MATHS: Modelling some proposed changes to Close Combat Part 1

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
For alternating attacks I'd like something like Death Maiden w/needle pistol and claw Vs stimmer with paired pulverisers.
Followed by the same death maiden v Stimmer with renderiser, resolved as if ADADADAAA.

That'll at least be a decent starter, contrasting a comparatively weaker but more attacks pulveriser Vs the hard hitting but fewer attacks renderiser. Plus toxin.

Okay, here is the napkin math.

1).
She hits on 2+ (0.83), wounds on 4+ and rerolls 1's for toxin (0.58 = 0.5+(0.5*0.16)), 5 out of 6 times furnace plates won't save stimmer (0.16) and 4 out 6 results on an injury dice result either in OOA or SI which are identical because of coup-de-grace.
Total probability to OOA with a single attack: 0.83*0.58*0.83*0.66 = 26%, so a chance for a single attack to fail to OOA is 74%. Stimmer retaliation will happen only if all 5 attacks (3 base, +1A charge, +1A dual melee) will fail to OOA the stimmer. Which is 1-(0.74)^5 = 22% of the time... Probability to OOA a stimmer through flesh wounds exclusively is less than 1%(~0.016% to be exact) so I won't even count it in.
A Stimmer gets to retaliate after a death maiden charge only 22% of the time. Talk about engaging experience for both players /s

2).
Same chance to OOA a stimmer with a single attack, 26%. This means that stimmer gets to retaliate 74% of the time! The cumulative chance of killing a stimmer after 3 attacks, before maiden gets to use leftover attacks is: 1-(0/74)^3 = 60%.
A Stimmer gets to retaliate after a death maiden charge a wooping 74% of the time. Good, but it feels that number of attacks don't matter much, because by the time we do 3 alternative attacks (ADADAD...) and get to use leftover attacks of maiden, one of the combatants will already be dead (60% from maiden alone, 3 retaliation attacks from stimmer rise this probability even more). So a big number of attacks does not matter, what matters is how hard do you hit. Not an ideal system, but way better than "I charge - you lose" of GW.

3).
The cumulative chance of killing a stimmer with 2 attacks is: 1-(0/74)^2 = 45%.
I've used this method of alternating attacks. A Stimmer gets to retaliate after a death maiden's initial charge 55% of the time. With this system, a number of attacks matters quite a lot. I'd even say that Unwieldy weapons need to give a +1A with this system, otherwise they become obsolete.

Conclusion: What is important with Alternating attacks system is that lethality is still high, one of combatants will die most of the time as a result of initiated fight, and I think it is fine. Close combat should be lethal. However initial alfastrike should be lowered a lot, making retaliation attacks (and traits that affect those) actually matter. Also it does not reduce the effectiveness of WS4+ fighters, since they they'll get to retaliate properly as well, hitting on 4+ and not some 5+ or 6+ which would make their usage against any melee champion questionable at best.
 
Last edited:

Al_Weeks

Gang Hero
Honored Tribesman
Dec 22, 2014
513
533
123
Bristol
Any chance of letting us know the stimmers loadout and what his chances are of OOAing the death maiden when he makes a retaliation attack?
 

Kiro The Avenger

Gang Hero
Apr 4, 2018
1,224
1,588
128
Bristol, UK
It sounds like your issue is primarily with reaction attacks not happening often enough?
I'm not sure that's the only or even the whole issue. If melee is just a coin toss for who gets the kill, why would anyone do it?
 

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
It sounds like your issue is primarily with reaction attacks not happening often enough?
I'm not sure that's the only or even the whole issue. If melee is just a coin toss for who gets the kill, why would anyone do it?
Yes, my issue is exactly with Reaction attacks not mattering. I have no issue with melee being lethal as hell, in fact it is the only thing that sets it apart from shooting. Fights should be resolved in a single phase, otherwise it opens up possibility to easily gang up on a charged fighter. Also you do not want to prolong the game too much.

Its not a coin toss though. Whoever has better weapon or set of weapon traits is more likely to come on top with alternating attacks. Killy expensive weapons have a better chance to slice someone apart in a single blow, parry has a much bigger impact, same with disarm... It creates interesting dynamic. Charging is always beneficial since you hit first. I'd say that there should not be traits or skills that allow retaliating before charge attacks occur, otherwise it ruins the whole point of charging.
 

