Featured Post

So It Begins, Part 9: Eomer's Return

  Good morning gamers, This is a scenario we did in Fantasy Fellowships and if I'm honest, this is one of the least balanced scenarios i...

Thursday, March 17, 2022

Dueling Calculator: Building A Monte Carlo Simulation, Part II

Good morning gamers,

In our last post, we viewed how to set up a Monte Carlo simulation to give us a better understanding of how models do when fighting each other (rather than just generating statistical probabilities). We only focused on the dueling aspect of the game because . . . it was really involved. Today, we finish the discussion (for now) to get the basics of wounding into the simulation. Let's get started!

Building The Simulation: Setting Up Inputs

In our last post, we had to roll dice for both sides, identify penalties applied to those rolls, determine if boosting rolls is a good idea, and then figuring out the highest roll for each side. More than 60 columns later, we found "the winner." Today, we don't have to do as much - but we DO have to generate a ton of columns. Rolling to wound in MESBG is done by one side during each duel (generally speaking - some models, like Sharku or Sharkey, can deal damage to their aggressors when they lose a fight) and it's often resolved within a few seconds of viewing the roll. But here's what's going on during those handful of seconds:
  • We figure out how many dice each model will be rolling (factoring in whether or not the enemy is trapped and we double our dice);
  • We identify the wounding difficulty we need to reach to successfully wound;
  • We roll the dice and look at our initial results and identify any successful wounds;
  • We identify if we have any rerolls that we can use (which could be rerolling 1s or rerolling any failed To Wound rolls);
  • We identify which dice have succeeded in wounding after rerolling;
  • We identify how much Might we'd need to spend to boost a wounding roll; and
  • We spend Might to boost wounding rolls that we want to work.
In order to walk through these steps is going to take columns - lots of columns for clarity - and we're going to have to make a lot of assumptions that go back to our inputs page. 


Our inputs page has the following information that we can use:
  • Base Attacks
  • Charge Bonus
  • Whether the enemy is trapped
  • Wounding Difficulty
  • Whether the model rerolls 1s To Wound
  • Number of full rerolls the model receives
  • Might allocated to boosting wounding rolls
The important thing to keep in mind is that the "Trapped" section is independent of knock-down - is my opponent trapped based on his innate positioning (regardless of whether you charged the guy or not). Setting this up right is important since we're rolling to see who has priority.

Notice that we've also set up different inputs for rerolling 1s and full rerolls - since we can reroll as many 1s as we want when we Feint (or however else you reroll 1s) and we usually have a limited number of rerolls, we will need to keep track of which dice have been rerolled on 1s before we do full rerolls.

Okay, let's get into it!

Monday, March 14, 2022

In Defense Of: Ruffians

Good morning gamers,

In our last post, we looked at how to make Harry Goatleaf work better in your forces (that is, in Sharkey's Rogues or the Chief's Ruffians LL). Today, we're continuing our discussion of Sharkey's Rogues by looking at the lowly Ruffian - quite possibly the worst warrior profile in the game (maybe second-worse to Hobbit Militia).

Why NOT To Take Ruffians
Photo Credit: Warhammer Community

Ruffians have . . . nothing to recommend them in their profile. They are F2/5+ (both stats are below-average), S3/D3 (average Strength, below-average Defense), 1 Attack/1 Wound (average), and C2 (below average). When compared to other Evil warriors, you can see that the price difference between a Ruffian and an Orc, Goblin, Uruk, or Man is pretty stark - and the stat differences are pretty stark too:

Evil "Core" Warriors . . . Ruffians are cheap and that's about it . . .

The only model that costs the same as a Ruffian with a piece of gear (6pts) is an Orc Warrior - and the Orc has +1 FV, +1 (or +2) Defense, and has access to Piercing Strike instead of Stun. Orc Warriors are available in a lot of factions (Barad-Dur, Mordor, Isengard, and similar models are available in Angmar - to say nothing of several Mordor-based Legendary Legions) and have access to spears, shields, and banners (all useful pieces of gear which Ruffians do not have).

