Loading presentation...
Prezi is an interactive zooming presentation

Present Remotely

Send the link below via email or IM

Copy

Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

DeleteCancel

Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

Stats game brainstorming

No description
by

Kevin Miklasz

on 13 January 2015

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Stats game brainstorming


http://www.brainpop.com/math/probability/basicprobability/
http://www.brainpop.com/math/dataanalysis/statistics/
Concept map
Teaching prompts
Statistics
Probability
Percentage
Population
Sample
Standard Deviations
by Gary Smith
How stats are used poorly to create data
Sampling
Biases
Observation
Causality
Mean
Median
Mode
Misleading
Stats

Analyze
Possible Outcomes of Rolling Die
More trials means better odds of rolling any number
Shark Attacks!
Rare
Fewer than 50 per year
Chances are 1 in 10 million
More unlikely than...
all these things
Shark attacks have risen...
Why?
Maybe there's just more people in the water
How do I not get bit?
Don't splash
swim in groups
don't wear flashy jewelry
http://www.brainpop.com/science/earthsystem/sharkattacks/
Outcomes
(results)

Probabilities
in real world
weather
rain
storms
lightning
hitting something
Lottery/
gambling
Biology
movement of water
not exactly random, but chaotic enough to appear random
Blackjack probability tables
Health
cancer development or recovery
genetic drift
mutations
Predation
Actually the question of how something finds food isn't totally probabilistic, but it's often hard enough to predict that we treat it that way
Langrangian vs. eulerian perspectives- eulerian approaches try to predict a systems response as a whole, and always involve probability
Pox (game)
Not taking a random sample of the population
Physics
Quarks
Position of electrons
Poker probability tables
Conditional probability
Distributions
Independent and Dependent Events
A BrainPOP movie
http://www.brainpop.com/math/probability/independentanddependentevents/
the probability of one event has no effect on the probability of another event
independent
ex. Dice
The chance of rolling a six on one roll has no effect on the chance of rolling a six on the next
dependent
the probability of an event changes the probability of a second event
in the brainpop movie, the probability of Tim choosing an apple affected the probability of Moby choosing an apple
ex. choosing without replacement
Compound Events
http://www.brainpop.com/math/probability/compoundevents/
the combined probability of two or more events
mutually exclusive events
the probability of both events having at the same time is zero
also, p(A or B) = p(A) + p(B)
inclusive events
it is possible for these events to happen in conjunction
also, p(A or B) = p(A) + p(B) - p(A and B)
Game Theory
http://www.brainpop.com/science/ecologyandbehavior/gametheory/
Mean, Median, Mode, Range
Range
A BrainPOP movie
http://www.brainpop.com/math/probability/meanmedianmodeandrange/
also, P(A or B) = P(A)
Deduce diseases from symptoms
Expectations/
Average of
outcome * probability
Network
depictions
Computer networks/
hacking
Stock markets, etc
Thermodynamics/
Properties of gases
(chaos?)
Clinical trials for medicines
twisting my ankle once makes it more likely that my ankle will be twisted in the future
Increased probability
Once I get sick with the flu, my body has built up resistance that makes it less likely to get the flu in the immediate future
Decreased probability
Autocorrelations
A statistic test that determines if an event is correlated with itself, in a sense measures the length of time in which future events are likely to follow the initial event. Also can find periodic patterns.
contains multiple choices
different areas have different infection rates
central limit theorem
Infections
Predict
Manipulate (distributions, initial states, objects, etc.)
Combinations
Permutations
likely game mechanics
http://www.spark101.org/video/using-expected-value-to-determine-life-insurance-p/
Dominant and recessive genes lead to ratio of traits being expressed in offspring
Distribution for a mixture of 2 different traits
"In asexual organisms, genes are inherited together, or linked, as they cannot mix with genes of other organisms during reproduction. In contrast, the offspring of sexual organisms contain random mixtures of their parents' chromosomes that are produced through independent assortment."
: Dependent and Independent Events
"Genetic drift is the change in allele frequency from one generation to the next that occurs because alleles are subject to sampling error."
: Stats concept
Use non-mendelian/blended inheritance, caused due to incomplete dominance or influence by multiple alleles, to represent continuous distributions, or change of values in future generations.

http://en.wikipedia.org/wiki/Genetics#Multiple_gene_interactions


reduced by large population sizes
isolated populations
bottleneck
founder effect
funny video about genetic drift, bottle necking and founder effect
Hardy-Weinberg Principle
Mendelian Genetics
1866, Mendel discovers heritable factors in generations of peas
true-breeding: when self-fertilization produces offspring that are all identical to the parent
hybrids: the offspring of two different varieties
cross: the cross-fertilization of two individuals

