What is the likelihood that ____ ? ( stats and luck thread )
TLCstormz
Age Unconfirmed Posts: 1,668
* And if the title is going to get this thread moved / trolled, please just change it, before you move / troll me. *
Soooooo......if the mathematicians can help us out here, I have some random questions / annoying observances.
Mine are :
1. What is the likelihood of Daken's blue causing a game ending green to appear IN THE EXACT SPOT that will cause critical after critical after critical, on the CPU's second freaking turn?
2. Same question as above, except substitute Daken's blue for Venom's snare?
Some community ones that I see over and over again are :
3. What is the probability of Wanda's blue appearing in a corner?
4. MNMagneto's blue and red in corners?
Etc.
Anyone else that has these head scratching conundrums happening to you on a "regular basis" ( I put it into quotation marks, cuz Lord knows the MPQ Calvary will feel the need to come in and say "you are suffering from consummation bias, and such ), pleeeeease feel free to ask your own quotations, as well.
:+D
Soooooo......if the mathematicians can help us out here, I have some random questions / annoying observances.
Mine are :
1. What is the likelihood of Daken's blue causing a game ending green to appear IN THE EXACT SPOT that will cause critical after critical after critical, on the CPU's second freaking turn?
2. Same question as above, except substitute Daken's blue for Venom's snare?
Some community ones that I see over and over again are :
3. What is the probability of Wanda's blue appearing in a corner?
4. MNMagneto's blue and red in corners?
Etc.
Anyone else that has these head scratching conundrums happening to you on a "regular basis" ( I put it into quotation marks, cuz Lord knows the MPQ Calvary will feel the need to come in and say "you are suffering from consummation bias, and such ), pleeeeease feel free to ask your own quotations, as well.
:+D
0
Comments
Needs more information - how strong the enemies are, how many crits, etc.
Assuming it can replace any of the 4 corner tiles, 4/80 or 1/20 or 5%.
Magneto creates 5 red tiles. Assuming all 4 are open for conversion and an average number are already red (we'll say 11) and that you're not converting any of the red ones to blue, there's a 62/66 * 61/65 * 60/64 * 59/63 * 58/62 to not generate a red tile in a corner. That's a 72.4% chance of not putting one in the corner, or to say a 27.6% chance of putting one in the corner.
Could be a myriad of things, one being bias, one being broken game. That said, it's a 1/4 chance for Magneto, so that's not particularly rare.
Actually the title is perfect, the content of the thread is why it has been moved to the Statistics forum
- the ratio of the enemy team getting on off-board cascade (that is tiles from off-board magically fall into a match automatically) compared to you getting one is 2.4:1.3. Meaning they're roughly twice as likely to get an off-board match than you fall out of the sky.
I started keeping other statistics only 18 games ago so fairly early to draw some conclusions but here's the data.
The ratio of a special tile you create vs. one the AI creates being matched in 1 move is 3.9:1.3 meaning the AI's special tiles are 4X less likely to be created in a matchable position and more likely to stay on the board for more than 1 turn.
The ratio of the AI vs. you getting a match of 4+ or greater fall out of the sky is 1.4:1.1
Once I have more data I can dial in on the numbers but I'm fairly certain the placement of special tiles generated from powers and the tiles that fall onto the screen are not random. The AI itself is fairly stupid and rather random in a lot of choices. So my guess is to make up for the random AI the tile placement of special tiles and tile order that falls onto the screen is always skewed in the AI's favour. This translates into the AI (on average) generating more tile matches and more power triggers, while at the same time limiting your special tile placements.
The only way to increase the likelihood of your special tiles not being matched in 1 move is multiple firing of the power on the same turn.
It wouldn't be completely impossible to answer these, assuming that we just limited it to three crits in a row, which is usually enough to totally throw a match in your opponent's favor, just with the amount of AP they net that way. I'd probably attack it using a Monte Carlo simulation of the problem. That's not something that I have time to do right now, at all. It'd be interesting to futz with the code in my head though