No One's Maths Thread (Mostly GSC)

[gigasnail]

eh. i'll look at them again when it's time to put a more serious list together, but, i'm not sold. if i wanted to play guard, i'd play guard or better yet, traitor guard. cult guard is nowhere in the same ballpark as traitor guard for flexibility and useful units.

number crunching may not win games on it's own, but it's better than guesswork and the power of friendship towards giving you an idea of how a given unit should perform in a given situation over a period of time. expecting every die roll to come up on its average roll every time is beyond folly. you all know this and it's half the point of this thread.

[No One]

gigasnail said:

number crunching may not win games on it's own, but it's better than guesswork and the power of friendship

Any objections to sigging? Because I feel this is totally appropriate to me (and I've got a couple of others of yours I want to save for posterity in my sig )

[mattblowers]

Mar 10, 2017 6:25:07 GMT gigasnail said:

expecting every die roll to come up on its average roll every time is beyond folly.

Probability vs "average" and how this plays into expectations has messed up more than one gamers plans. I don't claim to be a math-hammer guru but I know dice can absolutely ruin your game if you are planning your game around getting those average rolls. You haven't lived until you have failed a 3" charge or fail 12 straight 3+ saves and don't make a single FnP.

That being said, I never blame dice on losing the game (funny how you will hear lots of complaints about dice losing the game for players, but a win is always the result of brilliant tactical play). Even with bad rolls, if I lose it's because I didn't take enough variables into account.

[No One]

Ramble (& not exactly a direct response) post:

Expecting the average is OK - that's a good starting point, seeing as it is the average. Now, basing a strategy on getting the average, less so. If your game/turn/phase plan hinges around getting the average on pretty much anything, you're probably stuffed (especially if you repeat this over every phase - even keeping the matthammer going, expecting the statistical average every time is incredibly stupid).

But, how much are you prepared to lose to mitigate the risk of not rolling the expected result? While sometimes everything will pan out beautifully, and you can line everything up so you've got back up plans and contingencies without sacrificing other things, that's usually not going to be the case.

e.g. Say you've got a unit that you need to kill. The flyrant 'average' is 6 unsaved, you need 3. Are you going to dedicate 2 flyrants to it? What if there's another unit that you'd also really like dead, that a flyrant could be expected to basically kill on the other side? What if there's nothing nearby? Do you essentially 'waste' a turn of flyrant shooting just to mitigate unlikely bad dice? What if it doesn't need to die, or 1 unsaved would do it etc.

You mention failing a 3" charge - while that'd screw you over, there's often not going to be anything you can do about it. Not charging on the basis of failing a 3" charge isn't a good idea - hiding/shooting's going to result in a longer charge. etc. There might be mitigation measures you can take - but there might not without gimping yourself in other ways.


TL;DR - basing strategies around 'bad dice' has the potential to stuff you up just as much (if not more) than basing strategies around the 'average'. It is definitely something to always remember though.

mattblowers said:

That being said, I never blame dice on losing the game (funny how you will hear lots of complaints about dice losing the game for players, but a win is always the result of brilliant tactical play).

Eh, while a can't speak for others, I've definitely blamed dice for losing the game (on both sides), in my opinion totally legitimately. Definitely more so than 'tactical acumen'.

The 2 examples:
1. I had 2 flyrants on the board (cult in reserves). Shot his long fangs (flak) with both - only killed the sarge. His turn, 2 wounds (1 on each flyrant) - I failed both 3+. Then failed both grounding, charged, tabled. I lost because of bad dice - could I have played better and mitigated the dice in a way that wasn't really detrimental to my game? Probably. Does that mean that the loss was a result of him playing well? Heck no.

2. Against SM - he failed probably over half of his 3+. This was definitely the case early game, where he lost 2 squads to 2 flyrants, and had a 10 man assault squad+kitted captain run off the board. (Pretty sure I rolled entirely hits with the mawlocs as well, which definitely helped). I don't consider my victory the result of outplaying my opponent. Could he have played better and possibly pulled off a victory despite this? Maybe. Would I have still won if he hadn't had bad dice? Probably (I do think I played better, and my army dealt with his pretty well). Pseudo edit: Was more to the game than I remembered - but still, the impression I had a couple of months later was that dice stuffed him over more than great tactics did on my part (unless you count '4 flyrants' as great tactics ).

While usually dice hew close enough to the average (especially over the course of the game) that any major deviations on either side should be just considered 'part of the game', and losing because of that shouldn't really be considered losing to 'bad dice'. But, sometimes they don't - some games people just roll consistently equal/under (or equal/over), and major dice skew horribly one way. I'd definitely call losing in those circumstances at least partially the fault of 'bad dice'.

[mattblowers]

Mar 10, 2017 14:15:27 GMT No One said:

But, how much are you prepared to lose to mitigate the risk of not rolling the expected result? While sometimes everything will pan out beautifully, and you can line everything up so you've got back up plans and contingencies without sacrificing other things, that's usually not going to be the case.