Epidemic Models
SIS/SIRS (deterministic models)
Random assignment of initial infections, some probability of contact with neighbours and spread.
People are susceptible, immune, infected.
Population changes are modeled, immunity after infection can be included, […]
Speciation
Pandemic (board game)
Pandemic (digital game)
http://www.theguardian.com/news/datablog/2013/mar/15/john-snow-cholera-map#
John Snow maps spread of cholera visually, convinces people that it's spreading from some point source in the middle of town.
showing infectious disease rates by size of country
http://en.wikipedia.org/wiki/Epidemic_model
population bottleneck
founders effect
http://www.genetics.org/content/148/4/1667.full
neutral mutation- changes in base pairs which do not affect protein coding regions, and therefore organism fitness
non-neutral mutation are about 1/300 per cell division (i.e. 1 in every 300 cell divisions is a change that effects fitness)
In genetics, a mutation is a change of the nucleotide sequence of the genome of an organism, virus, or extrachromosomal genetic element. Mutations result from unrepaired damage to DNA or to RNA genomes (typically caused by radiation or chemical mutagens), errors in the process of replication, or from the insertion or deletion of segments of DNA by mobile genetic elements.[1][2][3] Mutations may or may not produce discernible changes in the observable characteristics (phenotype) of an organism. Mutations play a part in both normal and abnormal biological processes including: evolution, cancer, and the development of the immune system.
https://www.khanacademy.org/math/probability/independent-dependent-probability/dependent_probability/v/introduction-to-dependent-probability
Two events are dependent if the outcome or occurrence of the first affects the outcome or occurrence of the second so that the probability is changed.

drawing from a set without replacement is ALWAYS a dependent probability event.
http://www.mathsisfun.com/data/probability-events-conditional.html
Drawing a hand of cards from a deck- once I draw my initial cards, I am unable to draw those cards again
Nicotine is addictive- smoking once increases the probability that I will smoke more in the future
Once my dog poops, he is unlikely to poop again in the near future because it is out of his system
Examples of Inclusive events
Let's say you have to pick a dog at random. The probability of finding a black-haired dog and a long-haired dog is a mutually inclusive event, long-haired black dogs exist. On the other hand, the probability of finding a chihuahua and a respectably-sized dog are not inclusive events- chihuahuas are never respectably sized
Founder Effect Videos
Possible Human
Evolution
Hybrid Man
Illustrative on a very simple level...
Differences among
New area
Natural disasters
http://en.wikipedia.org/wiki/Non-Mendelian_inheritance
Positive
Genetic
Infections?
mutations affect resistance which affect
infection which affect mutation, and on
spatially dependent events
Sentinels of the
Multiverse
http://sentinelsofthemultiverse.com/multiverse/heroes/legacy
How might we teach that infections are spatially dependent probabilities?
who's this guy?
Legacy
and his daughter
Legacy
how might we teach that genetic drift, founder effect, and bottleneck effect differ?
how might we teach that genetic drift is a sampling error? +1
How might we teach that trait inheritance is a set of compound events?
How might we teach that the outcome of one event affects the probability of later events?
how might we teach that the probability of an "and" compound event is the multiplication rather that addition of those events probabilities? +1
Like in the Monty Hall Problem, every move should be precluded by some additional bit of information. So the person has to think if this influences his move’s likelihood of success or not, and how.
a la Pox – layout and interaction. Disease is spreading, there are probabilities of infections, and probabilities of vaccine/inoculation working. Each move select some people to try inoculating.

Somehow, the probability of infections spreading changes in different places? Or greater infected area/number of people increases strength of infection and probable places of next infection; calls for player to change path of action in response.
Risk (/Pandemic) – map of the world, infection is spreading. Like Risk, greater infected areas gives the infection more reinforcements. Some interesting probability fights to conquer territories, that are not dice-based?
Can a dice based interaction be made to exhibit or participate in non-trivial probability calculations?
General thought: Making this into a board game, or non automated game, especially in terms of effect of actions, would be much better I think.
// In dice games like Risk, does one think about the expected win chances of 3 dice and their army vs. 2 dice and the defending army? I know I haven’t, and even if I did, I remember having lost terribly ‘unfairly’, so to say, at times. ‘lies, damned lies, and statistics’.
Then the players themselves can be expected to manually figure out the composition of the outputs, and results of their actions, and sooner or later engage in prediction of their actions’ success as well. Because in solving for compound events, the general rule of finding different permutations, and so on, I think a significant factor is in understanding the scenario correctly and multiplying the right sets of numbers.
Move character around to avoid infection which has different probabilities at different places. (and at different times?) While trying to do tasks that require going to risky (wrt infection) places frequently.
A task like obtaining a specified population configuration 2-3/more generations in the future. Distribute traits in today’s population in a specific manner for this task (given properties like dominance/recessiveness about trait). Some other element? What ways could this be dynamic?

