Kirb your enthusiasm!

WEBSITE HOSTED AT: www.3plusplus.net

"Pink isn't a color. It's a lifestyle." - Chumbalaya
"...generalship should be informing list building." - Sir Biscuit
"I buy models with my excess money" - Valkyrie whilst a waitress leans over him


Tuesday, July 12, 2011

Fallacy 40k - Random Game Length


Random Game Length (RGL) appears to be thought of as a terrible game concept and helps win at all costs (WAAC) individuals abuse players looking for a good time and breaks the game. Well that's the way it seems when one reads reactions to 6th edition posts, BoLS and general opinions on tournaments, etc. The reality...it's one of the biggest balancing factors in the game and is one of the necessary controls to maintain this - it just happens to be random. Let's take a look.

Remember how we said before going first isn't always best? It certainly has its advantages but going second does as well. Going second when you know exactly how many turns are involved increases these advantages and makes some of these advantages even stronger (i.e. being able to move onto objectives last). Consider you know exactly when you have to start moving towards objectives/quarters, etc. and exactly how long you need to delay/hide units which are throw-aways/objective takers in these cases. Imagine Grey Knights with multiple psychic communions literally delaying their units for as long as possible, going 2nd and bringing them all on late game. Stupid, much like Ninja Tau but it could be annoying as hell in specific settings or be used to save one or two small scoring units for late game situations.



Now these cases can certainly be seen in normal games and you in fact see this happen. Their effectiveness? Varied. The difference? Their effectiveness is far less of a certainty. Imagine if all games always ended on Turn 6 and you could reliably and consistently bring your army in on Turns 4 and 5. You can now regularly ensure your opponent only gets a couple turns of shooting and your units can move onto objectives and control table quarters. But again, pretty poor strategy (references Ninja Tau) but what about being able to do this with just a few units such as three man Guardian Jetbike Squads or Acolyte Squads? Knowing exactly when the game ends means these guys become a lot better. This isn't a good thing as armies can then move away from 5th edition concepts such as multiple scoring units to bank on going second, etc.

This isn't the major issue in a game without RGL however and can be more or less effective for different armies (i.e. those who have reserve manipulation). The big issue is what a ton of people complain about - last turn contesting. A big part of RGL in objective (and table quarter) games is not knowing when exactly you need to be on the objectives. In 5th edition you score no points for holding an objective for six turns and losing it on the seventh but do gain a point for claiming that objective on the 7th turn. You don't however know when the game ends (though the max is of course seven) with the game possibly ending on the 5th or 6th turns. This means you must plan for the game ending then and this changes the dynamic of the game considerably.

Consider how not knowing when the game ends in the final turns affects your choices. Your opponent may have more deadly units you can target but you need to kill their Troops by the end of Turn 5 on certain objectives since you cannot contest them. If the game ends Turn 5 - awesome but if the game goes onto Turn 7, you've given the opponent's deadlier units more time to damage you and thus kill your Troops. On the other hand, you may move in to contest multiple objectives on Turn 5. If the game ends - awesome again and you'll most likely win but if the game goes on, you've potentially sacrificed shooting and have potentially exposed the unit to extra firepower. For example, zooming a Land Speeder 24" onto an objective is losing any shooting ability it has whilst bringing it within range of more guns. Again, game ends on that turn and it works out great but if the game goes on, that's firepower you've lost and increased firepower the unit is taking from the enemy.

This is what RGL brings to the game. You know the minimum the game is going to go for (five turns) and you know the maximum (seven) but you don't know exactly what turn the game ends on. This changes how you play drastically. Do you go for the objectives/easy kills/quarters early on Turn 5 and potentially hurt your strategy? Do you wait until Turn 6 or even Turn 7 and potentially lose the game if it ends earlier. You have to weigh these chances up and really prepare for a game ending on Turn 5. At this point you cannot be caught out if it ends early and work this understanding into your strategy. Without this random ending, going second would have a huge advantage and the game would become far less tactical as less decisions need to be made throughout the game but rather in just the last turn.

Comments (30)

Loading... Logging you in...
  • Logged in as
I agree that random game length is a great thing. However, they kinda did it wrong.