Kiro The Avenger

Gang Hero
Apr 4, 2018
1,224
1,588
128
Bristol, UK
I think there's currently only one rule in the game that allows you to attack before the charger does, and that's the new Courtier.

Alternating attacks makes any engagement between peer-adversaries pretty much a coin toss, only slightly weighted towards the charger.
It also means any single damage weapons become more or less obsolete if you intend to attack multi-wound models, since the fight will likely be determined long before you start being able to chip away at your opponent.

That's why I'm interested to see the maths beyond how likely one side is to make reaction attacks. To see the real impact of this
 

Thorgor

Of The YAQ
Oct 12, 2015
4,580
10,522
148
36
Sevres 92130 France
Charging someone who is as good as you in combat (and who hasn't been softened with ranged attacks beforehand) should always carry a sizable risk. There are a lot of safer targets on the battlefield for you to charge. And charging someone who is better as you in combat should be a desperate move you cannot expect to survive.
 

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
Any chance of letting us know the stimmers loadout and what his chances are of OOAing the death maiden when he makes a retaliation attack?
Chance of a single parried renderizer attack to OOA maiden is 28%. Chance of a single non-parried attack to OOA maiden is 34%. So first 2 parried attacks of a stimmer have a cummulative chance to OOA maiden of 48% "1-(0.72)^2 = 0.48". Third unparried attack will rise this cummulative chance to 64% "0.48+(0.34*0.48) = 0.64". 3 unparried renderizer attacks have 71% to OOA.

Alternating attacks makes any engagement between peer-adversaries pretty much a coin toss, only slightly weighted towards the charger.
It also means any single damage weapons become more or less obsolete if you intend to attack multi-wound models, since the fight will likely be determined long before you start being able to chip away at your opponent.

That's why I'm interested to see the maths beyond how likely one side is to make reaction attacks. To see the real impact of this
It is not "slightly weighted" towards the charger, in the case of stimmer and maiden the charger has a solid chance to stop the fight with a single blow 1/4 of the time (26% for maiden and 28% for parried stimmer). There is still a huge incentive to pull a charge off. Obviously multiple damage weapons become more important, but single damage weapons can make up for that with traits and/or price/ability to be carried alongside pistols. At first glance this looks like a balanced system to me and it has a lot of design space for skills and traits. In fact a lot of traits become way more useful without having to change them: knockback, parry, disarm, entangle. Also it gives both players a feel of agency, since they both get to roll dice and use tactics/skills midfight to influnce the outcome.

I'm not that good at calculating probabilities,so I don't know how to calculate who will come on top when different types off alternating attacks are used. @Thorgor, could you give a helping hand?
What I'm sure of though is that using comparative chart will ruin WS4+ fighters. Say, a WS4+ fighter with a toxin knife charges a stimmer. Currently each of his 2 attacks has a 10.8% to OOA a stimmer, ending up in cumulative chance of 20.4% "1-(0.892)^2 = 0.204". By making his WS6+ (-2 because WS4+ is double of a stimmer's WS2+), you are lowering a chance of each attack to whimpy 3.5% and a cumulative chance becomes 6.8%. With alternative attacks it is a 10.8% to end the fight with the first attack.

Edit: I liked NCE melee system because a big number of attacks did not guarantee a win. You could roll a lot of 1's and end up getting shanked by a juve who rolled averagely. Sadly, NCE melee system won't work in model by model activation system, so instead of bringing it into YCE (and resulting in a lot of traits/skills/stats needing a rewrite), why not emulate it by making a large pool of attacks not as decisive when charging?
 
Last edited:

Thorgor

Of The YAQ
Oct 12, 2015
4,580
10,522
148
36
Sevres 92130 France
Chance of a single parried renderizer attack to OOA maiden is 28%. Chance of a single non-parried attack to OOA maiden is 34%. So first 2 parried attacks of a stimmer have a cummulative chance to OOA maiden of 48% "1-(0.72)^2 = 0.48". Third unparried attack will rise this cummulative chance to 64% "0.48+(0.34*0.48) = 0.64". 3 unparried renderizer attacks have 71% to OOA.
Unfortunately, it's not that simple. Probabilities become harder to calculate manually when attacks are resolved one at a time. The first attack may fail to kill the Death Maiden but still wound her, leaving her with no Wounds and one Flesh Wound. This would make her easier to kill with subsequent attacks (I think the Wound roll stays the same in this case, but you will roll 2 Injury dice instead of 1, and she may end up with 2 additional Flesh Wounds, etc.) It can still be done given enough time though, but that's when a simulation would do a better work than me.
 