There are two models that are cheaper than Ruffians (Moria Goblin Warriors and Goblin Warriors from Goblin-town) and both have similar profiles (lower Movement, potentially higher Defense) but have access to spears/supporting rules (which Ruffians don't have). Goblins (either kind) are at least 1pt cheaper than a Ruffian with a piece of gear - and despite those savings, the Ruffians can only gloat over an extra 1" of movement.

Thursday, March 10, 2022

Dueling Calculator: Building A Monte Carlo Simulation, Part I

Good morning gamers,

Last year, we did a few posts on how to build an Excel-based calculator to determine your probability of winning duels (without boosting and with boosting) and the probability of wounding (either the average expected wounds you'll deal or the likelihood that you deal X wounds to an enemy). These posts, while very useful in providing a statistical view of combats in MESBG, are incomplete. You can tell the calculator that you assume someone has charged, or that someone is within range of a banner, or to some degree that someone has help winning the fight. But no matter how hard you try, there's more information that we might want - and it would be helpful if we could see examples of a combat. For this, we need a simulation - more specifically, a Monte Carlo simulation. And today, we build that.

Monte Carlo Simulations: Repeating A Scenario Many, Many Times

Monte Carlo simulations operate on the following theory: if you repeat a situation thousands of times, you can derive trends from them and get an idea for what to expect in a given (or changing-but-controlled) situation. This is particularly useful when we're rolling dice - you can mathematically figure out your likelihood of getting certain results statistically (as we've already done), but when you have dependent rolls (like dueling and wounding), it's sometimes helpful to see them in examples. And that's what we're doing today.

We're going to create a scenario (two sides fighting) and repeat the scenario 2000 times. Instead of assuming priority, we're going to roll for it - and that means we need to change the scenario inputs a bit. We already have things like Fight Value and number of dice to win the fight and wound on each side, but we need additional information too.

We're also going to allow one side to get up to five model types in the fight (Vertical) and the other side up to two model types in the fight (Horizontal). Because fights break up into one-on-one when multiple models engage from each side, most situations will involve one model on one side and one or more models on the other. The lone model could have a spear-supporting model (or 1-2 pike-supporting models), so I've let the Horizontal team have a second profile option. There may be situations where you'd need more than two profiles for the lower-model-count side (such as an Isengard-Easterling alliance with an Uruk-Hai Berserker supported by an Uruk-Hai Warrior with a pike and an Easterling Warrior with pike and shield), but for simplicity today, we're going to limit the scenario to 5 vs. 2.

We're also going to build out the dueling roll portion only - mostly because the process of accounting for basic dueling and wounding rolls is incredibly complicated and takes a LONG time to explain (I tried doing both dueling and wounding in one post and it was WAY too long). So, keep your shirts on - we'll be back next week to deal with the wounding portion. Keep in mind, though, that we can still learn interesting things from the dueling portion only (which we'll talk about at the end).

Finally, we need to acknowledge when making a simulation like this the first rule of spreadsheet tools:

Columns are free.

Columns in Excel continue to be added as you journey further to the right. We're going to generate a TON of columns in this spreadsheet - and by not constricting ourselves to a certain number of columns, we can keep our granularity as we go and it will keep our cells clean and our work simpler in the long run. So with that, let's dig in and start simulating some fights (well, the dueling portion anyway)!

Monday, March 7, 2022

In Defense Of: Harry Goatleaf

Good morning gamers,

Today we're continuing our discussion on models that don't receive much love from the community. Today's model has been described by Matt Iverson from the Unexpected Podcast team as "the only guy we are less likely to put in an army [than Barliman Butterbur]" - yes, we're looking at Harry Goatleaf, member of the Wanders in the Wild, Sharkey's Rogues, and the Chief's Ruffians Legendary Legion.

"What do you want?"