One of the factors that could influence infection spreading rates could be the connectivity of the node in which the infection is (imagining a graph of vertices and edges). If it connects to 5 other nodes, it is more likely somebody'll get infected soon, rather than a node with lesser connections. Or vice versa
Totally agree! I feel like that is the challenge. I don't think it is enough to just say that we are in a game that involves dice, QED there is probability here. How can we reinvent the game a it so that the probability gets brought to the forefront, strategy-wise?
What if the disease is spreading around a map, and your job is to allocate where your people should hide until the disease ends. The disease starts in one or two squares, and it's probability of spreading to adjacent squares depends on the type of disease (airborne, via rats, etc) and the types of terrain it's spreading over. You need to basically predict the spread of the disease and which spaces will be safest in 28 turns (28 days later). The earlier you place your people, the more points you gain if they stay safe
I kind of like the idea of having control over a population rather than a person. Reduced the cost to failure, which I think needs to happen in a probability game
I like the idea of having infections depend on population density, creates a nice push-pull. Group a lot of people in one are that might be safe, but because there are a lot of people the disease will spread fast if it gets there.
Or have a population with traits already distributed, and choose which members to move to an island, like in the lego video
I think we'd need to get two different traits to co-occur together- maybe there's a bunch of dogs of different colors and different hair length. If you need a long and yellow-haired, you have to figure out which ways to breed the dogs to get that-

I think the breeding would have to be constrained somewhat- maybe instead of being able to select only two dogs to interbreed, you have to select 3 males and 3 females, which we randomly mate and create a new population. You'll have like 3 cycles to produce at least 5 yellow-long-haired dogs.
Would be interesting to have some kind of game where the inheritance could switch from being totally exclusive (one gene does not effect the other) to a dominant/recessive genepool (one gene somewhat effects the other) to a linked set of genes (one trait always includes some other trait). And maybe the player strategies would have to change as a result of this difference. This would span the range from exclusive to inclusive events.
Thought of something like ticket to ride, where you have to follow one goal to completion, but by spending resources building towards one goal, you limit your opportunities to pursue future later goals.
Could be like dominion, you have a deck of cards, and you are trying to get certain combinations of cards each round, as you draw 5 cards. Each round, you also have the option to sub out certain cards for different ones.
Thinking about an ever morphing monster (like the AIDS virus, but that's a bad theme for a game). As you throw one type of attack at a monster you are able to weaken it, but then it will build up resistance, making that action less successful in the future.

In general, having actions change the state of the game that effects the effectiveness of later actions.
http://en.wikipedia.org/wiki/Monty_Hall_problem
http://store.steampowered.com/app/292200/
Spore

In the first level you play
as a featureless amoeba-like
creature. You pick up parts
from other creatures that have
died around you. You can
begin to use these traits to
adapt your creature (i.e., claws
for fighting, proboscises for
omnivores) according to
how you prefer



Crazy Plant Shop
Ecodefenders
https://www.filamentgames.com/fws2/projects/eco-defenders
External
Internal
Who is our audience?
Spore didn't really allow for environmental influences that players could/would react to
Perhaps we could develop a
game where there are natural,
random occurring events and
the player would choose how
to adapt according to their
knowledge of events that
would occur in the future
based on probability
I'm not sure what the game is but there are some very interesting interacting probabilities here.
mutating immunity to vaccine
probability of offspring spread of immunity
Apparently, different kinds of viruses have varying rates of mutation (mutation per round of replication)
probability of infection
Starcraft
Micromanagement of populations. You can
divide your units to move
to different areas of a
map and engage in
different tasks
The tasks could be
to respond to certain environmental events with decisions based on which evolved species is best suited to deal with the issue
mutating advantages against immune cells (much less likely)
Rapid Forced
Evolution
There's a probability that pest develop resistance to pesticides but a much smaller chance that they'll evolve traits that make them immune to their natural predators.
Pokemon
You raise creatures
that battle for you
with different types
that are better suited
for fighting other creatures
http://nautil.us/issue/17/big-bangs/how-world-of-warcraft-might-help-head-off-the-next-pandemic
WoW and
Pandemic
Simulations

http://bulbapedia.bulbagarden.net/wiki/Effort_values

http://bulbapedia.bulbagarden.net/wiki/Individual_values
EVs are traits that Pokemon
develop from exposure to other
Pokemon. Can be intentionally
developed at trainer's discretion.
IVs are traits that Pokemon are found with or inherit from parents. These can
be intentionally bred for.
Mega Evolution
allows you to force
evolve your creature
in situations to gain
increased attributes
Mechanic that can allow
the player to forcibly
evolve their population
in response to or in preparation for situations
'Pre'/User-Research
What is the 'context' for use?
Does the teacher have a say in the choice of when/which game is played? (
Will
)