The game continues in turn 5 on a 3+ (so, 1/3 chance of the game ending turn 5), then it continues on 5+ (1/3 chance of continuing) in turn 6 and then auto-ends turn 7. Seems logical, right, like a 1/3 chance of ending on each of turns 5,6, or 7? But it's not.

Really, you have a 1/3 chance of it ending turn 5, and then a 2/3 rds of 2/3 rds chance of it ending turn 6 (so 4/9ths, which isn't reducible, but anyway it's almost half) and then a 2/9ths chance of it ending in turn 7. So there's a legitimate chance it'll end turn 5, very likely to end turn 6, not very likely to end turn 7.

I'd have much preferred it was evenly distributed, which would have been continuing on a 3+, then a 4+. That would give you a 1/3 chance of it being any of rounds 5,6, or 7.

I often wonder whether GWs does not understand simple probability like this (and things like the Orbital Strike relay shooting in a 14" ring, tends to make me think that they don't) or if they understand perfectly well and did it on purpose / thought it was better / didn't want the average gamer calling them stupid for it ending on 4+ turn 6, thinking it "ends on turn 7 half the time".
4 replies · active 717 weeks ago
"...then it continues on 5+ (1/3 chance of continuing)..."

No, it actually continues on a 4+, which makes it 1/3 chance to end on each of turns 5 through 7.

3+ to go on after turn 5; 1/3 chance to end on turn 5
4+ to go on after turn 6; 1/2 * 2/3 = 1/3 chance to end on turn 6 and 1/3 chance to end on turn 7
OK, well, now I feel silly. I think I got thrown off by a lot of tournaments around here using 3+ and then 5+.

Nevermind, GW (and mycroft), you win this round! But next time........

Stupid Orbital Strike still hits in 14" circle, damnit. *grumble*
I really wish it was 25% to end on 5, 25% to end on 7 and 50% to end on 6.

To get that first 25% though we'd have to do something like roll 2 dice and if ether is 4+ you continue playing turn 6. A simple 5+ to play turn 7 would get us the 25% for that turn. In my mind this is just complicated enough that GW wouldn't want people to do it in standard play.
Thenakedkasrkin's avatar

Thenakedkasrkin · 717 weeks ago

Simple probability doesn't make sence in pseudo real world battlefield situations as it would assume that fighting strength and battlefield momentum decline in a linear fashion. The longer a fight continues the less armies can fight. Fatigue, casualties, ammunition shortages, etc take their toll on both sides of a fire fight but this decline happens more like a stress curve. Fighting starts and the intensity of the battle builds up quickly, it eventaully plateaus at a climax and then drops off considerably and suddenly at a certain point. This is how GW's RGL works. Turn one or two armies start to move and take pot shots, turn 3 and four closer quarters fire fights ensue, CC's take their toll, and casualties start to pile up, turn 5 continues with that but serves as the breaking point of the battle and each additional turn evokes a feeling similar to that of a prolonged fist fight. You get the feeling that the battle is trailing off and that you and your army are exhausted but still going. In a way, it is was a brilliant change from 4th ed.
I agree that RGL is a great addition to the game that (much like seizing the initiative) forces players to worry about things not going their way. I think the ire on the internet is not so much about RGL but about situations where a fast enemy gambles on the game ending and wins by last turn contest even if they are "losing" from VP standpoint.

In 6th edition I would like to see random game length continue, but I am intrigued by the idea of accruing points every turn for holding objectives. But to play devil's advocate, the downside of this is one player could already have effectively won the game by turn 3. They need to make sure there are still ways the other player can try to win (perhaps VP or tabling).
11 replies · active 717 weeks ago
Yeah, the move to RGL was one of the better changes of 5th ed. I routinely won games in 4rth ed with last second landspeeder pushes and it felt cheap even then.

I really do hate the idea of accruing points over turns, because it absolutely will benefit armies with troops that can contribute to the fight by sitting on an objective the entire game. Many codexs don't have troops that work that way.
Um... by my last count their were sixteen codices, of which six are imperial Space Marines and offer some variant of the Tactical Squad as a Troops Choice... add to that the Chaos Space Marines' and Battle Sisters' 'Tactical Squad', Tau Firewarriors, Eldar Guardians, and of course, Imperial Guard infantry hordes and I can't help but think the _majority_ of codices can offer troops that can contriute to a fight whilst takign and holding onto an objective.