Why NOT To Take Harry Goatleaf

Harry Goatleaf costs 40pts and is . . . well, not very impressive. Here are a few heroes that are around 40pts - and for thematic reasons, I have picked only Hobbit characters (see our post on Denethor earlier in this series for a larger sampling):

Lackluster profile much?

Harry Goatleaf gets the usual critiques that surround models from the Wanderers in the Wild list (can't lead troops, often make you lose your army bonus) - and to boot, his profile is really weak with only 1 Attack, 1 Might point with no specialized heroic actions, and Defense 3 with 2 Wounds/1 Fate. As we saw in our reviews of Denethor and Smeagol, there are heroes who have decent combat stats for 40pts or less - and Harry isn't on that list.

Harry also has a Lantern that gets some critique - like Blinding Light, it illuminates a 12" radius around Harry Goatleaf. However, unlike Blinding Light, you can't turn it off AND there is no "hit-on-6s" anti-archery bubble from it. Instead, you just illuminate an area - your models, your opponent's models, etc. This will only come into play (for good or for ill) in one scenario out of eighteen - but in that one scenario, it does nothing for you. The fact that you can't turn this off can be a huge liability - especially if you're only D3 with 2 Wounds/1 Fate in a scenario where the enemy will get +1 To Wound and can shoot at you from as far away as they wish. With all this in mind, he's more likely to be run in Sharkey's Rogues or the Chief's Ruffians LL instead of being allied into another force from the Wanderers in the Wild.

Thursday, March 3, 2022

Fyrd Formations: Reforming Shieldwalls

Good morning gamers,

In some of the recent formations articles (fighting on the ground and how to fight against shieldwalls), we've looked at shieldwall formations and how to both use them unexpectedly and how to fight against them. Today, we look at how to evolve your shieldwall formation to gain an advantage against an enemy shieldwall formation (or how to use a smaller shieldwall to take on a larger shieldwall). While we'll be using Minas Tirith and Mordor armies to illustrate, any armies that have warriors with spears and shields can do what we're going to do here.

Note: many of the principles we're going to cover here have already been brought up by Rythbyrt almost two years ago in one of his posts on spears and shield formations - check it out here. We're going to take some of the thoughts Rythbyrt brought up way back when he wrote that article and apply it using the principles that were used historically with the Fyrds in Saxon England.

What Were Fyrds?

A "fyrd" was basically a militia force of farmers that were trained by a local lord in Saxon/Mercian England. The Anglo-Saxons trained their local militias to fight in blocks with shields held tightly together and spears to thrust outwards at anyone who tried to break the shieldwall. The fyrd system worked really well against Scots/Picts, Irish, Welsh, and Danes - but didn't work particularly well against the Normans (something we'll actually be looking at next month).

In England during the first millennium, the usual modus operandi of military forces (be they Anglo-Saxon, Dane, or otherwise) was to form a shieldwall and ram up against the other guy's wall. You then strove with them all day, hoping to break a hole in it. If you managed that, then you could attack the men who were shoving with all their might against your fellows and kill them without them being able to put up a fight. If your wall broke, you were probably dead before you knew the wall had been breached. While there were formations designed to break through shieldwalls (like the Svinfilking), as a general rule, you just shoved our shieldwall against your opponent's.

Most of the time, shieldwalls succeeded or failed based on their length or depth - have either factor out of proportion to your opponent and your line could end up breaking quickly. If you had a short shieldwall and your opponent had a longer shieldwall that had the same depth as yours, your opponent would start to wrap around your men and kill you from the sides (bad day). If the shieldwalls of two armies were the same length but one was deeper than the other, the mass and strength of the deeper rank might break through the thinner wall. Either way, it pays to have numbers.

This principle (longer battle lines or larger numbers) actually does translate into victory in MESBG - but it is possible for a smaller force to defeat a larger one if it positions its troops well. To understand how that works, let's look at three different ways you can form up a shieldwall and why they might be useful depending on the size of the shieldwall your opponent brings to the table.