27 replies to this topic

[Obadiah333]

Couldn't finish the challenge. Tried tonight to get the last map I needed after the dumb down (200/300)- Canyon Network. I played from just after 5pm pacific until 9:30 pm. I got that map 0 times. Probably about 30 drops.

Over the weekend, I played numerous drops. Most of them ended up on the Bog, the map I hate most. But on to the point. Your RNG map rotation is not even.

I challenge you to pull my map stats for the weekend and show how many times I dropped on each map. I guarantee you that the numbers will show that it is outside of standard deviation. For those who don't know about probabilities, I'll break it down in case you don't know what that means.

Given the number of maps that were available, and an equal random chance to drop on any of them, a common amount of drops should average out between the maps. Then, there are stages outside of that common average, which are the standard deviations from the common average. These numbers (of map drops) must be within a certain percentage of that common average. Any number outside of that percentage means that the random number generator is not functioning properly.

Sorry if that is confusing, I tried to explain it the best I could the simplest way I could. The long story short - the random map generator is broken - the drop rate on each map is outside the standard deviation allowed.

PGI - you win. MVP dropship is now MVP RNG map.

I wonder how long this post will last before it gets deleted. I saw multiple threads deleted today that were negative concerning the event this weekend. Cover up incoming.

[Tristan Winter]

I wouldn't be surprised if there was an issue. I've been playing a lot this weekend. Can't imagine how many matches, but I'd guess something like 40-50. I only just got River City for the first time in my last match, which - thank the old gods and the new - happened to be a win and enough to qualify for Tier 2.

[Davegt27]

I rage quit last week before the event and looked up random number generator

Aka electronic dice roll in this context

I was going to send it to PGI but then decided it would not do any good

For those that don’t know a computer can duplicate 2 six sided dice with a random number generator program

[Stelar 7]

I have been playing for 3 days, and finally got the drop on HPG I needed, and enough hurt on the Reds to win. I can't complain though as I had two or three previous drops there, just didn't have good games on those drops.

Anyway the thing about random is you expect outliers. If they try something like this again they need to code a higher % of map choice for maps that are needed. Better yet though would be non-map specific contests. Then we only have to deal with the random teammates, random foes and the %chance my UAC will jam.

[Hit the Deck]

Obadiah333, on 21 September 2015 - 08:47 PM, said:

...
Given the number of maps that were available, and an equal random chance to drop on any of them, a common amount of drops should average out between the maps.
...

Each map's probability to appear is not equal.

[Tristan Winter]

Stelar 7, on 21 September 2015 - 09:01 PM, said:

I have been playing for 3 days, and finally got the drop on HPG I needed, and enough hurt on the Reds to win. I can't complain though as I had two or three previous drops there, just didn't have good games on those drops.

Anyway the thing about random is you expect outliers. If they try something like this again they need to code a higher % of map choice for maps that are needed. Better yet though would be non-map specific contests. Then we only have to deal with the random teammates, random foes and the %chance my UAC will jam.

Yeah, as long as you need random maps to complete, you're going to get disconnects. And they will increase over time.

[DaZur]

You have a 1 in 9 chance to draw Canyon presuming the weighting of all maps is the same. Since you only needed Canyon to complete you only had one favorable outcome (Favorable is actually a misnomer)

The easiest way to look at this is the more "favorable" outcomes you have the higher your possibility to draw one of them... With only one you are resigned to the least favorable outcome of 1/9. (If you had three maps needed you'd have 1/3 the chance to draw one of them as an example).

The unknown is how the depth of your PSR pool (number of available players) was/is in contrast to your favorable outcome... The more players playing actually further reduces your probability to draw that singular map.

Edited by DaZur, 21 September 2015 - 09:08 PM.

[Stelar 7]

Tristan Winter, on 21 September 2015 - 09:06 PM, said:

Yeah, as long as you need random maps to complete, you're going to get disconnects. And they will increase over time.

I don't disagree, though I don't condone the disconnects, I played every map I dropped on, and it probably cost some folks their games because I generally did pretty well.