Throw some number of viruses against a disease, some die, some survive and develop resistance. The more you send, the more you lose, but the more probability of getting resistance.
This does not
seem to be only
external events (i.e. volcanoes
etc). Could be in relation
tp one genomic event
Mean gene slot machine
need a high number of individuals, or a high number of events on a single individual
Multiplayer, one player controls viruses and tries to spread, one player tries controls people and tries to keep them way from viruses
If you are trying to
gain one trait
a dice or a probability
element will determine
what commonly
associated traits you
might acquire
20 traits, every time the game starts the traits are linked in a uniquely different way
Have players map the traits visually
Goal-
Rules-
Pieces-
Theme-
Strategies-
CC elements-

Extremes of not doing stats right in a games
Take a game like minecraft
The oppositional stats- in this case the stats really mostly function to oppose your desired goal. To produce grinding, in random. For instance in Minecraft, 1 in 10 blocks of gravel produces a flint. If you want flint, you have to mine a whole bunch of gravel. The implementation of stats just leads to grinding, not to strategic thinking about stats
The linear relationship- One way to get rid of the oppositional stats is to just increase probability to 100%- every time you mine stone, you get cobblestone. You are now better able to achieve your desires in the state of the game- but the stats have disappeared, in favor of deterministic relationships.
I think we want a game that avoids either of these states, i.e. one that involves stats and requires strategic thinking about those stats.
Thus, who all are the users and stakeholders whose opinions/needs influence the design of the game and the choices we make?
What prove-able need or problem is being solved?
What is needed – by teachers; for students wrt probability? (possible information from brainpop as to what videos are commonly used, or what things teachers have asked for or remarked about?)

What 'features' are expected in the target product? How do we plan to incorporate them?

Players play cards or moves that damage a common enemy (success of attack associated with probabilities). This changes the state of the common enemy. And make the consequent players recalibrate planned attacks. For instance, if player 1's attack removes the dragon's arm, player 2's attack on the arms has a lesser chance of succeeding. Something more complex so that checking depended of events requires thought.
To reconsider: (what's mentioned in the right on things that could be done wrong)
Probabilities in moves for individual play is still a matter of luck; somehow the person should be playing [for] the probabilities themselves (rather than, as earlier said, "against the probabilities", hoping they work out)
The Bean Machine – configure the bars above in different manners to get different distributions beneath. analogous to the, distribute trait amongst population for a final desired configuration.
More clearly reflective of statistics, it feels?

Also, appropriate representative of what playing *for* probabilities should be like. Where the desired output requires a certain probability of occurrence, which needs to be made to happen.
Yes, I think the phrases not wanting to play "against the probabilities", but play the probabilities themselves is desirable, was discussed.
I imagine one way of doing this would be: that actions should not function like they normally do – i.e. players want to conduct attacks as is traditional, and they're hoping against probabilities it works. Instead, an attack's (or move's) requirement itself should be a certain probabilistic scenario (say only 60% of the artillery should fall on the enemy for it to work correctly). And the player has to manipulate things to make that happen.
Literature research about teacher's or students' needs wrt understanding probability – what are common difficulties, etc. (
Vishesh
)

