Kirb your enthusiasm!
"...generalship should be informing list building." - Sir Biscuit
Wednesday, August 18, 2010
Battle Points: Why They're Bad for Competitive and Casual Players
Let's take a break from talking about particular armies and builds and units and instead talk about something much broader: tournament rankings. Specifically, battle points that many tournaments use. Usually this means some kind of primary/secondary/tertiary objectives system, or alternately a minor win/major win/massacre ranking. Battle points are added up over each round of the tournament and at the end, the player with the most wins. Seems straightforward, no? Drastically lowers the need for messy tiebreaker criteria- if you've got 57 BP, you're ahead of the guy with 56 BP. Nice and easy.
Easy, yes, but not nice. What battle points encourage is the opposite of what competitive AND casual players want. Let's review for a second: a competitive player is generally going to get the most enjoyment out of a hard-fought game against a strong opponent- someone who really challenges his play skills and pushes his abilities to the limit. (In the most ideal sense, they don't care whether it's a win or loss provided it was a good game, but the reality is that most everyone would rather win than lose, be they competitive or not.) Casual gamers would like to have a relaxed game where both players can enjoy themselves, chat, show off models, act out scenic battles, etc- there are a lot of different kinds of casual gamer, but the uniting factor is probably that they want a game to be about mutual enjoyment, as opposed to a contest. (Again, all other things being equal a casual gamer would probably rather win, but that is not necessarily their only goal.) Battle points are detrimental to both these styles of play.
Let's look at the competitive gamer and what he wants. Obviously, he wants to win the match and do the best he can- after all, he's here to compete. "Doing the best he can," in a battle points game, means that he wants to get every possible battle point. This, pretty much by necessity, means crushing his opponent as hard as he can, since most BP are a "either one person or the other gets them" sort of thing, and even when they aren't they generally require taking something from your opponent, like destroying one of their units. This means that not only is the competitive gamer forced to play like a jerk to get what he wants- taking every possible advantage and punishing every mistake the enemy makes in order to maximize his ability to get BP- in order to get an ideal score, he needs to be playing against a much weaker opponent. This is the first and foremost problem with BP: they reward "clubbing baby seals," to steal a phrase. If two highly-talented players are ranked up against each other, it isn't a contest of titans for victory in the tournament- it's a near-guarantee that neither of them will emerge the overall winner because they didn't score a massacre against their opponent. BP reward whoever gets ranked against the most low-skill opponents and crushes them ruthlessly. This isn't what competitive players want and it isn't fun for them unless they're huge jerks.
This is where the casual player comes in, and where his lack of fun stems from. If we are using a simple win/loss system, a casual player and a competitive player coming to the table is not necessarily a problem; both players can get what they want out of the game, provided the opponent is willing to work with them. The competitive player is happy to play out a good game with a clear victor- and if he doesn't win, it's because he didn't play well enough. The casual player can have his wacky scenario or epic character duel or whatnot as part of the regular game, and it doesn't necessarily cost anyone anything. But when battle points are introduced, it creates friction- suddenly, the competitive player needs to earn as many points as possible in order to have a chance of advancing- he can no longer afford to indulge in narrative fights when it has any chance at all of depriving him of a secondary or tertiary objective. His ability to compromise with the casual gamer is severely limited, as he's on a much tighter leash with regards to his goals. This means the game is going to more tense, more competitive, and less fun for both parties.
Battle points are a poor system; football teams don't determine a winner at the end of the season by adding up the total points they scored in all games over the course of the year, and for exactly the same reasons- it unfairly rewards teams who got easy matchups or got lucky and trounced an opponent badly, somehow defining this win as "better." Rather, they stick to a simple win/loss/draw- either you beat them or you didn't. Sometimes draws are avoided, even and everything comes down to tiebreakers, at which point "secondaries" become perfectly acceptable- if you had the same number of objectives, go to VP, or table quarters, or KP, or whatever. But battle points not only encourage unfun play, they can also create absurd overall results where the tournament "winner" was beaten by the loser in one of the earlier rounds (and never played them again.) Huh? My army beat your army, and somehow you still come out on top? That's a sign of a poor system right there, and one that needs to go.
