30 replies to this topic

[Obadiah333]

After the patch last night, I dropped eight solo matches. 7 of the 8 were on terra therma. I got that map 5 times in a row during the 8 match stint. 5 of those were conquest (I had selected any for battle type). 7 of 8 on terra therma. Yeah, that sucked. It's not that I mind that map, it's an ok map. However, I don't want to drop 8 matches and play that map 7 effing times during. I would have played more matches, but I wasn't about to sit through 5 more games of terra therma conquest. FIX YOUR MAP ROTATION.

[Sandpit]

I think they should make changes to the map rotation as well. I have found myself dropping in one map several times in a row and it's not uncommon. I think they should code it so that after twice in a row that map isn't in your rotation for at least 2 matches.

[Voivode]

This was worse when there were fewer maps. Early December patch gives us another map so the chance of repeat dropping will be even lower. As the maplist grows this will become more of an oddity.

[Sandpit]

Voivode, on 20 November 2013 - 12:10 PM, said:

This was worse when there were fewer maps. Early December patch gives us another map so the chance of repeat dropping will be even lower. As the maplist grows this will become more of an oddity.

The number of maps is one part of it but having a drop5-6 times in a row on the same map? That's not due to the number of maps we have. That's an issue with the RNG that picks the map. ANY kind of mechanic that relies on an RNG is going to have the exact same % statistically to pick the exact same number over and over and over again. A fix to that is telling that RNG that after it has picked the number 2 twice it cannot pick 2 again for xx amount of times

[Bront]

I dropped 9 of 12 drops on Caustic last night.

It happens. It's mostly luck of the draw, and even if they smooth out the % doesn't mean you'll always drop evenly.

[Sandpit]

Bront, on 20 November 2013 - 01:03 PM, said:

I dropped 9 of 12 drops on Caustic last night.

It happens. It's mostly luck of the draw, and even if they smooth out the % doesn't mean you'll always drop evenly.

I don't think being completely even is the goal or thought. Just that dropping on the same map more than a few times in a row, which for some casual players might be their entire night of playing, is not a good thing.

[Mercules]

Sandpit, on 20 November 2013 - 12:19 PM, said:

The number of maps is one part of it but having a drop5-6 times in a row on the same map? That's not due to the number of maps we have. That's an issue with the RNG that picks the map. ANY kind of mechanic that relies on an RNG is going to have the exact same % statistically to pick the exact same number over and over and over again. A fix to that is telling that RNG that after it has picked the number 2 twice it cannot pick 2 again for xx amount of times

On top of Elo and weight balancing you want the MM to add in, "Player X can't be on that map."?

[Bront]

Mercules, on 20 November 2013 - 01:13 PM, said:

On top of Elo and weight balancing you want the MM to add in, "Player X can't be on that map."?

Yeah, random chance means eventually it should work itself out, but people will see anomalies all the time.

[Sug]

Bront, on 20 November 2013 - 01:03 PM, said:

I dropped 9 of 12 drops on Caustic last night.

I actually forgot that map existed. Got it last night and had no idea where I was. I haven't seen it since tera therma came out.

Been playing consistently since CB and I've already played on Tera almost as much as the other maps.

And Tourmaline Desert.....why have I played on that map 100 more times than any other map?

Edited by Sug, 20 November 2013 - 02:41 PM.

[Sandpit]

Mercules, on 20 November 2013 - 01:13 PM, said:

On top of Elo and weight balancing you want the MM to add in, "Player X can't be on that map."?

I'm saying they need to adjust the RNG for the betterment of the game.

Example:

Casual Player A gets to play 1-2 hours a night. They get 10 matches in that time frame. They drop on one map 9 times. Don't you think they might get a bit bored with that?

Casual Player B gets to play 6 hours a week periodically here and there for 30-40 minutes at a time. This player spends 90% of their drops on the same map. Over and Over and Over and Over (well you get the point)

This game is already repetitive enough. One of the only things you have that offers some diversity (other than mechs and loadouts) is getting to play through the different maps. If a player spends their entire game time playing the same map repetitively, that's not good.

I don't care HOW they implement it but there really should be a mechanism that ensures that doesn't happen.

