MFZ/Rapid Attack: by the numbers

[Blorf]

Over the past few years, you guys have allowed me to collect some (anonymous!) data while using the Commander's Handbook online. This is pretty great, as it allows us all to get some insight into games. I felt it was time to give this back to the community and share!

There are a lot of caveats to the data, so take it more as entertainment than useful for any sincere analysis. For example, someone just checking out the game tracking section and not actually playing a game would have been recorded, and anyone using an opt-out to analytics or using the app while offline won't be recorded. Also, I'm rounding the numbers as I go, so things may not add up to 100%. But here it is anyway!

---------

These stats are based on completed games only.

Number of players:
63% of games were played with just two people. 26% have three, and just 11% are played with four. There were no recorded games with five! (I know my own 5-player games were tracked while I was offline.)

Game type:
The Skirmish is the most popular method to play, at 52% of games. Longer Battle games are as popular as demo games, with 24% each.

Number of rounds:
This falls into a typical bell distribution. Most games run 5 or 6 rounds.
1: 2%
2: 4%
3: 12%
4: 13%
5: 41%
6: 23%
7: 3%
8: 1%

Game length:
I wasn't recording this one, but I'm going to add it to the handbook re-release that I'm working on.

Winner's PPA:
Is it hardest to win as primary attacker? The data thinks so! There are quite a lot of wins as secondary, when you keep in mind that two thirds of the games are 2-player (nobody can have a PPA of 5).
3: 9%
4: 17%
5: 21%
6: 18%
7: 35%

Winning Score:
Winning scores are usually between 25 and 35. 35 is the most common value, 32 is the median, and 33.3 is the average.
00-10: 5
11-20: 8
21-30: 29
31-40: 24
41-50: 19
51-60: 5
61-70: 1

Tracking type:
77% of users opted for the most basic tracking type: just the assets. 14% used the maximum tracking available including spotting and activations. 9% used the mid-level system tracking.

---------

The rest of the data is not game-specific, but applies to all users. A basic cleanup was attempted to remove spambot traffic.

Country of origin:
Unites States: 61%
Finland: 8%
Australia: 5%
France: 3%
Russia: 3%
Canada: 2.5%
Germany: 2.5%
Brazil: 1.8%
Chile: 1.6%
Belgium: 1.4%
Singapore: 1.3%
United Kingdom: 1.2%
(all others <1% each)

Device:
Mobile: 49%
Desktop: 35%
Tablet: 16%

Referring websites:
The hangar: 75%
Mantisking's blog (the-mobile-frame-garage.blogspot.com): 7%
Other notable mentions (incoming links):
http://www.geeksleague.be/jeu-de-role/f ... me-en-lego
http://www.warmaniaforum.com/index.php
http://mfz-ra-rules.nanto.fr/en/home/more-resources

Popular rules reference pages:
Hit Chart
Frame Systems
A Frame's Turn
Quick deployment overview
Turn order
Example attack
Capturing Stations

Random stats!
723 companies were created.
2,022 frames were added to companies.
205 custom loadouts were created.

651 stations were captured.
594 frames were destroyed.
202 defensive ties were resolved(!)
305 games were started.
92 games were completed.

A simulated dice roll for a frame was made 4,571 times.
A simulated damage roll was made 750 times.

---------

I hope you enjoyed this (imperfect) look at the community and our games!

In case you don't know, I'm currently working on improving the app with games shared between devices, server-based load and save, and a player's network to help connect people and get games going. It's a long project (not my day job!) and its continued development is made possible by my supporters on Patreon. Thank you so much, you guys are fantastic! I've got only one last hurdle left before you'll get a hold of the initial beta version!

[Mantisking]

This makes me want to go through all of my battle reports and crunch the numbers.

[VitorFaria]

Interesting to know that the Handbook actually sees a lot of use!

[KungFujiApple]

Thank you very much for the numbers. Awesome work!

[Mantisking]

Blorf, does the dice roller part of the Handbook have any data Single Shot Rocket use?

[Blorf]

Probably not; it doesn't keep track of the systems set up when it gets used. Even if it did, there'd be no way of knowing when someone is rolling for real vs. just playing around. People who mash the buttons a lot can skew the data (and do already.)

What do you want to know? Maybe there's another way of finding out.

[Mantisking]

Probably not; it doesn't keep track of the systems set up when it gets used. Even if it did, there'd be no way of knowing when someone is rolling for real vs. just playing around. People who mash the buttons a lot can skew the data (and do already.)

Well, that sucks.

What do you want to know? Maybe there's another way of finding out.

We just had a moment in our last game when we were wondering how effective SSRs actually are.

[Blorf]

We just had a moment in our last game when we were wondering how effective SSRs actually are.

As far as mathematic effectiveness, the handbook itself can give you that. In the simulator section, you can drop an SSR onto a frame and see how its damage output changes. The numbers in the graph are the most likely number of systems of damage you will cause. The +1/+2 columns show what you can expect if you can commit one or both whites to the attack. Add and remove other systems here to see the differences.