Many of the arguments for battle points come down to the fact that games of 40K are long and players can't be expected to play for sixteen-plus hours in a single day; battle points cut down on the number of rounds "necessary" to determine an overall winner. Fair enough, and true as far as it goes, but they only do so by producing a misleading result. Shorter rounds (which generally necessitates smaller battles, rather than the 2500pt messes that are 'Ard Boyz over here in the States) and multi-day tournaments are the answer here: four two-hour rounds on the first day and semifinals/finals on the next are not an unreasonable expectation, and leave plenty of time for breaks for lunch and in between rounds, etc. A particularly ambitious tournament might run all four rounds AND the final two in the same day, but that becomes a truly all-day affair, clocking in at fifteen-plus hours, at which point everyone involved would probably be dead on their feet.
Battle points are not the most crucial issue facing the 40K tournament scene; certainly, there are plenty of others to worry about. But they are important, because they are harmful to the validity of the results and they are a deterrent to the "good" aspects of competitive play and encourage many of the bad ones. Whether you are a competitive player or not, the issue affects you (unless, of course, you don't attend tournaments), and not simply in the sense of winning or not winning the tournament; it changes how your opponents will behave and the sort of play environment that the tournament encourages. We owe it to ourselves, as a community, to make the game enjoyable for everyone, and the current system does exactly the opposite.
Subscribe to:
Post Comments (Atom)
31 pinkments:
Totally agree here. That's why I like Nova's system. There's no BP or soft scores to manipulate the results, you just have to win and keep winning.
Thanks for this take on battle points.
I would like to share how we deal with this in France.
Here 95% of the large tournaments are 2 days affairs, with 5 or 6 games depending on the format.
Single day tournaments are usually at the 1000pts level, with 4 or 5 games as well. Almost all tournaments also use the swiss pairing system to prevent seal clubbing.
First turn is random or uses comp to pair the players, then swiss for the other rounds, with comp as a secondary key.
There are no finals.
It usually allows a clear winner to emerge if a player wins all his games, if not then a secondary key is used.
However, 40K being a rock-paper-scissors game, I don't think that two players relative rankings should be determined by a game between them, but by overall tournament results.
I just discovered the blog and it's been most interesting.
40,000 is not rock, paper, scissors.
My army brings all three.
Switching to a Win/Draw/Loss format also helps lists like Tau, who realistically don't have the ability to completely massacre their opponents in the way that Orks or Wolves or BA do.
Excellent take on battle points.
@Kris
Tau are MORE than capable of massacring their opponents- the plethora of railguns, plasma, and missiles ensures that. However, your point is entirely valid- some armies aren't realistically going to table the enemy.
Excellent article Puppy. Another point about 'competitive' gaming. When playing against a good opponent with a good army list you want to beat them fair and square not because of silly mission rules, etc. Knowing that you played to the best of your ability against a quality opponent and won is good whilst if you lost there isn't going to be griping (assuming the player isn't a tool).
I think for larger tournaments when you have enough time to play enough games to eliminate everyone the win/loss/draw system is great.
Unfortunately, the tournaments I attend are single day tournaments, which is all I have time for anyway, and must determine a winner from only 3 games when 12 to 80 players have shown up. I have a feeling battle points won't be going away anytime soon for those tournaments.
Now that doesn't mean I don't wish there was a better system. I've been at a tournament where I got second place because I tied my second game so the best two players played against each other in the last game, and I got to play someone easier. It certainly wasn't fair that I got second place. However, the tournaments I attend aren't meant to be "competitive" at all. They are hobby tournaments that I attend with the soul goal of winning 1 to 2 games. I ignore comp and fill out everyone with perfect sportsmanship because I honestly just don't care.
The article made good points and really does show the problems with battle points. Unfortunately, I haven't read a very good solution for 3 game tournaments to not need battle points.
@Evil-Termite
Why not use Battlepoints as a secondary Tie-Breaker in 3 game tourneys. Better than a kick in the face?
Assing VPs to the objectives in the game and total VPs to tiebreak?
Rank the top tied players based on their (beaten) opponents end results. - ie. the guy that beat 4th, 8th and 12th is probably better than the guy that beat 5th, 18th, and 54th.
Thoughts?
@Icereane; we should move to France... VT2 is correct in saying 40k isn't rock/paper/scissors though. Whilst the armies available from 5th codecies are diversifying a lot and balanced lists cannot handle all lists equally as well, it's not a game of A > B > C > A > B > C, etc. Some armies are like this indeed but not the balanced onces. The biggest problem I think most tournaments will find is pairing in Round 1. Swiss is great but unless we use things like comp or something to do initial match-ups (since Rankings HQ is meh atm until tournaments actually become competitive), there's always going to be that bit of imbalance.
