Charles Soong. Feb 22, 2019
During the masters of the recent CaptainCon, an observer asked me a question that I couldn’t answer at that time, and put the train of thought into motion to try to answer that question. The question, as I recall, was, “How do you play so fast and always make the right decisions ?” And the person asking that was Etiene, whom I played in the Champions event 2 days ago. To myself, I don’t think of me as making all the right decisions, and every single game I could still point to many decisions and placements that I could have done better in retrospect. Therefore it was a shock to me that he would think I make all the right decisions, especially given how I played that particular game he was watching. It was Mag2 vs Thag1, where I forgot the advance move benefit of my theme, then I forgot to feat top of 2, which resulted in an assassination run of ~35% against Magnus which I survived by luck. (To people observing my games, I often make major mistakes not expected of a ‘good player’, whatever that means.)
In trying to answer that question, I thought back to the five finger forums, where I’ve been posting short summaries of my con games in a thread by Rob Shepard regarding merc dojo and bathhouse.. stuffs. I realized a significant number of them takes the form thus; I deploy like this, matching this unit with that unit across the table, we move up, shuffle around for 2 turns, and then he ran out of models. The end there was the key, he ran out of models, because in playing mercs I rarely get to assassinate(*), so the game plan is usually: kill every opposing model, then win scenario. My friends from Sweden tells me the same thing; the game is attrition. My merc game sheets, of the wins that wasn’t the result of Ossrum or Magnus locking the opponent out of scenario, often have above 100pts destroyed for that reason.
(*) A few things to clarify about this statement. First, I don’t build lists with primary win condition as assassination, or even a strong consideration in assassination; I build lists to kill models. Second, when given an opportunity for assassination, I normally don’t take it unless the odds are above 80%, because that is the ballpark chance I aim my attrition game at. Third, mercenaries don’t lend themselves as a strong assassination faction, as least I think so.
So how do we go about the business of killing every opposing models that matters ? I think the short answer is favourable trading. Where your pieces kills above their point (overall, not individually) while the opponent kills below theirs. To achieve this, allow me to break it down to its components.
Positioning – Otherwise known as threat ranges, is not a static number associated with each model, but an overlaying threat range of all the models under all combination of spells, feats and effects. A model can often have multiple threat ranges, projecting multiple damage outputs at each range. And while it may not be acceptable to give him a high value target at his highest output threat range, or possible to stay out of the threat entirely, you have to position your models in a way to minimize the opponent’s models’ output. Let’s use some examples for this. Bare in mind that while each example may seem simplistic, we are after the combined result from understanding all of them to maximize our trading ability.
Ex1. Molik Karn, with paingiver support, have a melee charge threat of 11” through open terrain. At this range, his melee output against a statline of 10/19, like a nomad, is (to simplify, we assuming it hits all the 3’s to hit) PS15 weapon master vs ARM19, free charge attack does 10, each subsequent attack does 6.5, and will on average kill the nomad’s 30 boxes, even assuming the shield goes down last, on the 4th attack (at 29.5 ballpark damage vs 30 boxes, plus the buckler consideration). So, the nomad will be toast, but karn will at most be 7” from where the nomad used to be if he animus, and if I can setup nearby models to ensure that I kill Karn after he pastes the nomad, then it would be a favourable trade. On the other hand, if I put the nomad further from Karn such that he has to spend sidesteps to get to the nomad, then I can even further the trade, because without the initial charge, 1 sidestep Karn’s output drops dramatically and will most likely take all 5 of his remaining attacks to kill that nomad, putting his final position in an even more predictable light. So, putting the Nomad 11.1” away from Karn seems like a decent plan. If the first sidestep model has decent defense, it might even cause a miss and another force, or if the nomad is 13.1” away, either way meaning Karn gets to it with 4 attacks, not killing it on average and without animus, that would be the best trade, assuming you have to put it in threat range. Deciding which setup will result in a favourable trade depends on knowing all of these numbers
Ex2. The feared Judicator, with its sign-and-portent attacks, has 3 common threat range numbers. A melee charge threat of 10”, a walk and shoot rocket threat of 19”, and a walk and spray threat of 15”. The latter 2 are linked because they are both range attacks and is mutually exclusive with the melee threat, although the melee threat has the highest output. Most of the turns it will shoot, so you need to understand how your models withstand its 15” and 19” threats. Breaking it down, the 19” threat is inaccurate, and rely on direct hits against hard targets and blast damage against soft targets, does not ignore stealth by default, and even with its sign and portents can often miss against models benefitting from cover or concealment. Forcing him to boost the hit further reduces his ability to boost damage. On direct hits, the rockets can be shield guarded to lower value models like an Ogrun or a Valkyrie, or at least spread out the damage among other heavies to minimize crippled systems. There are several defense against the rockets. The 15” threat is harder to mitigate because it has higher accuracy, is a spray and thus ignore majority of range defenses such as stealth and shield guards and intervening models, therefore you have less options to manipulate this attack’s output. The good news though, is that if the judicator can spray you, you are now within 10”, and majority of heavies can threat 10” (or at least you should heavily consider taking caster/heavies combo that do), and should they survive the range barrage, then you can plan for the counterattack, as the judicator is not a durable model with most casters. So the trick here is knowing what kind of range defense you have, whether or not your counterpunch model can survive the combined rockets/spray attacks and then hit back, or if you put multiple heavies within that threat range, 2 out of 3 of them might get to counterpunch.