This contest pitted us against each other a bit more than I like. For me to get my 400's. I had to wreck 3-4 mechs on every drop. Those folks then couldn't get their points, and those points were lost to my team.

[Hit the Deck]

DaZur, on 21 September 2015 - 09:07 PM, said:

You have a 1 in 9 chance to draw Canyon presuming the weighting of all maps is the same....

They are not the same as far as I know (I once read the probability list somewhere).

You can check your own map stats to confirm it.

[Stelar 7]

Checking your map stats will be a false result. You need a higher sample set for the numbers to even out. On low sample sets you expect outliers, that is what random means. If you see an even distribution without outliers, it wasn't random.

[Hit the Deck]

Stelar 7, on 21 September 2015 - 09:13 PM, said:

Checking your map stats will be a false result. You need a higher sample set for the numbers to even out. On low sample sets you expect outliers, that is what random means. If you see an even distribution without outliers, it wasn't random.

I expect forum warriors to be frequent players hence the suggestion (to check that each map have different prob. to appear - not for this event).

On a related note, PGI could very well alter the weighting just for this event to make some maps even rarer.

Edited by Hit the Deck, 21 September 2015 - 09:20 PM.

[EgoSlayer]

Hit the Deck, on 21 September 2015 - 09:11 PM, said:

They are not the same as far as I know (I once read the probability list somewhere).

You can check your own map stats to confirm it.

They are the same weighting to appear with two exceptions:
1) New maps get boosted rates so everyone gets to play them a lot and metrics are recorded for issues.
2 ) The maps with twins (old River city + River City Night, Old Forest colony + Forest Colony Snow, Frozen City + Frozen City Night) each have half the normal rate, so combined they equal 1 regular map

[Amsro]

yeah still have yet to see a match on mining collective. Literally 4 days, ~100 some odd matches. Not Once.

[Vlad Ward]

Hit the Deck, on 21 September 2015 - 09:11 PM, said:

They are not the same as far as I know (I once read the probability list somewhere).

You can check your own map stats to confirm it.

PGI confirmed that they changed the weighting so that they're all equal for this event.

If you're talking about the xml file on your computer, that's a reference sheet that may or may not be out of date at any given time.

If you're talking about your profile stats, then yes - there was a time when different maps had different weights. They like to do this to let people see new maps more often. That is not the case during the event.

[Cadaver603]

I haven't seen Crimson Strait once.

[slide]

I am sure that the RNG gods would have favored you (the OP) more fairly had you been the only player in this game that the match maker has to deal with.

As it stands their were 23 other players who needed to get a match, multiply that by the thousands of other players that need matches and I am sure that the incidence of maps occurring would be more or less equal.

Does it suck that you personally get the same map over and over, yes it does, but it does not mean that a map isn't on even rotation. I guarantee there is probably a whine thread for every map with the same theme. What you experience is not the same for everybody.

[Amsro]

RNG is the problem, how can you make RNG a requirement for completing a challenge.

Mind numbing.

[Hit the Deck]

Vlad Ward, on 21 September 2015 - 09:40 PM, said:

PGI confirmed that they changed the weighting so that they're all equal for this event.
...

Ah, thanks for the info! I was not aware of it.

Edited by Hit the Deck, 22 September 2015 - 01:16 AM.

[Eboli]

Congrats on everyone who have reached the target.

I chose not to play at all as I saw nothing but trouble for this event. Now, after seeing the anguish it has caused I am seriously considering asking for my IIC pack refund based upon PGI's inability to get things moderately done right. I will hold out to see how the Black Knight turns out first.

I am not a white knight for PGI but I have had a lot of tolerance for them over the years but now I am at a low...

Cheers
Eboli

[Sjorpha]

Random numbers don't guarantee a distribution within a limited deviation at all, especially not over a time as short as this.

That is the main problem with the event, the RNG is guaranteed to generate a number of uneven results where many players will not ever get certain maps. I'm sure that a few unlucky players saw things like getting the same two maps all weekend etc, which is to be expected with any random system.