Will this be for class room use? General distribution?
I agree. You should use stats
to make informed decisions
about the best actions to take out
of a set of actions i.e., which adaptations
given a table of info or something etc.
http://www.teacherlink.org/content/math/interactive/probability/numbersense/misconceptions/home.html
http://www.corestandards.org/Math/Content/6/SP/A/2/
Understand that a set of data collected to answer a statistical question has a distribution which can be described by its center, spread, and overall shape.
Develop a probability model and use it to find probabilities of events. Compare probabilities from a model to observed frequencies; if the agreement is not good, explain possible sources of the discrepancy.
http://www.corestandards.org/Math/Content/7/SP/C/7/
Design and use a simulation to generate frequencies for compound events. For example, use random digits as a simulation tool to approximate the answer to the question: If 40% of donors have type A blood, what is the probability that it will take at least 4 donors to find one with type A blood?
http://www.corestandards.org/Math/Content/7/SP/C/8/c/
Understand that statistics can be used to gain information about a population by examining a sample of the population; generalizations about a population from a sample are valid only if the sample is representative of that population. Understand that random sampling tends to produce representative samples and support valid inferences.
http://www.corestandards.org/Math/Content/7/SP/A/1/
Use data from a random sample to draw inferences about a population with an unknown characteristic of interest. Generate multiple samples (or simulated samples) of the same size to gauge the variation in estimates or predictions. For example, estimate the mean word length in a book by randomly sampling words from the book; predict the winner of a school election based on randomly sampled survey data. Gauge how far off the estimate or prediction might be.
http://www.corestandards.org/Math/Content/7/SP/A/2/
Understand that two events A and B are independent if the probability of A and B occurring together is the product of their probabilities, and use this characterization to determine if they are independent.
http://www.corestandards.org/Math/Content/HSS/CP/A/2/
Recognize and explain the concepts of conditional probability and independence in everyday language and everyday situations. For example, compare the chance of having lung cancer if you are a smoker with the chance of being a smoker if you have lung cancer.
http://www.corestandards.org/Math/Content/HSS/CP/A/5/
Evaluate and compare strategies on the basis of expected values. For example, compare a high-deductible versus a low-deductible automobile insurance policy using various, but reasonable, chances of having a minor or a major accident.
http://www.corestandards.org/Math/Content/HSS/MD/B/5/b/
Find the expected payoff for a game of chance. For example, find the expected winnings from a state lottery ticket or a game at a fast-food restaurant.
http://www.corestandards.org/Math/Content/HSS/MD/B/5/a/
(+) Analyze decisions and strategies using probability concepts (e.g., product testing, medical testing, pulling a hockey goalie at the end of a game).
http://www.corestandards.org/Math/Content/HSS/MD/B/7/
It should be fun to play
The concept should be embedded implicitly in the gameplay (i.e. the game should be designed such that effective play strategies require understanding and manipulation of statistics concepts)


The E-coli breeder game
Goal- You are a breeder trying to generate a brand of bacteria with a particular pesticide resistance ratio, to protect people from getting sick.

Rules- On each turn of the game, you are told how common 3 species of diseases are each game. You have a limited number of turns to breed an E-coli with the exact resistance ratio that matches the disease prevalence in the population. You spread the E-coli to 3 petri dishes. Each round, you can insert a certain amount of a certain kind of disease. That disease will both kill a certain percentage of the E-coli, and grant the surviving E-coli some resistance to that particular disease (according to some probability distribution), and reduce an E-coli's resistance to all diseases not encountered that round. All E-coli then triple their numbers, and the process repeats itself. After a certain number of turns, you need to pick one E-coli that matches the desired disease ratio.

In later more difficult levels, we can add conditional probability, such that being infected with disease A will make more more likely to die to disease B, or more likely to develop resistance to disease C.

Pieces- 3 petri dishes. A population of E-coli. 3 kinds of infectious diseases.

Theme- It's the future, and E-coli are all the craze. You are a geneticist, breeding Ecoli in your lab all day. E-coli can fight the most common diseases on the planet, but only if they are bred right. You are given the desired profile of E-coli to create, and limited time to do it. The health of the world depends on you!
Makes me think of forest fires potentially, fires are more likely to spread down different paths than others, spatially dependant probability events
I'm starting to think that spatially dependent probabilities are too subtle- yes they are dependent event, and their probability depends on their spatial configuration, but I think the dependancy between the events is too subtle, too nature to players. Usually the thing next to something gets effected by it. Teaching dependant probabilities in a non-spatial way might be better, education-wise
Host vs Parasite
Perhaps we can do interplanetary worlds to provide for a host of different environments to adapt to
Events:
E. coli goes rogue; need to counter
Resistances developed against certain strains
Expected negative abilities when trying reach for higher tier abilities
yes
via BrainPOP's GameUp
if we hit common core, we are reaching the needs of teachers
3 use cases on GameUP, work for all 3
-outside school on own
-inschool, part of lesson
-inschool, free time


outside school
In-school – teacher assigned and like-homework
Extra time in school

Alignment with content (wrt curriculum), and other site content
www.brainpop.com/games
way to visually represent how traits in the next population are related to the current one
like a final fantasy tactics job tree. Or games where using a weapon more an more makes ou an expert in that weapon, opening up later job trees
job tree adaption design circle
(note for Kevin to flesh out later)
Make the game fun!
http://www.brainpop.com/educators/community/bp-game/the-sports-network-2/
Be able to manipulate
your probabilities
statistical
specialization
Cost-Benefit
Analysis
What are real life
"microbial-jobs"?
High risk-high reward
Rogue-like breeder game
i.e. Rogue Legacy, FTL, Don't Starve
i.e. Rogue Legacy, FTL, Don't Starve
Playtesting
Had three petria dishes with populations of bacteria, each had to survive being hit with a disease. when hit with a disease, each individual would gain one of three possibilities: survive and reproduce (double in number), gain resistance and not reproduce, or die. Probability of each of each occurring was determined by the disease, each disease was given difference ratios as shown below.