Knowing the tiered threat range and their respective damage output is one way to use positioning to your advantage. Another would be spacing.
Ex3. A slayer (10pt) vs 9 steelhead halbs (9.9 pts). In a vacuum, the advantage is with the steelheads, and the way you press those advantage is knowing the slayer’s charge threat range is 10”, with 1” reach at its final position, limited by the charge angle. The halbs, with their 11” threat, can easily position in a way to bully the slayer like this; forward 3 halbs comes within 7.1~10” of the slayer, space out laterally at least 3”, and the remaining ones staying outside of 10.1” of the slayer. The 7.1” spacing means the slayer can not walk to between 2 halbs and must charge, and the 3” spacing means that when it charges, the change in facing will cause it to have LoS to only 1 halb after the charge movement. If the slayer comes in and kills 1 halb, the remaining ones can almost all charge and with good odds of obliterating the slayer. (each individual one need a 6 to hit at dice off 6, do single attacks at those odds). If the slayer does not commit, it must back up to avoid being charged by 3 halbs with the remaining as 2nd wave, thus ceding table space and scenario presence. This is a very one sided trade and illustrates why no one plays slayers against infantry at their base stats without layered buffs provided by casters like Gaspy3. If this unbuffed slayer goes in to kill 1 halb, the remaining 8 charges it and have a 39.4% to kill it, position in a way that majority of them are at 2”, the crippled slayer kills 2, then the 6 remaining one most likely kills the slayer. Because the advantage is so tilted, the halbs can tolerate less-than-optimal positioning and still win this.
Ex4. A bit more nuanced. A full unit of immortals with UA, plus partial cost of an ancestral guardian support (cashing in at ~18pts) vs 16 halbs (a min and max unit, also 18pts). Here the AG is just there for the threat range support and will not actually contribute to the positioning. So you have 16 models against their 11, numerically you can trade 1 for 1 as the merc player, as long as you keep the overlapping threat range game in mind. Again, in a vacuum, on the approach both sides threat 11” prior to triggering vengeance, and as the halbs have the numerical advantage, will offer up the ‘bait’ first. For Occam’s sake, I’ll just generalize the 16 halbs as one unit. The halbs would put up 3~4 models within the 11” threat, while the remaining stay out of 11” from every immortal. The immortals now must choose to take the bait or cede table space, and since scenario is always a thing, usually are forced to charge in. So an immortal at mat6 vs a set defense def13 halb, need a 9 to hit. On average 1 immortal vs 1 halb will not kill it. So for average, you put in 2 immortal per halb, so 6 goes in, each pair have a 47.8% chance to kill their charge target. Let’s say 2 of them kill it, and the remaining 4 immortals and the UA stay out of charge range of the halb horde. So now the odds are 11 models vs 15, but 6 immortals are in charge range now. Let’s also say the immortal UA spend a soul to pop defense and put them at 13/17. 8 Halbs charge in with some overlap, plus the 1 remaining guy who was up there doing a non-charge attack, 6 halbs hang back. Charging halb at mat7 ps11 each has a 65.5% chance to kill an immortal, so 8 charge, with whoever didn’t have a charge target combining with other non-charges, and the 9 halbs fairly easily kill 5 or all 6 of the immortals. It is now 15 halbs vs 4 immortals + UA and you can see how no amount of reroll on the immortals side can win that fight from there.