Everything. Winning consistently in 40K is about having a game plan for when things go wrong. If you NEED to kill something (ie. the flyrant comment) I absolutely will dedicate 2 flyrants to it when 1 is usually enough. Just last night the nominated unit that I needed to kill I only got 2 hits out of the TL 12 to stick. Wounded on neither! Luckily I had redundancy built in and the next one rolled above average and took out the remainder.

Redundancy is crucial to win 40K. It's why I really don't like arguments built around comparing things in a vacuum. Plans always go awry when things go sideways in a game. It's not a mistake that all the top 8 LVO were high ranked ITC players. They consistently find ways to mitigate dice that get cross eyed so they can still pull out the win. It's the reason why people are happy to take Fateweaver at 300 points.

That doesn't mean I won't take chances. I absolutely will take chances and have games come down to a single roll of the dice. If I can get firstblood and warlord on my first turn I'm happy to dedicate 5 flyrants to get the job done when 3 should work on average dice. So maybe I waste a turn of shooting with 2 flyrants (asssuming they can't shoot a sub-optimal target) but better that than failing to get those points. If the dice start to go sideways and the game gets tight those 2 tertiary points could easily win the game. A lot of players won't do that, but it's because they are expecting average dice on this turn rather than backing up your plan with redundancy and playing for end game from turn 1 on. I've had a lot of opponents get focused on the here and now and forget end game. This gets them out of position and then they struggle to get things righted for end game scoring.

And then some days things just go badly and you get tabled because you are playing Tau. Dice happens.

[gigasnail]

Mar 10, 2017 10:19:38 GMT No One said:

gigasnail said:

number crunching may not win games on it's own, but it's better than guesswork and the power of friendship

Any objections to sigging? Because I feel this is totally appropriate to me (and I've got a couple of others of yours I want to save for posterity in my sig )

Work out.

[kazetanade]

Mar 10, 2017 6:25:07 GMT gigasnail said:

eh. i'll look at them again when it's time to put a more serious list together, but, i'm not sold. if i wanted to play guard, i'd play guard or better yet, traitor guard. cult guard is nowhere in the same ballpark as traitor guard for flexibility and useful units.

Dont hurt a guy for trying. =/

No One I tested a medium sized Flyrants + Mawlocs+ Sub Up vs the new Traitor Legion gladius - word bearers with 5 rhinos, basic troops/special weapons tac squads, bikes with Lord and Apostle (on foot), 2 units of cultists, 2 ML3 DPs and a ML2 Sorc no bike. 

1850 points, 615 points of Sub Up in 4 units (the rest in Nids). Results are atrocious - I rolled 6 once for Sub Up the entire game. You cant set up safe enough/close enough to threaten the Rhinos without being absolutely ballsy and risk losing your entire ground presence in a single if not 2 turns. I feel like the mathhammer about Cult Uprising makes it impossible to use as anything more than a "I'll set up here first", and it's basically all luck if you ever get to charge first with it - constantly having your allied portion cycle through it for CA fishing is giving up any hope you have on the board, unless you really have some kind of scoring system you can land on and grab at every turn. 

Worst part was the summoning - the guy didnt summon any Tzeench horrors like I was expecting, he summoned almost exclusively doggies - T4 W2 A2 fearless nightmares with 12" movement and (please do not swear) deepstrike. 

While Mawlocs werent helping, I actually question whether more Sub Up would have helped, as I was constantly getting swept in every engage - lose by 3 models, fail leadership, fail init get swept. I am now firmly believing that Fearless is far more important than Hatred as an allied force. If I had taken the points from the Sub Up and put them into a FC with token obj sitters in a GSC CAD, I may have been able to blocked out the middle ground and just grind the entire army down with Flyrant support.

One thing I noticed is also that the raw number of bodies you can bring out with the points of the mawloc + the Sub Up also lets me bring 3 Maguses, 2 Neophyte troops, and a 6 man Doting Throng which puts almost 80 bodies with both buffs on the table to choke the midfield. This also would have basically won me the game(s) with how long it would take for boxes to break through and for melee to chew through, but then I lose any chance of dealing against invis deathstars. But I think just 600 points isnt enough to do this and I rather bring the FC which is far more hardy (fishing for Shrouding psychic power, may get invis too). But at this point, I dont even think I should keep the Flyrants, and should just go pure Cult to fully abuse the CA flexibility given to such vast numbers of bodies. 

Just some thoughts for you, since you are running a Flyrant + GSC allied for a tournament.

Of course keep in mind my side is so stupidly EW focused - even for this "modified" mission, the only additions were VP for killing enemy units in your turn, and VP for controlling double or more objectives than your opponent does (which is almost impossible vs a Gladius). 