In regards to tournaments which cannot produce an un-defeated player, rankings should be based upon W/L ratio first and like fester said, use something as a tie-breaker. For example if for a one day tourney with 3 games, 8 people got 3-0, those are the top 8 players and then rank them according to some secondary system. Then there's 24 who went 2-1 and they are ranked 9-24, etc.
W/L ratio should be the primary determinant of tournament winners.
Isn't this kind of the point of seeding? After the first two rounds, you should be playing against people of similar skill, thus allowing you your hard competitive games. Having run tournaments before, I once used a system where players got BP's for only loss, win and draw. The overwhelming feedback that I got was that there should be distinctions for minor and major win/losses.
If you're playing anything less than three games, than you're not playing a tournament, you're playing a game of can I get lucky matchups, but the more games that you play, combined with seeding should give a clearer winner.
Most Australian tournaments are 5 games over two days.
@Eltnot
Yes, it is, but seeding is only partly helpful in that regard. Consider:
After the first round, where matchups are random (or close enough to, since comp rarely does anything useful about determining which armies are good), you've only selected for armies once. There are still going to be lots of "poor" players who scored well by virtue of luck, good matchups, etc. Someone who gets matched against one of them is better off than someone who gets matched against a "good" opponent- which is essentially my complaint. Battle Points makes it so that you don't want to play anyone skilled, because that effectively knocks you out of the tournament. And, as Stelek has noted many times, it can allow someone who didn't go undefeated to beat someone who did, which seems like a poor representation.
>The overwhelming feedback that I got was that there should be distinctions for minor and major win/losses.
I think the problem is that most people don't understand _why_ this is bad. I don't have any problem with recording major/minor wins for tiebreaking purposes, but it's a poor system for determining who the winner was.
@Evil-Termite
Three rounds plus semis/finals is workable up to 32, I believe; if you're running more people than that, I think you can afford to dedicate the time for a fifth round or something. I mean, 40+ people is bigger than the regional 'Ard Boyz I was at for the semis...
">The overwhelming feedback that I got was that there should be distinctions for minor and major win/losses.
I think the problem is that most people don't understand _why_ this is bad. I don't have any problem with recording major/minor wins for tiebreaking purposes, but it's a poor system for determining who the winner was."
I also think that a Minor win is better than a massacre as that often represents two lists of similar power run by similar general levels running into each other and having a great game compared to a good list trumping a bad list or lucky rolls, etc.
I wonder how well something like this would work at the local RTTs. They're 12 players, max, which makes it kind of difficult. I've always wanted to organize one, though.
It is incumbent on a TO who is designing a set of tournament scenarios with battlepoints to make them not particularly about grinding the enemy into paste and rubbing their face in the dirt. I use battlepoint objectives that tend to have goals other than just abusing the other player, and these seem to work out pretty well for both the hardcore and ?softcore? players.
I don't think pure W/L makes for any more pleasant of a game inherently. I've certainly had games where I was demolished and had a great time, and won games by narrow margins only to find my opponent very unhappy. I experienced this at a tournament earlier this year and my opponent said his worst loss was a close game - it takes all types.
@MSP
Twelve is kind of an awkward number, although I suppose you could do something with a bye. That feels off, though; maybe three rounds, and then run the best players in a finals game? I dunno. I'm not a statistician nor an expert in game theory, so I can't say what the "best" system would be for that number of people.
@Jwolf
Whether or not it makes games better is certainly arguable, but I don't think you can argue that is discourages good players from wanting to match up against each other, which seems bad in my view. Isn't that exactly what you want, to have the best players play each other and one of them come out on top? BP don't really allow that to happen.
I agree with the original post, there should be no gray area with points. Won Lose draw, only three possible outcomes! playing a tough tactical battle and winning against a pro should not be compared to tabling a "fluff bunny".
Use of a swiss draw is the only logical way to ensure people get to play against worthy opponents, and you can decide a winner at the end of the tourney. Finally countbacks can be used to compare people on the same final score through opponent %ages, This is how they have run MAGIC tournaments and shows how if you have defeated tougher opponents than a player on the same final score you should come out on top.
Awarding different points for major victories etc, is a slipper slope. (towards battle points).
I was recently asked to input about an upcoming tournament and I advised a win/lose/draw system. Earlier this year I saw a friend finish behind people who won fewer games solely because he didn't "pants" the weaker opponents and partly due to his army's structure which could not control 5 objectives come turn 6-7. Not fair in my view? YES?