Ex4.2. Let’s say the immortals don’t charge in, and do pop defense and stay out of majority of halbs range. Then halbs press the position, 3 forward ones charge in and kill 2 at 65% per pop. Each unit leader maintain cmd7, while the remaining halbs stay out of 14”. Immortals vengeance to kill the 3 halbs, charge in 4 to kill 1 of the unit leader, and the attrition goes poorly for immortal from here again.
This example spans 3 rounds because that is how most infantry skirmish happens, with pieces being put forward to hold zones or as bait and others as 2nd or 3rd wave. Spells, feats and other support abilities all swing those math, and scenario might force you to overcommit to hedge a favourable odd against the opponent scoring, but the fundamental is still the same; position yourself to minimize your losses, make the opponent’s position predictable, and counterpunch with greater impact. You should note that in a 10 vs 10 situation, you should almost never place all 10, or even a majority of your models within threat of the other unit unless you have specific defensive abilities such as stealth, polarity field, or Rask’s feat.
Matchup – rock paper scissor. Warmachine was often compared to rock paper scissor in its reduced form; battlegroup beasts gun line, gun line beats infantry, infantry (especially weapon master variety) beats battlegroup. Every model/unit in the game has a certain matchup into another model/unit. Given that both players knows how to position their models against their prospective target across the field, model stats and abilities determine the matchup odds. Matchup changes dynamically due to buff, feats, terrain, and other models intervening across the board and must be evaluated turn after turn. Often you have a group of models fighting another group of models, and the expected output or effect of each individual model in that ‘isolated’ scrum can be different from the model next to it. Knowing the odds of each of your attacks and how they stack up to impact the table takes some memorization of unit stats and dice odds and some hedged bets, but that’s what you have to do to make the right choices as the table changes.
Ex5. Taking the above example, an equal point of slayers vs steelhead halbs, the halbs crush that matchup. But the halbs don’t always have buffs put on them on behave of them being a cheap infantry unit, while the slayer is often only taken in a list that layers buffs on them. A 10pt slayer in Gaspy3’s battlegroup vs 10pt of halbs, and that matchup swings. The slayer now have a 9” walking(facing) threat, 12” charge threat (which exceeds the halbs) with pathfinder (also something halbs don’t have), and the unyielding benefit means even if it get charged by 8 halbs it will probably survive. The halbs now need 6 to hit and is dice off 8, and needs to do 2-man CMA to have better odds, and the 4 pairs end up at 3.9% to kill it. Average damage is 18 assuming all hits with most likely no crippled systems. The slayer kills 2~3 (now almost ignoring free strike at arm23), the remaining 5 might get 2 charges plus 3 non-charge, doing a 2-man charge and 3-man non-charge, doing another 5.5, probably leaving 1 arm and head attack. This seems like it still benefits the halbs, because they are one of the most cost efficient infantry in the game, but if the terrain or other models blocks part of the halb charge, or an overlord lurking behind the slayer comes forward and sprays that cluster, then the wounded but functional slayer could easily change role and go on to cripple another heavy with his ps22 combo strike. The other factor is that the halb takes up more space and is easier to deny with terrain or other models, and their positioning is harder to duplicate across the board. So 6 slayers vs 54 halbs could go either way based on the board.
Ex.6. That same slayer (who is now bored and irked of being used as an example so many times), vs a same number of gun line infantry, with carapace in play, is not even a fair fight. Slayer will advance up with terrain advantage, eats 1 big CRA (because at those odds it’s the best output) for a few damage, goes in and kill a few dudes, eat a smaller CRA, rince and repeat. Slayer is already favoured in this matchup even prior to gaining carapace, now it becomes a cakewalk.
