30 replies to this topic

[Bront]

Individual user data is anecdotal. You'd need to track 10000+ drops to get better tacking numbers. I've gone though weeks where I never saw one of the maps, and other weeks where I droped on that map 3 times in a row. Random chance doesn't state that you'll drop on every map equally, just that you'll have a chance to drop on them all equally next time.

[Henry Morgan]

Bront, on 05 December 2013 - 11:55 AM, said:

Individual user data is anecdotal. You'd need to track 10000+ drops to get better tacking numbers. I've gone though weeks where I never saw one of the maps, and other weeks where I droped on that map 3 times in a row. Random chance doesn't state that you'll drop on every map equally, just that you'll have a chance to drop on them all equally next time.

In a very general sense, you're correct about randomness.. However there's also probability that comes into play. True, Random chance doesn't say it will be 100% equal across the board. But, over time, it also should level out somewhat as probability also comes into play. That is, if the 'random' part is truly random.

Unfortunately, since we don't have access to the PGI data warehouse to mine, we're stuck with individual user data. And, however anecdotal it may be, over time it should be fairly consistent. And one thing that does appear to be consistent is that the high heat maps appear to be more heavily weighted than the low heat maps.

[Mercules]

Henry Morgan, on 05 December 2013 - 05:11 PM, said:

But, over time, it also should level out somewhat as probability also comes into play. That is, if the 'random' part is truly random.

https://yourlogicalf...amblers-fallacy

[Henry Morgan]

Mercules, on 06 December 2013 - 10:04 AM, said:

https://yourlogicalf...amblers-fallacy

http://en.wikipedia....al%27s_triangle

http://www.mathsisfu...s-triangle.html

[Bront]

Henry Morgan, on 05 December 2013 - 05:11 PM, said:

Unfortunately, since we don't have access to the PGI data warehouse to mine, we're stuck with individual user data. And, however anecdotal it may be, over time it should be fairly consistent. And one thing that does appear to be consistent is that the high heat maps appear to be more heavily weighted than the low heat maps.

Anecdotaly, I'm not finding that true as of late (Past week), numbers wise I haven't tracked since the last patch, but my overall play numbers I don't think support that.

The other problem you have is you're chasing data on the forums, which is a small (and often quite angry) subset of people, so expecting their data to be representative, or even honest isn't enough to show the true numbers.

Beyond that, perceptions tend to notice the bad more than the good. Since most folks think heat maps = bad, they tend to notice them more. Someone drops 10 times, but get 2 drops on Terra Therma, they could easily think "This is my second TT drop, it's coming up more often than it should", ignoring the 2 frozen city drops.

Also, to be fair, many of the cold maps have issues as well (Alpine being huge, Forrest Colony Snow being small, Frozen City has people who either like or hate it), while the hotter maps are OK maps with annoying heat (Heat I think is the biggest reason folks hate TT).

[Diego Angelus]

Bront, on 08 December 2013 - 06:56 AM, said:

Anecdotaly, I'm not finding that true as of late (Past week), numbers wise I haven't tracked since the last patch, but my overall play numbers I don't think support that.

The other problem you have is you're chasing data on the forums, which is a small (and often quite angry) subset of people, so expecting their data to be representative, or even honest isn't enough to show the true numbers.

Beyond that, perceptions tend to notice the bad more than the good. Since most folks think heat maps = bad, they tend to notice them more. Someone drops 10 times, but get 2 drops on Terra Therma, they could easily think "This is my second TT drop, it's coming up more often than it should", ignoring the 2 frozen city drops.

Also, to be fair, many of the cold maps have issues as well (Alpine being huge, Forrest Colony Snow being small, Frozen City has people who either like or hate it), while the hotter maps are OK maps with annoying heat (Heat I think is the biggest reason folks hate TT).

There is nothing to notice stats says more then enough for example my most noticed maps are one I enjoy but stats say its Crimson straits follow by thera terma(I don't hate it because of heat i hate map design and stupid terrain) difference between those two is like 15 matches so Its not only hate its a fact as well.

[Bront]

Diego Angelus, on 08 December 2013 - 07:07 AM, said:

There is nothing to notice stats says more then enough for example my most noticed maps are one I enjoy but stats say its Crimson straits follow by thera terma(I don't hate it because of heat i hate map design and stupid terrain) difference between those two is like 15 matches so Its not only hate its a fact as well.

Give me the map stats of 100 different random players (and definitely not folks you group with, as that skews the data) and then tell me.