What I gain from that exercise is that adding an SSR to an existing single-direct-fire-equipped frame only provides a minor benefit of about one-third system of damage. If you use all three at once, the benefit only climbs to 0.5 systems.

Adding an SSR to a frame that does not already have a direct fire attack makes a much bigger difference, going from zero to 1.5–1.8 systems. (2.2 with all three at once.) In this case your damage output increases dramatically, but ONLY if you would not have otherwise fired in that round. If you would have otherwise fired artillery, you basically gain nothing. If you would have otherwise attacked hand-to-hand, you're at a disadvantage by using the SSRs (yes, H2H even beats using all three at once).

So as usual, the question depends on the situation and how you use them. If you play with people who add them onto existing soldiers, they probably aren't doing very much for the outcome of the game. If you fire off all three with a dual-direct system, you'll add an expected 0.4 systems to your company's total expected damage output for the entire game. But if you fire them each from frames that would not have fired on that turn otherwise, your total game damage output might rise by 5 systems—more than enough to swing a game's outcome, for sure.

[VitorFaria]

...If you play with people who add them onto existing soldiers, they probably aren't doing very much for the outcome of the game.

What about people that mount SRRs into soldiers, as it allows you to drop all systems to damage and still fight at range?

A can be a force to be reckoned with during the late game!

[Blorf]

Yes, of course. The comparison is mostly between firing alongside a direct system vs. not otherwise shooting.

But you're still better off switching to hand-to-hand whenever possible.

[Mantisking]

So I went back and collected more data on the games my group has played.

We've played an average of 4.67 games per year since 2013. We have an average of 4.15 players per game, and games last an average of 3.96 turns. 57% of those games have been Battles and, the secondary Attacker wins 64% of the time. I have a little more data if anyone is interested.

[Blorf]

I have a little more data if anyone is interested.

In many ways, this is better data than mine. Yeah, what else have you got?

[Mantisking]

I have a little more data if anyone is interested.

In many ways, this is better data than mine. Yeah, what else have you got?

Okay, so it seems some of my math was off in the prior post. The average number of players per game is 4.11 and the average number of turns is 4.

As to players, 3% of the time we have six, 32% of the time we have five, 36% of the time we have four, and 29% of the time we have three.
29% of the time the number of turns per game is five, 42% of the time the number of turns per game is four, and 29 % of the time the number of turns per game is three. We've played a total of 112 turns.
57% of the time we choose to Battle, 36% of the time we Skirmish, and 7% of the time I don't have enough data to decide on a category.
The Defender wins 18% of the time, the Secondary Attacker 64%, the Primary Attacker 14%, and 4% of the time I don't have enough data to decide on a category.

I'm going to go back and dig out the PPA for each player/game and add that to my data.

[VitorFaria]

The Defender wins 18% of the time, the Secondary Attacker 64%, the Primary Attacker 14%, and 4% of the time I don't have enough data to decide on a category.

Woah, that's very weird to me.

I don't count my games formally, but I'm pretty sure my games have more balanced win stats for all the roles, I myself tend to win a lot as a defender.

[Blorf]

Woah, that's very weird to me.

Nah, that's the math of a reasonable balanced game. His games are typically 4 or 5 players. If each player had an equal chance of winning, you'd expect to see the number between 50 and 60%. (2 of 4 or 3 of 5 players in the game are secondary.)

That it's a little bit higher might be explained by stronger players aiming for that role in that group, or just natural fluctuation in probability.

Vitor, I think your games are more often just three people, yes?

[VitorFaria]

Vitor, I think your games are more often just three people, yes?

Yup, you're absolutely right!

I guess that explains it reasonably enough.

Stop having more friends than I do, you show offs!

[Mantisking]

Just crunched some more data. I'm only going to release my info as I haven't asked the other members of my group about theirs.
Out of the 28 games our group has played, I've been in 25 of them.
40% of the time I had 7 PPA.
12% of the time I had 5 PPA.
28% of the time I had 4 PPA.
20% of the time I had 3 PPA.

[CmdrRook]

Sounds like you strongly prefer the role of defender. Had that strategy yielded beneficial results?

[Mantisking]

Sounds like you strongly prefer the role of defender.

I wouldn't say "strongly prefer", it's more a case of everyone else showing up loaded for bear and not willing to take a chance to have the crosshairs on their back. Honestly, it's only the last two years where I've approached games with something like a plan as opposed to throwing together a bunch of cool-looking, odd-loadout Frames and calling it a squad.

Had that strategy yielded beneficial results?

Considering my winning percentage is 11% I'd say no.

[Mantisking]

Oh, here's a fun little fact. Outside of myself -- I organize and host most of the games so counting my playing streak would be unfair -- the person with the longest number of consecutive games is Occam's Spork at 11.