[Henry Morgan]

Voivode, on 20 November 2013 - 12:10 PM, said:

This was worse when there were fewer maps. Early December patch gives us another map so the chance of repeat dropping will be even lower. As the maplist grows this will become more of an oddity.

Actually a few weeks ago I was noticing the same thing. It seemed like I was getting certain maps more frequently than other maps. So, I started tracking the maps I got with the random map selector for matches. The data is for my last 100 matches, and I was running as a PuG with "any" mode selected.

I was also tracking Assault and Conquest modes I got, but I think there are other variables I wasn't tracking such as time of day, number of pre-mades online, etc... that could impact those numbers. As a PuG, I'd be 'filler' for pre-made groups, and those tend to lean towards Assault mode only maps.

But here were the numbers I came up with over 100 matches:

The map breakdown with the Assault and Conquest also tracked. I used Red to highlight the "Hot" maps, and Green to highlight the "Cool" maps. The maps with no highlight are the "Standard" temp maps.




Then I just charted it out to make the visual trending a bit more easier. Red are the "Hot" maps, Blue are the "Cool" maps and Yellow is the "Standard" maps.



Based on the numbers, I had;

40% chance of getting a heat "Standard" map
39% chance of getting a heat "Hot" map
21% chance of getting a heat "Cold" map

Which I thought was interesting as there are less heat Hot maps than the other types. Or that my least visited Hot map was still greater than the most visited Cool map.

Granted 100 matches isn't a large amount in the overall view of how many matches are done every day. However I only started tracking it after my perception was that I was getting a lot of a certain type of map, but really couldn't provide any evidence beyond "I think". And, I didn't want to go back into the Stats part and dump the Maps data, as I don't know how far back it does, and if it will be unbalanced with historical maps getting more hits overall than a newer map.

But it would appear the Map "Random" Rotation does appear to be weighted for the hotter maps. At least in my case.

[Sandpit]

I have no doubts that over the long term the maps become relatively even with their spread. I'm more concerned about the casual player who has limited time and plays for short periods at a time. If a player plays 10 hours a wek and 75% of those are on the same map for that week that just compounds the repetitive play we currently have. The casual player isn't going to be on long enough to see that maps even out.

[NRP]

I still say PGI should let the players vote for a map from 2 or 3 choices during the matchmaking screen. Giving players a choice is a good thing.

[Roadbeer]

As many of you know, I track my matches, so here are my last 40:

Forest Colony: 3
Forest Colony (Snow): 3
Caustic Valley: 1
Frozen City: 2
Frozen City (Night): 4
River City: 2
River City (Night): 2
Alpine Peaks: 1
Tourmaline Desert: 1
Canyon Network: 7
Terra Therma: 8
Crimson Strait: 6

And while Terra has the highest number in the rotation, not once was it streaked, where the Crimson was in two batches of 3 in a row, same with Canyon Network.

Random does not mean equal distribution, it means random, and there are chances of "streaks"

[Sandpit]

Roadbeer, on 23 November 2013 - 10:28 AM, said:

As many of you know, I track my matches, so here are my last 40:

Forest Colony: 3
Forest Colony (Snow): 3
Caustic Valley: 1
Frozen City: 2
Frozen City (Night): 4
River City: 2
River City (Night): 2
Alpine Peaks: 1
Tourmaline Desert: 1
Canyon Network: 7
Terra Therma: 8
Crimson Strait: 6

And while Terra has the highest number in the rotation, not once was it streaked, where the Crimson was in two batches of 3 in a row, same with Canyon Network.

Random does not mean equal distribution, it means random, and there are chances of "streaks"

yup, that was my meaning behind the RNG post earlier. RNG means every number has an equal % of coming up every single time. It doesn't bother mme personally when I do get some streaks and drop on the same map several times in a row. I'm playing the game and having some fun regardless of what map I'm on. My concern is that casual player that isn't "in love" with the PI and plays well casually lol
If they drop in game and they tend to get those streaks then they're essentially playing on the same map over and over and over in their eyes. Which, I think, might also lead to some of the map hate threads we see from time to time. If they just set up the rotation where a player has a map eliminated from their queue after hitting it maybe 3 times in a row so that 4th game is guaranteed to be a different map it would alleviate some of that. Then after that 4th map it goes back to business as usual (unless that 4th map hits 3 in a row and then it gets eliminated after that 3rd match)
The only problem I an see with that might be queue times increasing sometimes. It's a trade-off and it would jsut depend on which way the player base would rather have it.
Repetitious maps or longer queue times. I personally think once they saw longer queue times they'd much rather have map streaks but they'll continue to complain about it until they see the other option. I did see where the devs are working on a map vote feature and if that's the case this entire conversation will be moot anyhow.