Let's consider the following scenario:
3 players IG, Eldar and Chaos Deamons, plus an SM player:
turn 1 IG wins against Eldar
deamons wins against SM
turn 2 Deamons wins against IG
Eldar wins against SM
turn 3 Eldar wins against deamon
IG wins against SM
Best case scenario : everyone played everyone else.
Who wins?
That seems like a rather unlikely scenario- for one, you're playing three rounds and only have four people. Why not just go single-elimination (in which case you would have a very clear ranking of Daemons, IG, Eldar, SM following the end of round two, since the two winners played each other and the two losers played each other.)
This is a case where you _would_ have to resort to major/minor wins, because you have set up a rock/paper/scissors play between the three armies makes it effectively impossible for any one of them to claim to be dominant. But, again, this seems like an amazingly unlikely result- try and achieve such a scenario with a 16-person tournament and you'll find it to be nigh impossible. (There _are_ still a small number of configurations that would result in the same kind of tie, but the chances of them happening decrease exponentially as the number of players increases.)
More games != better results.
AbusePuppy - I don't see how Battle Points keep the best players from facing off against one another, nor do I agree that any system makes good players want to fight each other more or less. I do agree that any system can be run improperly and create anti-fun incentives for the players. Part of the job of the TO is to balance the mission goals to make an overall fun experience for the participants that is accurately competitive. Anyone who says they know the way to do this and there isn't another one probably hasn't run an event of any size. It is not a similar thing to run a 20-30-50 man event for locals and a 150 person event for a widely varied field. What works on a 1-day rogue trader amongst your club members doesn't really scale up very well, but I hear many people who haven't ever even done that espousing their righteous and holy knowledge of how tournaments should be run.
Complex scoring models are harder to understand and use well than simple WLD systems; that is true. That truth does not translate into WLD is better; it just means a bad TO with a WLD system will run a much better event than the same bad TO with a PST system. A good TO can do well with either, so personal taste is more important than the system for a good TO.
The point Puppy is making is as follows. Assume there is a field of say 32/64 players which could potentially have a single un-defeated player assuming the tournament is run that way.
Player A & B & C are all very good with good lists. Player D is terrible but has a good list. Player A plays versus player D and crushes him and gets say 20 battlepoints. Player B & C play and B narrowly wins so gets 14 battlepoints. Does this show who's the better player? Player A could now play Player E who is also crap but with a good list but played against Player F who was crap with a bad list, etc.
Obviously this isn't eliminated in a W/L system as some people can 'sneak' below the Radar and some individuals could have 'easy' runs but there's minimal difference between crushing your opponent and scraping a win, the end result is the win.
BP have their place I think in tournaments where you're not aiming for an un-defeated player at the end but I'd be hesitant to call that event 'competitive' due to the increased abilibty to inflate your score and not actually be the best there. When players who are 5-0 or 4-1 place lower in terms of generalship to players who go 3-2 I think there is a problem.
Battle points might be good in determining tie-breakers but I think your primary W/L should determine your ranking potential (i.e. 3-2 cannot beat 4-1).
@JWolf
BP don't keep the players from playing against each other, they effectively punish anyone who plays a skilled opponent. If you barely manage to beat me, you probably are not picking up the extra points for doing extraneous stuff, which means your win is "worse" than someone who just tabled a n00b.
This is where the problem lies. We want a system that encourages the best players to face each other; battle points do the opposite of this.
I haven't claimed I have "holy knowledge" or that a W/L/D system is some sort of perfect gift descended from the heavens above. But if you do have that 150+ person tournament, you should be able to run a two day event and end up with an undefeated champion, not four or five rounds and then pass off a prize to whoever managed to get paired up against the worst spread of players.
I used that scenario to show 2 points:
First with direct elimination if you change the first pairing, you change the end result.
Whoever plays with SM first is going to win the tournament.
Second, if you only want an undefeated champion, then the randomness of the pairings is more important than individual skill. It will all depend on wether you face this or that army in the KP mission or the objective mission.
If you face a good opponent in the 2 markers mission, a draw is much more likely than in a KP or objectives mission.
Finally: why would you want to encourage good players to face each other? They will face each other if they are paired, their will has nothing to do with anything.
If you use a swiss round over 2 days with 6 games, you can have 64 players and be almost sure to have a clear winner with a WDL system (remember, draws can happen). If you have more than 64 then there is no way you can have a clear winner with less than 7 games unless you use battlepoints (or a secondary key, which is also dependent on the spread of players you played).