Ex.7 A derp turtle, vs almost any heavy of less than 20 pts, or even 2 heavies at 10 pt each. The turtle can use its superior range and reposition to stay at 15” from the target while peppering it with pow15 guns, and when the opposing heavy comes in, wreck it with its powerful melee output. Facing 2 budget melee heavies, it can wreck or cripple one, take the beating of the other, and use the rage token generated from that to wreck the second one. Its output is just that good that almost no unsupported model at the same price range can top it. A unit of dawnguard sentinels, with vengeance, iron zeal minifeat, and being 5 pts more expensive, might still struggle to kill the turtle in a vacuum, because at that point it depends on the terrain and other elements. With just those units, a turtle kills 1~2 with the gun on the approach, stay out of vengeance range. Sentinel runs forward with minifeat. The turtle charges in, with impact and tail kills a cluster of 3 man and repositions to deny charges from another 2~3. Now the remaining 5 charge and 2 non-charge has a 51% to kill it, if it doesn’t then the unit is most likely wiped. This is a 17pt vs 22pt matchup and is barely even on both sides, with the sentinels supposedly having a good matchup into a heavy target. Derp turtle is OP.
The matchup we usually talk about is a list against another list, but that is basically the sum of all individual matchups of every model in one list against every models in the other list. A matchup between 2 units in a vacuum can tilt one way, and with terrain or a buff can tilt the other way as well. So you will have to look at the table and imagine how each list will play, and how you could counter that with either of your lists, with models being offered to absorb enemy attacks and others as counterplay.
Scenario – the equalizer. Without scenario, the game breaks down into a micro series of positioning and matchups, and the player who is more familiar with them will not only be able to win the game, but win it with far smaller casualties. Scenario forces you to commit certain positions or resources in a way that may not be optimal, and it is your job to make that as uncomfortable for the opponent as possible. In the best case scenario of scenarios, you lock the opponent out of scenario by pushing the line of battle outside of the zones and with certain defensive abilities so that your front line doesn’t get completely destroyed before you can score enough points. Scenario and killbox makes the game more predictable and you can use that to narrow down your opponent’s responses, and use that to optimize your positions. To do that, you need pieces that can clear a zone or a flag with certain, decent odds.
Ex.8. A recent game I played at 2018 WMW’s invitational. There’s some background to the game(*) and how it arrived at that list pairing and individual models, but let’s look at the positioning and the reasoning. This is a multi-turn game plan based on the scenario on the table. Thanks David for being used as an example.
I won the first turn, which was crucial, because how far Stalkers can reach if they go first. I move up aggressively, putting 1 talon from each flank within the threat of stalkers (talon and stalker both threat 13”), 4 nomads fairly central to respond if he comes in. Of course, if he comes in, 2 stalkers with their base stats will not kill 2 talons under Mag2, and if he feats it won’t do anything useful either. However, if I had put 2 nomads also in range, he might be tempted to feat bottom of 1 to kill 2 heavies and 2 lights. These are the options I have, and based on my choice will change his options. Blue arrow denote within 13”, red arrow outside of 13”.
So, the offer of 2 talons is obviously too small, plus he has not had ritual sacrifice focus available. He shuffles around and positions for an alpha. Due to the fact that my jacks do not have pathfinder (even with the pathfinder OBJ), he puts some models in the threat range of talons and get ready for the next turn. It was important to note that when he positioned his models, they were based on having threat range to cover the entire zone. For example, stalkers have measured 13” to either back corners of the square zones. This is crucial because if he did not threat the entire zone, then I would easily deny the scoring option there. He made sure he has 2 stalkers, a wraith engine and some extra for each zone, enough damage output to probably kill most of my battlegroup if I commit.
On my turn, options are limited on a spectrum of how much stuff I commit to the zones, which in response will draw in a proportional amount of his models. Both rectangle zones and my OBJ are in play, circle zone is not. (the circle zone is not in play because he is spreading out to the flanks to avoid Magnus’s feat) It’s unlikely that he hold the feat for another turn because I have sniper guns, but if I undercommit, the wraith engine might not need to go corporeal, which will limit my retaliation. I would rather be able to cast harm than have to magic weapon a heavy, which doesn’t kill it at all. So I put in 1 talon on the right flank, 2 on the left, nomads shuffle back a bit because they threat shorter than stalkers and I don’t want to lose them yet. 3 talons and the objective. The odds are that 1 jumping stalker, with feat, doesn’t kill a talon. So that he would need to commit at least the left wraith engine, both stalkers there, and 1 stalker plus engine on the right. Anastasia came in the right and took out an arm of the right most stalker to try to stall that one. She was conscious to stay out of 2” from the wraith engine so she doesn’t become a machine wraith.
