Skip to main content

Don't blame the dice! Understanding probabilities

 

Something I often see is people on Reddit asking for tips on affiliations that don't rely on attacking as they "always roll bad" as well as people getting frustrated and blaming their dice for losing a game. 


1. The Negativity Bias

It's basic human nature to remember bad experiences more and better than good ones. It helped the development of our species to avoid deadly dangers and the like. The chances of getting eaten by Sabertooth Tiger during a game of MCP are rather slim though (even if your models might be getting killed by an enemy Sabretooth). 

What that translates to is that you are much more likely to remember that one critical roll, that went bad, then the 30 others you had, that were perfectly average. 

What can you do about it? It's a lot of work but it can help to keep a full on log of every dice roll you take in a game. If you want to go the extra mile, you could even compare your results to the expected averages. Jarvis Protocol offers a very helpful dice calculator for exactly that. 

In general an attack die has an expected average success rate of 50% before exploding crits as you have four dice faces that are success: hit, hit, wild and crit. Adding in the possibility of a crit adding a dice and that also rolling into a success increases the overall percentage to 56%. 

Defense and dodge dice are strictly worse as you only have 3 of 8 die faces being a success: block, wild and crit. That combines to an average of 38% success before crits exploding and 43% after.


1.a) The explosive nature of Marvel dice 

Crits exploding is what makes Marvel dice so spikey. That's just an undeniable fact (and the reason Hex is one of the very best special conditions to hand out). What that means is that spikes will happen. The more dice you roll the higher the chance that one of those rolls will go off. If you have a feeling like you never really spike go back to the journal. Because what's most likely the thing that really happened is that you spiked at the wrong time. Maybe you spiked a defensive roll where the attack didn't have any success anyway and your spike didn't matter. But that doesn't mean it didn't happen. 

 

2. Decision Making 

This is what decides games one way or another much much more than dice rolls. It's no coincidence that good players top tournaments constantly. Everyone of those will tell you that actually winning an event will always include a portion of luck but it's only a small factor compared to making the right decisions at the right time. It's also a key part of getting better at this (and any other) game to understand when you made a wrong decision and why. But be careful of this:

2.a) Trying the same thing and expecting a different result is the definition of madness. But:

Let's say you have Cosmic Ghost Rider, with his Psychosis Token attacking an enemy Carnage with his Hellfire Blasters. 6 dice energy, counting skulls vs. a single defense dice. And it somehow comes out to 0 damage. You have to ask yourself here: did I make the wrong decision in attacking Carnage? Coming back to the expected averages and we see the chance of not dealing any damage in this situation were a grand total of: 1,77%. In other words a statistical anomaly and nothing you should be expecting to happen a lot. 

The trap many people fall into is stopping to attack when a couple of them flopped. The (irrational but understandable) reasoning here is that "my dice just don't show up today". What you have to remind yourself of is this: dice don't streak. They will always average out to the same results. The amount of rolls you take in a game might sometimes not feel like it but that's down to sample size and nothing else (unless the dice are weighted wrong). 

The same thing applies the other way around of course. Expecting weak attacks to do more than you should and being disappointed if they don't. Let's say the SHIELD Agents use their Pistols attack against Shang-Chi and are in range 2 of him, which gives him his Martial Artist innate superpower. So we are looking at 4 attack dice into 4 defense dice which count blanks as success. The expected damage here is 0.41. So doing nothing is the most likely outcome. 

So what we learn here is that you should kind of know what your expected damage output in a given situation is and value that against the board state. And accept that sometimes the right play has the wrong outcome due to random chance. And if it does, do repeat it.


2.b) Playing To Your Outs

This is a phrase often used to explain that you should think about ways you can impact the board state in a way that furthers your chance of winning. If you're in control of the game this will sometimes feel easy but there's still a key point you have to be aware of. The game ends if someone scored 16+ points (and the other player didn't tie them). Being in control of the game can lead to you being less careful in your play and/or being unnecessarily aggressive. Count the VPs on the table and calculate how many VP each player will score at the current board state. If you're ahead on Points calculate how many points you still need to win and how many you can "let" your opponent have before it becomes a problem. If you're behind on points search the board for ways to change that. And that leads to: 

2.c) Taking Risks

In any given game it will inevitably come the time where one players outs are only going to be achievable through taking risks. That means you have to go against the averages. Take the low percentage to do much damage attacks if they have a chance of gaining you VPs or stand your character on that secure and pray to Odin that it'll survive what's coming. At the heart of it this still is a dice game and not chess.