The only ones who would appear to be truly bad at it are Orks, Necrons, Daemons, and Tyranids. But, orks have the numbers to let one or two mobs of boyz hang back; Necrons have a host of other problems anyway and desperatly need all manner of other fixes; Tyranids and Daemons... well... they are meant to be very specialized armies anyway, and a clever player should be able to figure out something.
Breaking a squad of Sisters off to sit on an isolated objective is throwing 200+ points out the window unless a GEQ walks by (SoB squads will 99% of the time be Melta/Melta or Melta/Heavy Flamer and can't be less than 10 models, so no cheap objective holders). Same with Chaos to a degree (depending on the troop choice).

If you do turn-by-turn objective tallies, it absolutely benefits units like Missile Marine tac squads, DE blob DL Warriors, and so forth. If you have an army that needs troops that will contribute in the fight at less than 36" you are going to be unfairly penalized unless there are more objective central to the board than in deployment zones...at which point an overabundance will start to affect gunline armies.
5 Stormtroopers with 2 plasma guns = 70 points of perfect objective campers. What remains to be seen is whether or not they'll still be an option with the new release. Ive run 2 squads of those guys for a few years now and they never disappoint so i willbe very sad to see them go. No one can argue Marines camp better though, better saves and a heavy weapon far surpasses anything the SoB can utilize.
You have obviously never seen an army of all invo save, all AP1 (on 6's), sisters.
I play Sisters and I'm not even sure what that comment means, it has nothing to do with what I said -which is that leaving a squad of Sisters back to hold camp isn't much of an option when you have 12" for your effective shooting and need to blow units off in one go (meaning, you need multiple squads for solid units).

And those abilities cost Faith points, they aren't just free, so no, I haven't seen a whole army of ++/AP1 Sisters.
I understand how the faith points work. Suffice to say, the last two sister armies I faced had more than enough to make it work.
So 4 armies (plus Sisters as noted before) out of 16 (just under a third) get a big "fuck off" ?

This seems especially bad since they're Xenos (or not MEQ sisters) armies; worsening the divide between the "haves and the have-nots"
Demons have plaguebearers, which do just peachy holding objectives (it is essentially the only thing they're good at), and yeah, Orks have numbers, and by and large fearless.

And tyranids, well, with Troops MCs, cheap cheap termagaunts, and the ability to spawn more, I would go so far as to say objective squating is one of the few things they do really well.

Yes, I think every army in the game is built with troops that are either durable enough, or cheap and plentiful enough, that they make fine objective sitters. This is (probably) not an accident. The only folks you could say really lack in this department are Dark Eldar, but even they have wracks, and anyway, isn't it kinda the theme of the army?
I don't think you understand where the discussion about holding objectives came from. The issue is that having turn-by-turn objectives favor armies that have troops choices who contribute without movement and so would be able to sit on objectives.

For example, Guard vs Daemons or DE or SoB all of whom are forced to close with the Guard. X objectives (lets say five) spread around the board. Starting first turn the Guard camp on objectives and score points, the other armies either have to forgo claiming turn-by-turn points for said objectives, or dash between objectives (which may or may not be possible depending on spread) as they obviously can't afford to camp because the Guard will blow them off the table with little damage in return.

The introduction of that sort of rule would require a much more restrictive means of determining where objective markers went to keep the balance between armies who require a lot of mobility to succeed in general and those that require none.
It might have been clearer if you had said, "Points over time for objectives will benefit Gunline armies". I can see now that that point was in there, but it was a little broken up and in-obvious.

As far as whether it's true, or not, *shrug*. It currently kinda awesome to play fast, hard assault troops that can score. Rules changes will lead to play changes, obviously.
I've been wondering, for nearly fifteen years now, why GW has always clung to the "I Go, You Go" format. Given the scale of the standard game of WH40K can easily see ten units and fifty to one hundred infantrymen per side, a straight-up system of Alternating Activations wouldn't work... but, perhaps players could roll for priority/initiative/whatever-youname-it at the start of each game turn? Battletech, Firestorm, and numerous other game systems do this.