So he came in hot. 2 cav models and the stalker came in and finished the right side talon. Right engine killed the OBj. Left side, 2 banes and a stalker left the first talon crippled, forcing the wraith engine to commit. He did and finished that talon and did some to the other. The other stalker finished the last talon. Cryx scored 3. From here, I was able to concentrate my resources, which anticipated the stalkers’ positions, and clear out most of the models in the left zone. Right side I killed a stalker and did a bit of damage to that engine. Mag2 moved to a position to catch a few banes and the right engine in feat so that he doesn’t die and Tart doesn’t come closer.
After killing most of his key models I was about to win the attrition 2 rounds later. This game in particular, virtually all model positions were derived from scenario zones. He positioned to threat scenario, I positioned to anticipate that and to push his commitment high while limiting my losses. Due to the fact that his models are all melee based and does not move my models (as long as I don’t let him generate machine wraith), his position after killing my models can be predicted. There were some poor dice on the left flank that forced the 2nd stalker to commit, but in hindsight would not have changed the overall game state much.
Had I lost first turn, the game plan would be drastically different, but still revolve around scenario and piece trade. I would be forced to offer more models on turn 1 or 2 to try to tempt him to feat, and he can easily hold back and force me to come into the zones. He can spread out to minimize my feat to one flank only as well and I almost can not stop him from generating at least 1 machine wraith. Again, scenario will drive the positioning. Each turn I must put sufficient models in his threat to tilt the odds into him feating, or else I risk being locked out of the zones entirely and lose on turn 3 or 4.
(*) The backstory of this game, was that I knew I was most likely fighting David on round 2. I have not played the Magnus-Skarre1 match before and in my head feels terrible because he match threat with my talons and out-threat my nomads. So I modified my usual Magnus list to 4 nomads and 3 talons, with the talons being a threat he must eliminate due to the stall lance, and talons being durable chaff I can throw away. 4 nomads gives me enough hitting power, and Hutchuk/Ragman/Aiyana help me deal with the ARM buff of her feat. This was a list tailored for this matchup, and had I gone 2nd would still be strongly favoured for him.
All the previous talk of positioning and unit matchup must be considered in the context of scenario, as it is pointless to win the models fight but lose on points.
Sometimes, losing first to go first on a live scenario against a fast army is so crushing that it puts you on the backfoot the entire game. This happens when the opponent on the top of 1 can move up and project threat well past the scenario elements (typically 30” from their table edge) with enough forces that makes moving forward difficult. You would move up, maybe take some shots, present some targets, and then based on the amount of present and the relative threat ranges between the 2 armies, he can either go in and lock the line of engagement near your deployment, or back up out of your range and force you to sacrifice models to contest. The game has pushed this aspect so far that many games are considered won or lost on the initiative roll. (won or lost being, it swings the estimated percentage by maybe 20% or more). Lists like Irregulars with Ossrum/Mag2, Gaspy3 slayers, Xerxis2, Several flavours of Devours’ Host, Mage hunter theme in Ret, and many others, all look to capitalize on projecting threat almost to your deployment line to force the scenario that way.
Ex.9. This game was Ossrum vs Skarre1, at SOO 2018. Skarre won the roll to go first and ran up. Easily threat to the back of my zones with several models. I moved up and positioned in a way as to shield my heavies with my lights, but it did not work and I lost 2 lights and 3 heavies to the alpha because I did not feat, as the line of engagement was near my AD line and I could not reliably push in to contest his flag and his zone (also due to some mistakes and poor plays).
