ring to Mordor, then sacrificing everything you've got to achieve that goal is worth it. Yes, the cost is high, but failure results in significantly greater loss.

    But if the goal is to hold that hill over there so we can launch a stronger offensive tomorrow, acquire those resources over there so we can resupply our troops, or cross that river so we can escape the British, then losing 50-80% of your forces probably isn't worth the "win" (aka your Pyrrhic victory).

    Basing points on kill scores simply rewards those with nuclear weapons. Why achieve the goal when I can just annihilate you and thus prevent you from scoring? It's certainly a valid strategy, but it doesn't reward "Generalship". Would Alexander have been as successful if Persia had been armed with automatic rifles?

    I like the 20-0 system as Mike describes it; where achieving goals is what earns points. I do somewhat question the bonus objectives being counter to the scenario goal though. If the point of the bonus objectives is to help separate out the scores, then shouldn't you encourage the accomplishment of those objectives? If the secondary objectives require me to sacrifice the win to achieve them, then I won't bother with them.

    Still, it's better than a kill-based separation mechanic. Kill points should only be used as a 3rd-4th tier tiebreaker, IMO.