[almostmercury]

I don't know man. Set up far back, the bearers come to you and you can literally be anywhere. With a primus, 9 ambush dice each turn is no laughing matter, but running without fearless is a setup. When I went against 3 knights and (please do not swear) lost in the worst way possible, I still knew my list would perform from the stathammer. 10th percentile just happens sometimes.

Moreover, they're only rhinos. Pop them with flyrants.

First Curse is over doing it, but a Patriarch in a CAD is probably worthwhile.

[gigasnail]

For nid allies I would roll flyrants and mucolids only, do the rest of the GSC, then check and fill with nids only as/if needed.

[gigasnail]

I would think that going full-cult and running the board with doods is going to run into time issues. It's probably the strongest option, but I dunno that it's viable b/c of state of the game/rules bloat.

[kazetanade]

Mar 10, 2017 18:59:38 GMT almostmercury said:

I don't know man. Set up far back, the bearers come to you and you can literally be anywhere. With a primus, 9 ambush dice each turn is no laughing matter, but running without fearless is a setup. When I went against 3 knights and (please do not swear) lost in the worst way possible, I still knew my list would perform from the stathammer. 10th percentile just happens sometimes.

Moreover, they're only rhinos. Pop them with flyrants.

First Curse is over doing it, but a Patriarch in a CAD is probably worthwhile.

Setting up far back basically gives him control of the board - not always a bad thing, but with 2 objectives each and the midfield objective, all he needed was to bum-rush one of my two objectives and force something off for 2 turns and I'd auto-lose from the points difference. 

The rhinos were okay, I started popping them in Turn 2-3 and wrecked 4 out of 5 of them by the end of the game. Problem was the summoning, he was outputting 3 or 4 new units on the table every turn while the DPs and Sorc were alive. At the 4th turn he even pulled a surprise summoning from his warlord, who got turned into a psyker from the traitor legion warlord traits table. I got luckily and killed both DPs by top of T2, so in total he managed to pull out around 5? or 6 dog units throughout the entire game - if he had his DPs survive to T4, I would have been buried under dog bodies regardless of CA or not. 

The other problem was that with 9 ambush dice, I rolled only a single 6 after 5 turns. By turn 4 I had run out of opportunities to fish for CA 6s as the hounds and bikes were closing in on both my objectives and had to try and cut them off before they could reach - so I spent an entire game to fish for the formation's biggest benefit and came up seriously short. While out of the ordinary, this is nowhere near the 10th percentile problem and is actually completely possible to happen again. This tells me that you need a very large number of units to be using CA in order to pull out a consistent impact from it - again unless you are scoring for just being somewhere. So either you need a max size SU (being 4 acolytes and 3 metamorphs of whatever config), and probably another one for good measure, or you play around with a different army strategy. The fearless tarpit is what I'm going for next (although Renegade zombies do the tarpit better, I dont have access and I dont play humies), unless I figure something more effective with First Curse. 

Lastly - why I think the need for consistency is more important than potential, in any tournament setting the one mistake that happens is pretty much going to screw your entire campaign. When dice happens, better have a back up - and I just realized that SU doesnt have a back up.

Hopefully I get a few more games to test - I like Mawlocs because they threaten so much but they do so badly against freaking gladius esque lists. I kind of need to know whether to drop them as dead weight or not. 

[kazetanade]

Mar 10, 2017 20:25:38 GMT gigasnail said:

I would think that going full-cult and running the board with doods is going to run into time issues. It's probably the strongest option, but I dunno that it's viable b/c of state of the game/rules bloat.

It's a time waste strat. I was thinking of doing it as even 40 bodies + mawloc/flyrant makes my game run for 2++ hours, but I might get called for "intentional stalling" and get disqualified. 

[gigasnail]

Yeah that's not a road I would want to go down. I like to play a game, not try to game the system.

[almostmercury]

Yes, rolling 9 dice each turn and only getting one 6 over 5 turns is possible, it doesn't involve consistency.

It just as likely that his summoning farm rolls perils over and over and falls into the warp in addition to whatever damage you do to it. It doesn't make summoning inconsistent.

[kazetanade]

Mar 10, 2017 22:07:37 GMT almostmercury said:

Yes, rolling 9 dice each turn and only getting one 6 over 5 turns is possible, it doesn't involve consistency.

It just as likely that his summoning farm rolls perils over and over and falls into the warp in addition to whatever damage you do to it. It doesn't make summoning inconsistent.

You're comparing a 1/6 chance of success vs a 1/72 chance of similar impact failure. Word bearers have a 3+ for conjuration, spell familiar to reroll failed psychic, then take 4 turns to perils themselves to dead unless they roll 1 and fail ld10.

Point is that its very consistent compared to hoping for 6s on a small allied force.and the main thing I want to tackle here is how to tackle CAs consistency issue so that even if you roll badly you can cover it.

I think the issue can solve itself just by bringing more points in it. But if you're not willing or not able to, pretty, hard pressed to think of an alternative.