So Big Jim at Galaxy in Flames has been asked to help make a tournament scoring system with composition. He turned to the blogosphere for input in this post. It's certainly the subject of the week at several places.
For what it is worth, I think composition scoring and/or codex-specific restrictions are outmoded. The game is much more balanced than it was a decade or two ago, the background deeper, and Codexes have been designed to be more inclusive of different styles and play preferences. Even more importantly, there is the internet... the tournament community is able to communicate faster, broader and more openly. I think a national or even world baseline of tournament structure, suggested missions, scoring, and other such is now a plausible reality. Errors of codex "creep" could be fixed or toned down, commonly asked rulings could be standardized, and even typos could be corrected... were someone given the official authority to step up and do so. (Yes, I'm looking at you, GWUS.)
But as this august body doesn't exist, tournaments are run by the guy on the spot. Some of them choose to use comp, some run a free for all, and some try re-writing or limiting every Codex in a vain attempt to balance things. I've played and even run all of these. The problem is that they're all based on non-authoritative subjective judgments.
Now while I don't think I'd run a tournament with composition scoring again, I WILL still play in ones that do. I'll even modify and gear my lists for this. Why? Somewhat because I'd rather play than not. Also, I've been there and appreciate how thankless it can be to run a tournie, especially compared to the amount of work it can absorb. But mainly because that's the chosen rules of the tournament. When you go to a tournament, you aren't just playing a game of 40K... you are playing the tournament. If all you want is a game of 40k, do a pick-up game or stay at home and play your friends. I find it ironic how often people complain about composition scoring "nerfing" their codex, yet don't blink when fed a homebrew mission that screws them in the ear. In the same vein, if you don't like the way things are being run, get off your duff and run one how you think it should be.
Anyways, I've digressed. Were I in Big Jim's place and designing a tournament with comp, I'd harvest from what I used to run. Forgive the length, but I'm also trying to explain my rationale too. The actual system is fairly clean and not too complex. I'm certainly not saying it is "perfect", just that I was able to run it for several years. And if nothing else, as the saying goes, "You can please some of the people most of the time..."
First, my ground rules for any tournament:
1. Keep scoring simple. Use checkboxes with fairly easy True/False questions.
2. No surprises. The scoring information should be available to all players well before the tournament day(s).
3. Balance between hobby and game. Composition, Painting and Sportsmanship should be organized such that your "average" army and player is capable of scoring most to all of the points. Also, these should be given enough weight to influence without being overwhelming. A great general with an unpainted net list should be more likely to end up "Best General" rather than "Overall". In the same, a nice-guy general with a beautiful army and mediocre games shouldn't be a shoe-in for "Overall".
4. Transparent scores. At the end of the day, give the awards out fairly and let everyone see their scores and standing. A tournament organizer needs to be (and be seen being) completely honest in all dealings.
These rules came from being both a player and an organizer. I remember one GTT I went to where I was the only player to manage 5 Massacres that weekend, yet friends of the TO won all the top prizes... Said TO also wouldn't show us the final scores nor even talk to me. If you can't look everyone in the eye at the end of the day, you've done something wrong. Not everyone has to leave happy, but don't let anyone leave feeling like you cheated them.
As to scoring? I always liked a 0-100 total system as it's easy on the brain. Presuming a 3-game day, I would run something like:
Composition. 10% of total score.
25%+ of points spent on Troops. +2 points.
40%+ of points spent on Troops. +1 point.
No more than 40% of points spent in any one non-Troops Force Org. +2 points.
No more than 25% of points spent in any one non-Troops Force Org. +1 point.
No triplicate non-Troops unit choices. +2 points.
No duplicate non-Troops unit choices. +1 point.
No minimum sized units. +1 point
Notably, I designed this in 3rd edition as an attempt to put the focus back on Troops, mitigate mix/max, and reduce spamming. It lasted well transitioning to 4th and I've seen it hold up fine in 5th ed too. Composition scores were done at checkin, at the same time the list was checked for validity. These are then used to determine the initial pairings.
Painting & Modeling. 10% of total score.