Just a note on randomness:
1) PGI has never admitted it, but I expect newer maps get a slight bump in frequency to work out any final bugs AND to let folks give it a run down. TT seemed to show up a lot when it was new, and CS is the newest map we have.
2) The maps with 2 different modes seem to be down in frequency individually because they seem to be splitting drops between them in my observation. I don't have data to prove it, but most stats I've seen do seem to show that River City + River City Night and Forest Colony + Forest Colony Snow drops tend to be similar to the most frequent map choice for most people in total drops, so it's possible they're splitting their chances, which makes each individual map come up less often, skewing data towards maps that don't have a pair and thus drop more often.
3) If the above 2 things are true, the cool vs hot balance is hurt by having half a cold map (FCS) vs a new hot map + a hot map getting a bump in commonality (and admittedly, about 300 matches in, Caustic was the map I dropped on the least, so there may have been something to that).

Now, I'm not drawing any hard data here beyond what I've seen posted and my own drop stats, so untill I can see PGI's stats, I can't definitively say if any of that is true.

[Henry Morgan]

Bront, on 08 December 2013 - 06:56 AM, said:

The other problem you have is you're chasing data on the forums, which is a small (and often quite angry) subset of people, so expecting their data to be representative, or even honest isn't enough to show the true numbers.

Initially I was looking at the forums, after I was thinking "Wow, I seem to get a lot of hot maps", and found a number of posts by people complaining about the same sort of thing. Thus, I started keeping track of the maps I loaded into, to provide something of a data source for my own experience, rather than a 'gut feeling'. Which is a little different than the generalities stated in most of the forum posts.

Bront, on 08 December 2013 - 06:56 AM, said:

Beyond that, perceptions tend to notice the bad more than the good. Since most folks think heat maps = bad, they tend to notice them more. Someone drops 10 times, but get 2 drops on Terra Therma, they could easily think "This is my second TT drop, it's coming up more often than it should", ignoring the 2 frozen city drops.

Which is why I started tracking the matches. I agree, we remember the maps we dislike more, and I wanted to see if the data from tracking was in line with, or different what my perception was. Perception is one thing. But getting hard numbers to start showing a pattern goes beyond perception.

I did start keeping track of the matches after the last patch separately, in case there was a change. After 124 matches, they're trending very much in line with the same trend before the patch. And the variance between what I'm getting, and what the normal percentage should be is too great to chalk off. Especially since the non-hot/cold maps are pretty much in line (both before and after last patch) with what they should be.

[Sandpit]

Bront, on 08 December 2013 - 07:44 AM, said:

Give me the map stats of 100 different random players (and definitely not folks you group with, as that skews the data) and then tell me.

Just a note on randomness:
1) PGI has never admitted it, but I expect newer maps get a slight bump in frequency to work out any final bugs AND to let folks give it a run down. TT seemed to show up a lot when it was new, and CS is the newest map we have.

I thought this was pretty well-known common knowledge. They bump rotation of a new map.

Map stats under your profile will give you exact number of matches played. You'll see trends there sometimes but over 1000 matches you'll see things are pretty even in distribution.

[Bront]

Sandpit, on 10 December 2013 - 04:40 PM, said:

I thought this was pretty well-known common knowledge. They bump rotation of a new map.

It's a common assumption, though it's not common knowledge (mostly because people complain about it).

[Henry Morgan]

Sandpit, on 10 December 2013 - 04:40 PM, said:

I thought this was pretty well-known common knowledge. They bump rotation of a new map.

That's not a real surprise. As maps become more numerous, and a new one is added, they'll want all the players to get a good opportunity to play it a number of times before it becomes a 'standard' rotation. But, I purposely started tracking my numbers after the latest maps were already released and in the stream for a bit. I wanted to track when the maps had a equal (supposedly) opportunity to cycle, and not run into variables such as map availability.

My primary curiosity behind this was not so much an individual map's frequency, but rather the "hot" versus "cold" maps, as groups, and their frequency. At the start, I only had my perception that it seemed like I was getting "hot" maps more frequent than the "cold" maps, even though there are more cold maps than hot ones.

I understand that random is random, and it won't be exactly perfect. Especially when talking about an individual map. However, when talking about grouping maps as a type, and tracking them, the randomness should be more prevalent, which is not what I'm seeing.

After almost 500 matches since starting the tracking, the "hot" maps as a group are continually a higher percentage of appearing in a match than a "cold" map. To the point the percentage is pretty much reversed as what would be normally expected. That's why I think something is borked with the map rotation, and that hot maps are being weighted much heavier than cold maps. Not just a single map, but as a group.