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
Fair enough. Thats why I've called it a napkin math 😅

I feel that it does not rise the probability too much, since more than half of outcomes on injury dice are OOA anyway, so if you got to injury dice, you've probably made a kill (especially with pulverise trait). The point was to show that retaliation from alternating attacks create more of a back and forth experience.
 

Al_Weeks

Gang Hero
Honored Tribesman
Dec 22, 2014
513
533
123
Bristol
Unfortunately, it's not that simple. Probabilities become harder to calculate manually when attacks are resolved one at a time. The first attack may fail to kill the Death Maiden but still wound her, leaving her with no Wounds and one Flesh Wound. This would make her easier to kill with subsequent attacks (I think the Wound roll stays the same in this case, but you will roll 2 Injury dice instead of 1, and she may end up with 2 additional Flesh Wounds, etc.) It can still be done given enough time though, but that's when a simulation would do a better work than me.
Thanks for summing it up neatly. I will get round to this, real life is getting in the way atm.
 

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
Thanks for summing it up neatly. I will get round to this, real life is getting in the way atm.
Stimmer and maiden have an awful lot of traits that affect probabilities. I think it is easier to calculate probabilities for mirrored champions.
I'd compare identical melee champions: WS2+ T3 W1 3A no armour.
One uses a D2 weapon which wounds on 2+. Other uses two D1 weapons which give +1A and wound on 4+. Then compare how they fare against each other when charging/being charged in vanilla ruleset, alternating attacks system and advanced alternating attacks system. Maybe I'll get to that sometime.
 
  • Like
Reactions: Petitioner's City

Kiro The Avenger

Gang Hero
Apr 4, 2018
1,224
1,588
128
Bristol, UK
I know that it's not that precise a calculation, but assuming each attack had a 25% chance to OoA the opponent is reasonable as a napkin calculation for discussion.

In that instance, it's a competition for who can roll a 4 on a d4 first. That's a slight advantage to the person that goes first, but really not that great.
And that's peer combatants, you mentioned that a ganger has a ~10% chance to cause an OoA, imagine them going up against a Stimmer when they have *one* attack before the Stimmer starts his best down.

This really emphasises one aspect of reaction attacks that I think people are missing. They cost the defender nothing.
When charging with alternating attacks, I would be using my precious action points to give the enemy free attacks against me, and all for the benefit of a small chance of knocking them out first.
This is very different to shooting, with shooting the enemy gains no actions and, in fact, loses an action if they get pinned.
That's why I think combats should be heavily in favour of the attacker. Because they're the ones 'paying' their actions to make it happen.
 

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
And that's peer combatants, you mentioned that a ganger has a ~10% chance to cause an OoA, imagine them going up against a Stimmer when they have *one* attack before the Stimmer starts his best down.
10% to cause an OOA is exactly for the initial single attack. With second attack that a WS4+ fighter will use if they survive stimmer's retaliation, they have the same cummulative chance of taking out a stimmer ~20%. Also surviving initial stimmer's retaliation is easier as well.

This really emphasises one aspect of reaction attacks that I think people are missing. They cost the defender nothing.
When charging with alternating attacks, I would be using my precious action points to give the enemy free attacks against me, and all for the benefit of a small chance of knocking them out first.
This is very different to shooting, with shooting the enemy gains no actions and, in fact, loses an action if they get pinned.
That's why I think combats should be heavily in favour of the attacker. Because they're the ones 'paying' their actions to make it happen.
I agree, this is why I've proposed appropriate tweaks for alternating attacks. If you have more attacks than opponent (usually from charge) then your next sequence will use up 2 attacks instead of 1. This rises chance from 1/4 to 1/2 in case of stimmer/maiden and makes charging more important. (AADAADD if 5A maiden initiates a fight on 3A stimmer)

TBH the action economy for melee was always borked in many GW games because in order to engage you have to be within B2B. If you have 5 separate shooters standing near each other and a single charger, most of the time charger will be able to attack 1 shooter at a time, meanwhile each shooter can contribute to killing the charger.
 