Had a some interesting play dynamics, and visually it felt very cool to have the resistances of each individual displayed visually. Treating resistance as a bit of a binary state, individual by individual, didn't feel like the greatest dynamic, not did it embed stats the most-effective way. Since new resistances just overwrote old resistances, the conditional probability didn't feel totally with it.

Also, I talked it over with my girlfriend a bit post-testing, and it did not in fact treat genetics and evolution in the correct way. Exposure to a disease doesn't create genetic resistance, it simply makes those in the population that have resistance more prevalent in the population. Generally, variation needs to exist, for natural selection to have something to act on. Which meas we need to be starting with populations that contain some resistance to each disease.
mid-game
end-game
Version 2, we thought of after playtesting. I think more biologically accurate, and will treat the statistics in a more interesting way. Not each petridish doesn't contain individuals, but one giant population. That population has 4 statistics: number of individuals, and % of the population that is resistance to disease A, B, and C. By exposing the population to one of the three diseases, you generally lower the total number of individuals, but increase the percentage that are resistant to a particular disease. In general though, individuals with resistance grow slower than those without resistance (having the genes to grant resistance creates a metabolic cost), such that the % of the population that is resistant will tend to go down with time. The goal of the game is the same, to hit certain benchmark numbers of % resistance by exposing petri-dishes to certain diseases before the time limit runs out.
Evolve(r)
Main interaction – select groups of people (/beings) and make them procreate with each other, for certain problems coming up a few generations later

You’re controlling a village of people who are grouped by traits. Reorganization, or selection of a certain trait according to which people aren’t grouped, requires specifying desired traits, and an option to select people with that trait.

The aim is to move people around into these sub-villages where they'll procreate and exchange spread across their traits according to the numbers they're distributed in.

The target traits they have to evolve could/should change with respect to how they evolved. In that if some mutation is required to be spread across 1/3 of the population (which solves some problem) - the new problems are in response to the new mutations the population has. (similar to the different worlds and people one obtains in Spore?)

<Just moving people around might be overly simple? Is there a better interaction or more complex system to work on?>

---I think allowing for an initial movement, and then allowing for other movements at elect times, or a limited number of times, would be ideal- moving people around should feel like a limited resource and thus require strategic thought
What if there are a few things to respond to- disease, predators, and heat. Each population is in a different environment, and so has a different survival requirement based on those three traits.


What would time scales be like
for this game? Millions of years
to accommodate for the changes
that would take place in a population?
Perhaps they could be a space faring race
to include more exotic environments; perhaps a look into future possibilities of human evolution if they moved around the
universe. grandiose, but a thought.
Time moves in years/decades. After rearranging people, you can start(/resume) time, and see the population composition in terms of traits change live. You'll have quests of composition, and challenges to face. Possibly in parallel. // The challenges could have rewards of different mutations you can distribute across your population.
And you can rearrange people while time is moving as well. And certainly change tracks (substantial rearrangement of people/beings) when a new challenge or quest comes up.

More information:
Apart from the composition of the people, there could also be some measurement of the population's traits' feature ratings: values on immunity, agility, diet/metabolism etc. (this appears like it could become overcomplex) but some of the components could also influence the rate of growth of population, being able to face challenges like predators or diseases, and different characteristics/abilities.
(Tentative) Rules
Although the game is about statistics that involves strategy,
can we include chance/random probobility to change game
dynamic for a game?
Is this multiplayer?
Is the game beatable after few plays? (is there a probability pattern that can be used eveytime to beat the game?) or does the difficulty adapt or/and change?
I see this game featuring the
job tree mechanic the most out
of any of the three

It seems like a tower defense kind
of game where you upgrade your
units every round
How does this/can this incorporate
statistics though? What are cellular level
statistics? Population proportion die-offs
given before each action is taken so you
can do a cost-benefit analysis
Will it be possible to play multiplayer?
job tree for the parasites, the host, or both?
Is there a way to include conditional events in this? Connect this with the teaching prompt on the left, that new information intermediately could lead to different chances of the final results?
I like the idea of a tower defense feel a little more, the current game feels a bit like you'd be investing a lot of time in micromanaging the spatial position of your units, and not leave enough time for thinking about stats strategy. If attack paths and defense points are predefined, that can free things up a bit.
i would say both. if multiplayer, then both players
would have trees, and knowledge of other's trees.
if single player, then player can foresee and adapt. the game can indicate percent chance the opponent will choose next 3 or so traits and player can plan accordingly
yeah i'm imagining more
static turn-based type
of tower defense if anything
Building off of netrunner, what if the host can build a bunch of defenses, but those are not initially known to the parasite until they attack the host and have to encounter the defense. Then you have to weigh statistic unknowns in planning your attack.
Suddenly, a certain village is affected by some weird storm that possibly affects their traits, or how it is passed on – is it preferable to change configuration or not?
turn-based or real-time?
build out tech tree, define core mechanics, and define potential actions for both sides
going back on job tree at a cost but to adapt; set action points
Christian
http://www.ologames.com/Free_Games/Who-Wants-To-Live-A-Million-Years
Benchmarking
Listing other probability games;
their features and shortcomings;
and comparing our plans or imagination with respect to them