[Roadbeer]

There is also a bit of confirmation bias in play, when you don't like a map, every time you see it reinforces your dislike and feeds the "I'm seeing it too much" view, just as when you like a map, you don't feel you see it enough.

I love Canyon and Crimson and I don't feel I see them enough, but looking at my rotation, I actually see those two quite a bit more than the others.

[Navy Sixes]

Roadbeer, on 23 November 2013 - 10:55 AM, said:

There is also a bit of confirmation bias in play, when you don't like a map, every time you see it reinforces your dislike and feeds the "I'm seeing it too much" view, just as when you like a map, you don't feel you see it enough.

I love Canyon and Crimson and I don't feel I see them enough, but looking at my rotation, I actually see those two quite a bit more than the others.

For the past two nights, I've been stuck in Forest Colony/Frozen City (with a few side-trips, here and there). I hate these maps, and was ready to hit the forums to drop some post-rage, when I went on a streak of Caustic Valley/Crimson Straight runs in the wee hours this morning. Unfortunately (for my stats as well as the poor folks who had to play with me... Sorry, all.) I was half-asleep by then and finally gave up and had to crash.

So, yeah, I agree with Roadbeer. Streaks happen and they suck, and I agree there is a reinforcement bias when you get maps you hate over and over. But try to remember that waiting forever to drop sucks more. If the matchmaker (already of questionable functionality) has to keep track of everyone's map history and make sure you don't get the same map twice... Search Warrior Online.

Edited by Tycho von Gagern, 23 November 2013 - 12:40 PM.

[Henry Morgan]

Just a follow up from my previous post in the thread. I've continued the map tracking and here are my current results.

Up to this point of time I've tracked 333 matches. All of which were tracked after the latest maps were added, so the numbers wouldn't be skewed due to availability.

While there was some movement between the individual maps since the last time, the general theme did seem to remain the same. That being 'hot' maps hit the rotation more often than 'cold' maps, even though there are less hot maps.

The latest numbers (click to enlarge):


And the individual map totals plotted out in a chart (click to enlarge):


I understand things like streaks can happen with maps, as the random number generator is probably not a true random number generator. However, after 300+ matches I would expect to see some normal leveling starting to take place.

Based on my tracking:
Chance to get a 'cold' map: 22.2%
Based on true random, Chance to get a cold map should be: 33.3%

Chance to get a 'hot' map: 34.8%
Based on true random, Chance to get a hot map should be: 25%

Chance to get a 'normal' map: 42.9%
Based on true random, Chance to get a normal map should be: 41.7%

There seems to be a leaning towards getting hot maps over cold maps, even though there are fewer of them in the rotation.

Even within the heat classifications, there is some imbalance in the rotations. Take the normal heat maps, and the difference between Canyon Network (15.6%) and regular Forest Colony (4.2%). The true random percentage should be 8.3% for each individual map. Canyon Network is 2 times the true random chance, and regular Forest Colony is half the true random chance.

Overall, from the first 100, the percentages of chance to get a map by heat classification seem to trend pretty steadily. Which I wouldn't have expected when adding another 200%+ units of data to it. I understand 'streaks' can happen, but the streaks should be random as well, if it is a truly random system.

I still suspect the map rotation isn't truly random, and that some maps have higher weights with a greater chance of appearing than others.
Which, if true, can also have some direct input into the use of ballistics over energy weapons. If I'm almost twice as likely to get a hot map over a cold map (even though there are fewer), then ballistics makes the better choice over time.

[80sGlamRockSensation David Bowie]

Map Rotation is not broken,

Its just highly weighted towards specific maps and Working as Intended™


Why? PGI Reasons™

[Vegetal Maldito]

It IS weighted. I played half the matches in Alpine than in any other map. And I love Alpine....don't know why they do that.

As highlander pointed out, PGI Reasons....