Last edited:

Thorgor

Of The YAQ
Oct 12, 2015
4,580
10,522
148
36
Sevres 92130 France
That's why I think combats should be heavily in favour of the attacker. Because they're the ones 'paying' their actions to make it happen.
They are also the one choosing who fights who. Combat should not be the best tool in the box to kill melee specialists imo, you should shoot them. Combat should primarily be used to take out targets that are not good at combat.
 
  • Like
Reactions: Petitioner's City

Petitioner's City

Gang Hero
Nov 15, 2017
914
1,517
133
Edinburgh, UK
They are also the one choosing who fights who. Combat should not be the best tool in the box to kill melee specialists imo, you should shoot them. Combat should primarily be used to take out targets that are not good at combat.

This is a good approach - a rock paper scissors approach.
 

Petitioner's City

Gang Hero
Nov 15, 2017
914
1,517
133
Edinburgh, UK
True, I do agree Kiro - it is a broken balance. We need to nerf shooting, and as I've said, I think a key issue is at the start of a campaign everyone just is too good. Like shooting is not hard, but it should be harder in the game, as these gangers aren't marksmen. Enforcers had much better starting stats, as it both gives them a lot of room to develop, and gives a sense of trained but not exceptional marksmen and close combat fighters (versus untrained gangers and juves).

In my own campaign ruleset, I recently removed Improbable shots, but I wonder if always hitting on a 6 is too good for shooting, so I'm dithering on using the oldmunda Improbable shots table, which feels like a good gradient.
 
  • Like
Reactions: Spenetrator

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
The problem with 'rock paper scissors' is that shooting fighters are just as good at shooting each other as they are melee fighters.
So you haven't got a triangle at all. You have one deformed limb.
I agree that shooting is too powerful, but that is a separate topics. There are a lot of strings that we could pull to equalize shooting and melee. Making it easier to connect the melee charges is one of them, but that'd be a separate topic.
There is no Rock-Paper-Scissors dynamic in the game because mainly you kill people with melee or shooting. In my opinion melee should be a lot cheaper than shooting (to allow inclusion of meatshields who soak up shots via target priority tests), but just as deadly or even more deadly once the melee begins.
 
  • Like
Reactions: Petitioner's City

Kiro The Avenger

Gang Hero
Apr 4, 2018
1,224
1,588
128
Bristol, UK
I agree that shooting is too powerful, but that is a separate topics. There are a lot of strings that we could pull to equalize shooting and melee. Making it easier to connect the melee charges is one of them, but that'd be a separate topic.
There is no Rock-Paper-Scissors dynamic in the game because mainly you kill people with melee or shooting. In my opinion melee should be a lot cheaper than shooting (to allow inclusion of meatshields who soak up shots via target priority tests), but just as deadly or even more deadly once the melee begins.
I agree that that's how to solve melee.
But alternating attacks does precisely the opposite. It makes melee less deadly and less reliable, because every time there's a good chance you just let the enemy kill you for free.
 
  • Like
Reactions: JayTee and Al_Weeks

JawRippa

Gang Hero
Mar 31, 2017
1,069
1,380
133
Saint-Petersburg, Russia
I agree that that's how to solve melee.
But alternating attacks does precisely the opposite. It makes melee less deadly and less reliable, because every time there's a good chance you just let the enemy kill you for free.
Even if you die to reaction attacks, your opponent still has to spend actions doing a coup-de-grace on your fighter, since reaction attacks do not grant free consolidate/coup-de-grace. Also I think that alternating attacks make melee champion's fights actually interesting compared to snoozefest that we have now. Obviously any changes to melee should be done after shooting is heavily nerfed...
A lot of guns are too easy to hit with and they can be used effectively on long and short range. It'd be nice to cut damage of shooting, so shooting is consistent (due to ability to attack targets from safe distance), but finite (jams should be a constant threat) and lack lethality, while melee is more unpredictable and lethal and lasts indefinitely throughout the fight.
 
Last edited:
  • Like
Reactions: Petitioner's City