This whole idea of playing with ecosystems of microbes is fascinating.
Host v Parasite Playtest
The parasite started on the right side of the board with three 20 sided die. Each die represented a swarm of 20 viruses.

Orange rectangles were immune cells. They ate 5 viruses per encounter. It was not possible to develop immunity to them.
The host started on the left with three cells and 3 Mutation Points. Each turn, he got 1 MP per cell.
The
objective of the game
for the
host
was to keep at least one cell alive for 30 turns. The objective for the
parasite
was to destroy all three cells.
Green/blue squares were antibiotics. They had kill percentages of 50 and 40 percent, respectively. Anything not killed in an encounter became immune to that color. The square moved under the die to denote its immunity.
Each side took turns moving their pieces. Each piece could move 2 spaces.
Parasites reproduce by getting 5 viruses
to a cell. The cell becomes infected for two
turns. At this point, it is destroyed and
produces two new swarms, which inherit
immunities of their forefathers.
In three turns, a new cell emerges.
When viruses and antiobiotics collide, we rolled a die for each member of the swarm. All that didn't die, became immune.
Results
Compelling gameplay. I think it's the start of a very interesting system. I knew next steps would be increasing the number of choices that were based on careful consideration of not just single probability but nested probability.
Yellow squares were cells.
Host uses MP to buy immune cells (3 MP), blue antiobiotics (1 MP), and green (2 MP).
Host vs Parasite v2
Mostly redesigned the upgrade and resistance-developing system to be more in players' control.
Playtest 2
Based on version 2 of the game, as described above- it felt more tedious to keep track of percentages, and a less interesting game overall. The tediousness would be totally alleviated if a machine were to do the calculations for you. With only two things to track percentages on (death rates with and without a resistant gene), the play mechanics felt a bit static. To get to the desired percentages, you have to know what are the resistant percentages in your population. And then you simply add more of the type of disease which is furthest from the required percentages, to kill a bit more of everything else while leaving that gene intact. Very uninteresting strategy.

A more interesting game might be one where you are instead trying to eradicate a population of bacteria using antibiotics. Maybe you are a doctor and you have to prescribe a sequence of antibiotics that can cure the disease, given a patients prognosis (which includes a description of the resistance percentages). After each treatment, the bacteria randomly mate and double the population in size.

To add complexity, maybe the population can be given a 3-4 gene signature, which the genes not each corresponding directly to an antibiotic. There can also be 3-4 antibiotics, which each have a differing effect based on the genes present. For example, antibiotic 1 might be really effective against bacteria with genes A and B, but really terrible against C and D. Antibiotic 2 might be extremely effective against C, and mildly ineffective against A, B and D. Each round of the game would present you with different available antibiotics.
This is the setup I chose from understanding the rule description.
3 petri dishes; 30 E.Coli are given to begin with; Diseases A, B, and C such that A - kills 50% of who it attacks and gives resistance to the rest; B kills 25% and C kills 70%; and both give resistance to the rest.
And then I chose that all E.Coli have 2 units of resistance against A, B, and C each to begin with. And gaining resistance on fighting any one disease, implies those E.Coli get a +1 resistance unit against who they fought, and -0.5 against the others. And the target resistance ratio was 3:2:1.
This was interesting to try, but I felt harrowed after one-two moves due to excessive data keeping and not being able to keep track of all the things that I probably should be aware of.
After trying one-two moves itself, I was pretty confused and found it hard to keep track of all the information. One of the primary points of confusion was, what did the target resistance ratio mean? Do we need any strands of E. Coli, even one or two individuals, to have that resistance ratio, or is the 'average' resistance ration of all the surviving E. Coli supposed to be 3:2:1 (or chosen target).
Also, as problem setters, is any combination of the diseases' kill ratio and desired target ratio achievable, or not?
Here, I simply put all the E. Coli into one petri dish, and 'inundated' them with the different diseases one by one, just to see what sort of resultant resistance ratios emerge.
Multiple villages and traits seemed again, slightly overwhelming to compute and keep track of manually. So I simply tried to see what can be achieved in one village, with two [pairs of] traits: in 64/60 people, I assigned everybody to be 'purebreds' as in both alleles of their genes are the same, and we're considering two genes – concerning eye color, and height; and distributed the 4 kinds of people in equal numbers.
I tried to figure out ways to reduce the population of the tall people (assuming the tall allele is dominant), and it's interesting to see that it takes three generations for this to be manageable, assuming previous generations go away. This was so far the only observation of an achievable task I could make.