 

3. Sometimes It Really Is The Dice

With the aforementioned small sample size of a single game of dice rolls you can plain and simply beat the odds (in a positive or negative way). It's important to not get too salty/ overly cocky when that happens. I've tried to completely for go any comments about the dice being particularly good or bad on either side. What I will sometimes do is mock my characters for being lazy or unwilling to fight ("looks like Zemo brought his plastic sword today") or comment on them for doing something really good like "damn Hulk really is mad tonight". If it annoys people, let me know but I feel that is a much nicer way than to moan or brag about your dice and even worse, or opponents. 

It's also absolutely valid (and useful) to talk through the game after it's ended with your opponent to see what the critical moments where. And if dice rolls where a part of it. If they were take the time to calculate the expected outcome. It will help you understand better if it was the right decision based on percentages, a calculated risk or a lucky punch that's been taken advantage of. 

That all said: have fun in your games and don't get too stuck up on all this here if you're just here to have your little superheroes smash their faces in. 


PSA:

The Original Across the Bifrost hosts TT and Pat had probably the best MCP Podcast Episode ever on the subject: Luck in Wargaming. But sadly that is currently lost (please try to bring it back Norbert and Jaime! Update: Jaime confirmed that they tried to very hard but couldn't manage to do it)


That's it for today. 

It you'd like to support me you can do so at patron.com/sgprotocol 


Cheers from Germany 🍻 



Comments

Popular posts from this blog

Affiliation Overview: Web Warriors

  Why should you play the affiliation  Peter Parker is the single most popular character in the world. Miles and Gwen share between them the focus of the two best superhero movies of all time and are extremely popular characters in their own rights.  On the table Web Warriors are a supreme objective play team building off of their core of incredible 3 threats. They are big enough nowadays to offer different approaches too though.  How to start the affiliation  Untypically I don't think getting the affiliation pack is always the best way to start here. While Miles and Gwen are absolute core models to the affiliation, 4 threat Peter Parker will very rarely (if ever) see the table and Venom is decent but really up to personal taste. And seeing how the only real benefit of the affiliation pack is getting the additional Peter for free you can really decide if that's worth it for you. If you own core set 1 you already have the character and if you own core set 2 you h...

Affiliation Overview: Defenders (updated)

  Why should you play the affiliation  Defenders are a giant affiliation containing both the more mystic focused Defenders of Doctor Strange and also the Marvel Knights which made up the Netflix show Defenders.  On the board Defenders have a super deep roster with a tonne of great characters and some variety in play style.  How to start the affiliation Update: In Q2 2025 AMG releases the Defenders affiliation pack which uniquely comes with five models in Doctor Strange, Wong, Daredevil, Luke Cage and Iron Fist. You have to pick this pack up first and absolutely should.  After that I would first pick up Shang-Chi and Silver Sable followed by Hulk and then Namor and Black Panther, Chosen of Bast. You have your core very much there.The affiliation is very deep and has a lot of good characters but with those mentioned you have your probably most used models in hand already. Also all of the affiliated TTCs. Just note that Pentagrams of Falafel (let me have fun with t...

Affiliation Overview: Midnight Sons (updated rewrite)

Why should you play the affiliation? These are Marvels best Earth based characters. This is a definite objective fact and not just my personal feeling.  They also offer a wide variety of play styles and affiliated models.  How to start the affiliation Update:  Blade, Moon Knight, Doctor Voodoo and Ghost Rider will be combined into a MS affiliation pack, with their current stat cards in all four langues and new art. It releases in April.  Your first purchase here is probably still Blade (leader) and Moon Knight but you could also now start with Elsa Bloodstone (leader) and Man-Thing. Both are good value boxes as both come with a leader and a good affiliated 3 threat. So in the end it's probably best to just get both. Next up you'll want to look at getting a high threat model for which you have 3 options that all have their unique strengths and reasons to be picked. The biggest threat investment but also by far and large his best home comes in the Immortal Hulk. Ghost ...