How could have that game gone differently ? There are a few ways based on predictable model positions. Plan A. Move forward but stay out majority of his threat range except the lights. This forces him to either back up slightly (as the spray bunnies out-threat his stalkers and can easily cripple a system on a spray), or feat to come in and only kill lights. If he backs up, then we play a dancing game from there and I get to decide again how much to put forward and whether or not to feat. If he comes in then I should be able to counterpunch with my feat and more heavies, and be able to weather the following turn much better. Plan B: move up into his threat range and feat. This will rely on better positioning than I did here, but I should be able to avoid losing more than 1 heavy, and be in a much better position to counterpunch than what happened. In reality I did neither and the poor positioning plus lack of feat meant he had a productive alpha and eventually scored scenario win by keeping me out of his side of the board. (there was also the fact that Annie rolled snakeeyes to kill his bane officer, then got Tart’d) If I did feat forward I would need to be even more forward than in the picture, because his response could be to come at me, or wait out my turn. If he wait out my turn, then he out-threat me again and I don’t have pathfinder so I need to be either forward enough to fight a scenario game from there, or I’ll be forced to pull back and weather his alpha again.
Another way to play from the backfoot is to offer them an assassination. Imagine if you figure the long game to be at 25%, then it would make sense to offer them a 50% assassination. If they take it then you play the game at 50%, and if they don’t then you might have gained better positioning by doing so.
Plan ahead – putting it all together. After all the talk of position, matchups and scenarios, now you have to consider all of that, and imagine how the turns play out 2 or 3 rounds forward. Terrain looks a certain way to limit model movements, which model will score zones, how will they matchup against the models the opponent will try to score zones with, how does that individual fight plays out with or without buffs. Sometimes a fight on the flank between a few models can swing on a single die roll, and turn a favourable match around, you need/should have a backup plan for that. Based on the first player’s deploy, second player should deploy to counter them, then first player’s position presents options for the second player to choose from, and so on. It takes a lot of knowing what the odds of each branch and sub-branch is to predict correctly how you should play this current turn, and certainly no one can do that repeatedly and accurately. Nonetheless it is an exercise you must engage in if you want to play the game within your control.
Ex 10. Recent game at ATC finals. Thanks to Tom for being this example. Background info was that it is spread the net, and no one on the team wants to play against Magnus2 on the opposing team. Personally I know Magnus2 better than my team mates so I took that match. Knowing Magnus’s plan is to feat on a large portion of a melee army, the plan is to give him the worst table (which we did as B position) and spread out to the flanks as much as possible. He won the roll as usual, and chose to go first, and was given the side with many rough terrain.
He ran up, hemmed in by the forest and the water. Magnus failed a charge to get up the board 8”. I positioned like so. On the left flank, one model of the immortals was in 10” run range to get into his zone, the central supreme guardian (PICKLE RICK!!) is within charge range of his Toro, supported by an agonizer. Bronzeback stayed out of the zone and just out of 20” from Magnus, and the right flank immortals also stayed out of charge range of the eliminators. An archidon and a gobber chef got assigned “Anastasia Duty” by putting their backs against the board edges to prevent backstab. The position here planned against Magnus’s feat by giving him almost no useful models to feat on, the surpreme guardian with the agonizer has virtually 0% chance of dying to a 3 focus Toro and I want him to come at it. I was able to position this way because I know the stats of Magnus’ models intimately, their threat ranges, damage output, and what he is trying to achieve, and planned my position to deny his plan while keeping the scenario live. Due to the poor feat prospect, Magnus did not feat, and I was able to clear my zone and start scoring on bottom of 2 before he could. Here, I had plan to contest, score, bait, avoid the control feat, and dictate his movement, partially because the terrain made his moves even more predictable.
Let’s wrap up; understanding positioning and threat range (and the reduced output at longer threat range), individual model matchups based on their stats, how those numbers change with spells and feat and sometimes terrain, how that overlays with scenario elements, to allow you to plan ahead and predict the game state out of several possible outcomes. Taking this experience back to list building will let you improve your lists iteratively and be more consistent on the table, and that is how we improve as players.
p.s. Apologies to the unwilling subjects in my examples. Those are the games that came to mind when I was trying to get the point across.
p.s.s. I do not practice what I preach; there are ideal circumstances when you know all of those things, and there are the majority of the times when you have to wing it on the fly. That’s why targeted practice is so crucial.
p.s.s.s. This is not meant to be a comprehensive guide to, well, anything. This was spurred by trying to answer a question from my perspective on a complex subject, and I did not expect it to balloon to this size. If you manage to slug through it, good for you, let me know how you think.
p.s.s.s.s. I’m not a robot