Playtest
100 B (Black eyed) - T (Tall)
& 100 B-s (short)
100 B (Black eyed) - T (Tall)
& 100 G (Green eyes)-T
So I got a friend to move people around, and tried to work on the calculations, and sadly couldn't move beyond three moves because the calculations became overwhelmingly tedious.
In that limited an experience, my friend said that she couldn't form much of a clear picture about what is happening, or the possible impact of a move. Also, she tried to attempt some of the calculations herself, and those seemed pretty daunting too, and she said that as a player, she wouldn't want to deal with that kind of math.
So this wasn't a fun experience in practice, this time. I think mostly because of how I took 15-20 minutes of computation between moves to report the next status.
I feel like it might be possible to generate a vague intuition as the player practices on the game, but I'm not sure what else can be done to make this more fun.
Predicting traits
People in different 'localities'/villages/regions are moving around, and you have to predict after some movements, the likely composition of a trait.
I'm doubtful about the fun-ness in this game, but essentially reversing the above game, seemed like a possibly interesting concept – in that it makes the calculation and thinking about probabilities more explicit.
Making the Punnett square for different sets of traits/gene pairs in the population, and finding out the resultant genes and their compositions, although largely tedious, was moderately challenging, and felt like something that *might* be an engaging task in itself, in some form.
In retrospect right now, and looking at the other playtests, this game might be more playtestable, as well as playable/approachable if I didn't work with large populations in hundreds, but very small ones, say less than 10.
The demerit here, of course, is that the random event is probably a random number or a die throw, and not being sampled a lot, leading it becoming less predictable in terms of scale, I think (as we discussed with respect to the other dice games).
But I'll try to do this tomorrow to see what I can get.
This, for me, raised the repeated earlier question, of the extent of explicit or formal calculation we want the player to engage in, if ever. If yes, how does this work in comparison to the related trade-off with fun (can this computation ever be fun, somehow?).
Or is a primary focus more for this to be a fun game, such that the underlying concepts of probability can be formally described and taught later, and this only works as a representation to help prime a player?
Couldn't think of a good way to do the calculations easily, so didn't actually playtest the game, but I did draw out a potential gameboard. The idea was that the game would involve a series of levels, and each level would have a different arrangement of islands, and number of traits. Player moves involve opening and closing migration routes between islands.
^What the player was told
^Background information and calculations I (/the computer) worked on
nice, math!
Next Generation
Population calculator
https://docs.google.com/spreadsheets/d/1qWsGB5rhD8GuRpFdlzRMS_5qhfkFk49aFFUeVjt1MoY/edit?usp=sharing
is ^here.
Enter the populations of specific allele pairs in the top row, and the resultant population (though often in decimals), shall be given in the bottom row
Calculations from the computer
Level 1, a simple island configuration
Mid level difficulty island configuration
Playtesting v2
New eradication game- end of round view from playtesting
This game felt very puzzle-like. There didn't feel like there was enough complexity in the game to feel interesting. There was also the suggestion of tieing this more directly into antibiotic treatments- so instead of multiple diseases, only one disease, and needed multiple treatments, enough to cure the disease, but not so many that you exhaust your limited resource pile
http://www.carolina.com/teacher-resources/Interactive/online-game-cell-structure-cellcraft-biology/tr11062.tr
Cellcraft
hero academy
Wave attacks a la Tower Defence – After some generations, village 2 is going to be attacked by a composition of 40% archers and 80% militia.

The aim is to maximise survival, knowing that you can procreate mages and knights which are good against militia and archers, respectively.
Calculator V2
https://docs.google.com/spreadsheets/d/1XvdqL6lrRdKHoxRjZpSD6EcKJlXLHVlB4SSRT2RA-So/edit?usp=sharing
or https://drive.google.com/file/d/0ByEKm4TLSaMuajlUTzZlak1rSUk/view?usp=sharing
When you start the game, you need to enter values in row 6. After that, you simply need to enter changes (how many people of this trait set moved away from or into this population). The resultant values will be printed in row 22.
Copy row 22 into row 6, and the values to be reported to the player, will be updated in row 2.
Where is the stats in the strategy? Needs to be clearly elaborated
Full transcript