WYSIWYG, +2 points.
No bare metal/plastic, +1 point
3 colors, +2 points
Detailing, +1 point.
Flock or other base work, +1 point.
"Cohesive" look or theme, +1 point.
Conversions, +1 point.
"Above and beyond", +1 point.
Scoring was against the greater majority of the army, rather than every single model. A couple guys "in-progress" wouldn't count against you for this, scored at 90%+ of the army if there was a question. WYSIWYG was the one exception in that you needed a full 100% to get that.
Basic scoring was either by one judge or an averaged score of multiple. My one rule for multiple judges was that every player got judged by every judge; no splitting the work.
The top 25% scorers (or anyone with a 7+ or some other selection criteria) would be invited to display their army during lunch and compete for "Best Painted". The one rule here was that the player have painted the majority of the army themselves... no "winning with a checkbook". Players could vote their favorite painter on their Game 2 results card. "Best Painted" went based on these votes and ties would go to judging score or a second vote.
Why did I do it this way? I once went to a large RTT with my extensively converted, nicely painted, near display quality Space Wolves. The tournament had 3 judges splitting the work and scoring was purely subjective. Judge A, the TO, was in a hurry and gave me a 6/20 because "the details just don't stand out at six feet away." My friend with the barely past basecoat Blood Angels got a 12/20 for "good effort, but needs work" from the much more lenient Judge B.
While this scoring system doesn't eliminate subjectivity or favoritism, it places that burden of that judgement in the hands of the players.
Sportsmanship. 5% of total score.
Was your opponent reasonably knowledgeable of the basic rules? +2 points.
Was your opponent courteous and civil? +2 points.
Even if you got destroyed, did you enjoy the game? +1 point.
This gets reported every game and then averaged. Optionally weightable by the TO and deliberately kept low points so that vindictive or "old-friend" opponents have less influence. Averaging also keeps feedback honest but lets you be open with publishing the final score. Most players scored perfect 5's every game.
Notably, this was just to maintain basic courtesy at the tables. The tournament rules also had a rider that the TO could disqualify or even evict a player from the store for unsportsmanlike conduct. I only had to use that three times in over 4 years of monthly tournaments and was totally justified when I did.
Players would note their favorite opponent of the day on their game 3 results card. "Best Sportsman" would be based on these votes and average score.
Battle Points. 75% of your total score.
The primary mission objective is worth 0-18 points. Got Massacred = 0, Major Loss = 3, Minor Loss = 6, Tie = 9, Minor Victory = 12, Major Victory = 15, Massacred Them = 18.
Additionally, each game has up to 7 bonus points for secondary objectives. Typically 2-4 of varying complexity and weight. For example, "+2 points for killing an opponent's HQ" or "+1 for having a Scoring unit alive within 6" of the center of the board at the end of the game". This can be switched up every mission in a smaller tournament. In a larger (20+ players) I'd keep them the same every turn to make it easier for all parties. With these, also make sure it's something achievable regardless of the opponent's list. For example, if an objective is "Kill your opponent's most expensive Fast Attack choice.", also add "If they do not have any Fast Attack then, in order, switch this to Elites, Heavy Support or then Troops." Also, try to have these be a set amount of points per game, rather than dependent upon something variable like KP.
Notably, running 20 points main and 5 points secondary also works. I just like having a bit more emphasis on secondary items. "Best General" was based strictly on these scores.
Obviously, "Overall" went to the highest overall score and would over-ride any of the "lesser" three prizes.
In the rare case of a tie where both parties hadn't placed in another category, I'd usually just split the prize.
In the final analysis: The biggest thing is to be open and honest at all time. The less subjective you are as a judge, the more people will realize you gave them a fair shake.
Cheers and hope this gives you guys some ideas...
***Image thanks to GIS and this site on fireworks... all too appropriate for this explosive subject***
UPDATE: I forgot to mention, winning Overall put you out of the running for the other top prizes. Also, a prize would often go to the person in last place; typically a new brick of dice. For large events, door prizes, goodie bags, shirts and other stuff randomly or universally given out are also nice.