If you have more players
@Abusepuppy - Are you just letting players choose their opponents or something? Because in any system that forces those who do well to play each other, the best players will face each other, barring wierd dice. I don't see any difference between W/L and P/S/T BP (which is really gradiated and multiple level W/L)in that the best players should naturally end up facing each other over the course of the tournament, given enough rounds. And it's not because I'm not doing the math - what I mean is there is no significant statistical variation in the outcomes. The starting pairing vs. mission for each player in either system is significantly more important than the system itself for determining winners; this is an area where a lot of work can be done, and one I hope to be able to cooperate with other TOs in generating a decent seeding system.
@Jwolf
Only if you define "best player" as "whoever has the most battle points." My contention is that if two _skilled_ players face each other, they both suffer in the rankings, since neither is able to earn many BP from the match, whereas crushing a weaker opponent places you high in the standings. W/L/D doesn't do that- they guy who beat a rock-hard BA list and the guy who phased out some Necrons are equal.
@Icareane
>First with direct elimination if you change the first pairing, you change the end result.
I was actually suggesting that you eliminate the _last_ pairing, since it's unnecessary.
>Second, if you only want an undefeated champion, then the randomness of the pairings is more important than individual skill.
O_o What. No, what? Are you saying that the random variables (what army you get paired against, what mission is selected) are more important than the controlled variables (strength of list, skill of player)? I'm sorry, I will not accept that, because doing so effectively means that the entire tournament is a coin flip. Do those variables affect the outcome? Certainly it does, just like luck of the dice can screw a good player just as easily as a bad. But a strong list is prepared to fight many different kinds of opponents and is prepared for all of the different missions.
>If you face a good opponent in the 2 markers mission, a draw is much more likely than in a KP or objectives mission.
Sure it is, that's why most people thing that's a poorly-designed mission. But it's not like there's no strategy to it, or that you can't win it; I've won lots of Capture and Control games, and not just by virtue of crappy opponents.
And again: 64 players is a lot. If you have the 128-player tournament, seven rounds doesn't feel unreasonable spread over two days. I'm not a tournament organizer, and I don't run Adepticon, BoLScon, or any of the others, but if you're expecting multiple hundreds of players... you probably need some kind of qualifier event or something. Unlike M:tG, you can't cram a game of Warhammer into a 1hr time period, it's just not doable. Qualifiers, break the tournament up, whatever, I feel these are all better solutions than a flawed system of rewarding victories against the wrong kinds of players.
I have run several large GTs and never seen a player sneak into 1st place due to their soft scores. For me and how I run a GT that is another Internet myth. I think the best way to go about the pairings each round following the first is by use of Swiss style pairings. The players scoring the highest should be playing each other otherwise you run the risk of implementing a system that rewards players for clubbing baby seals.
G
So examples provided by others across multiple blogs of where individuals won less games but scored higher means what then?
Using points allows for very flexible / multi-tiered missions, and you can always reduce the points to a simple W/L/D model if you want. I have an article posted about this subject.
Oh fuck yes. (can I say that? I'm a bit too used to HO these days).
Battlepoints also DRAMATICALLY decrease the odds of the the winning player being the person who got to the top table and stayed there the whole time. Tiebreakers based on who you faced, on the other hand, are PERFECT for this.
Would you rather play in a tournament where the guy who got the nub opponents has an advantage, or the one where you have to have beaten people at the top tables to take first?
Swiss pairing like MTG
@Evil-Termite: If the tournaments aren't competitive, then there's no need to deterime rankings, and thus no need for battle points. Get rid of the rankings, play three games against random opponents, have fun, and go home happy.
@Jwolf: Hyper competitive players who can't stand loosing are certain to be miserable when they encounter an equally skilled, or more skilled player. And win or loose, they will probably make their opponents miserable too.
In most (all?) sporting tournaments, you have a regular season or round robin in which, ideally, every team plays every other team. When that isn't possible, matchups are generally based on regional groupings and/or results from previous tournaments; however, the more competitors you have, the more round robin games you'll need to get statistically valid results. From this pool the best 2^n players are chosen to compete in n rounds of playoffs. If you have a tournament with 100+ players and 2 days, I doubt there is any way to derive meaningful rankings. If you want meaninful rankings, and only one or two days to play, your touranment probably has to be limited to 16 or 32 players, maybe 64 at the most.
If you must run your tournament in a manner that makes meaningful rankings impossible, then perhaps your tournament would be better off without rankings? If marketing considerations force you to have rankings anyway, nobody should take those rankings seriously.
If you don't care whether or not your ranking are meaningful, then why bother with rankings at all?
Post a Comment