You could assign each codex army a bonus, maybe include some (special) characters that have greater or lesser bonuses t5han the defautl for their codex, and inject some tactical uncertainty into the game.
3 replies · active 717 weeks ago
erm, that's really one of the most fundamental decisions about designing a game system. You can't change something like that without having it be a massively different game.
That format is the entire reason I hate playing table top Battletech. Re-randomizing the turn order every turn is complete BS. If you know you are first, you can plan accordingly, same for going second. With so much randomization thrown in, it's incredibly unpredictable and ultimately unfair. If the turn order ever inverts, the player who goes from second to first has just gained a massive advantage.

There is a reason most strategy games have a turn order that is consistent throughout the game (40k, M:TG, etc).
And what happens when someone goes 2nd in T1 and 1st in T2? It would be a fundamentally different game and as it works now, it wouldn't work. Blocking and movement would become far less useful and add in extra randomeness which isn't needed.
Angelic Despot's avatar

Angelic Despot · 717 weeks ago

More than being necessary and fair, RGL also makes the games more interesting and fun - and 'realistic'. In real life you don't win battles by being in the lead at 11pm, regardless of how your opponent has been doing before, or will likely do after.

In 40K, RGL forces players to fight and contest the objectives more consistently throughout the game. As you a) don't know when the game will end, and b) how long you might have to hold the objectives for, you are forced to try to defeat the opponent during the first 5 turns of the game, husbanding your forces and then clinging on to your lead.

It would be pretty dull if the two armies sat back, out of range for 5 turns, and then leapt forward on the 6th, almost ignoring each other in their haste to get to the objectives.
1 reply · active 717 weeks ago
I see you've played 4th edition!
I don't mind random game lenght in itself, but 5 turns is just too short a lot of times. I simply don't like that. 4+ to get a turn 7 would be enough for me. (so you always get a turn 6)
1 reply · active 717 weeks ago
Only a single turn of randomness isn't so great though. This is the one thing at Centurion I think needed to be changed as fester had T6 always with T7 on 4+. Having the range of T5-T7 is much better and games generally go on to T6 regardless, but there is that chance for it ending on T5.
The possibility of a mere 5 turn game keeps gunlines in check. As an IG player, I know I drawed a few games because I played too defensively, got movement blocked, and failed to reach the objectives in time. This is a good thing.
Random game length (coupled with the order of play reflecting the order of deployment, combined with the 1/6 risk of seized initiative, and voluntary reserves) is really important for diversifying Nash-equilibrium strategies in the game, and therefore avoiding the problems with 4th edition where games often broke on whoever won the roll for the first turn.

Game theory can help you understand this by considering multiply-iterated versions of the Hawk-Dove game. The Hawk-Dove game involves picking one of two strategies: Hawk, where you take $1 from your opponent, and Dove, where you give $2 to your opponent. While the second best outcome for both players is to play Dove so both get $2, the Nash-equilibrium strategy is to play Hawk.

In a multiply-iterated version that lasts a finite and known number of iterations, basically game turns, the Nash-equilibrium is still Hawk. That's because the last iteration's Nash-equilibrium is Hawk, and you work back to front from there so that each option at each iteration will be Hawk.

However, if you don't know when (or even if) the game will end, then mixed strategies heavily weighted towards playing Dove are favoured. Indeed, instead of both players just mindlessly playing Hawk, you have a wonderful mix of complex mixed strategies. I say wonderful because there's no absolute one-true-strategy like there is with the finite defined game, and so coming up with mixed strategies can be a fun game in itself.

However...

One can treat Warhammer 40,000 5th edition as having a defined game length if you're going by the book, rather than any fancy Battle Missions or home-brew tournament weirdness. After all, the game lasts between 5-7 turns with each game length having a set likelihood of occurring. In other words, you calculate the best strategy for each of the three eventualities. Still, even the option of three different points at which the game can end is better than one, if not optimal for practical reasons (Warhammer ain't cricket!).
Definitely agree that random game length is good. It means you have to chase objectives early instead of playing defensively most of the game
More importantly where can I get the guinea pig with the hat?
1 reply · active 717 weeks ago
From my cold dead grasp.

Post a new comment

Comments by

Follow us on Facebook!

Related Posts Plugin for WordPress, Blogger...