RWG Gateway Overkill

[Riamus]

So I loaded up an RWG map into Teragon to get a preview of the map and noticed how many gateway tiles there are on each town.  The gateway tiles are the light green tiles.  There are a TON of them.  I know they wanted multiple roads connecting to towns and I appreciate that.  But doesn't this seem a bit much?  Especially when some are actually touching corners with one another.

 

This also lets you see the disparity between how towns are in the forest biome compared to other biomes.

[zztong]

It looks like more gateways empowers not having weird and multiple roads running around settlements. This way there's more than one tile providing an exit from some settlements.

 

With that many gateways as exits from settlements, the few gateway T tiles available will be getting a workout.

[Riamus]

26 minutes ago, zztong said:

It looks like more gateways empowers not having weird and multiple roads running around settlements. This way there's more than one tile providing an exit from some settlements.

 

With that many gateways as exits from settlements, the few gateway T tiles available will be getting a workout.

Yeah, but this?  Lol!

 

There should at least be a minimum distance between them of 2-3 tiles, imo.  Like I said, I do like that there are multiple connections for roads, especially on larger towns.  I just think this is overkill. 

[theFlu]

45 minutes ago, Riamus said:

Yeah, but this?  Lol!

Seems a little bonkers, but look at the road system that spawned them there; that's all knots

 

Probably not intended to look like that, with or without the checkpoints; but randomness is just hard to get "right". Hoping they'll get it 'perfect' eventually, ofc

[YourMirror]

In what way are they in your way while playing the game?

[8_Hussars]

4 hours ago, Riamus said:

Yeah, but this?  Lol!

 

There should at least be a minimum distance between them of 2-3 tiles, imo.  Like I said, I do like that there are multiple connections for roads, especially on larger towns.  I just think this is overkill. 

To be fair, it is RWG and random outliers will happen.  It appears this issue is identified, based on a sample size of 1.  Perhaps, give it 20-30 maps to get a feel if the gateway tiles are too densely used or if this map contains an outlier.

[theFlu]

4 hours ago, 8_Hussars said:

To be fair, it is RWG and random outliers will happen.

I used to fully agree, nowadays though, "random outliers" come with any RNG system that does not properly weed them out. If you're using an RNG, you have designed every random outlier to be part of your intended feature set - it's just as much a bug as if a designer made it manually.

 

As in, Yeah, rare enough issues will happen, but that doesn't actually change anything, it's still either a bug or a design flaw

[Riamus]

3 minutes ago, theFlu said:

I used to fully agree, nowadays though, "random outliers" come with any RNG system that does not properly weed them out. If you're using an RNG, you have designed every random outlier to be part of your intended feature set - it's just as much a bug as if a designer made it manually.

 

As in, Yeah, rare enough issues will happen, but that doesn't actually change anything, it's still either a bug or a design flaw

Exactly.  If it is not intended to have so many in a group, then you do what I said earlier and have a minimum distance added to the RNG generator.  RNG doesn't have to have such outliers and can be programmed to remove them once identified.

[Ramethzer0]

I think if you make a map where the town density and the map size allotment meet at a very dense cruxpoint, then you're always gunna generate maps like this.

 

One thing I've learned after generation about 30+ maps with varying settings, is that the those settings matter more than people think.

 

Edited July 4 by Ramethzer0 (see edit history)

[8_Hussars]

1 hour ago, theFlu said:

I used to fully agree, nowadays though, "random outliers" come with any RNG system that does not properly weed them out. If you're using an RNG, you have designed every random outlier to be part of your intended feature set - it's just as much a bug as if a designer made it manually.

 

As in, Yeah, rare enough issues will happen, but that doesn't actually change anything, it's still either a bug or a design flaw

Random outliers are part of any system.  Yes, its a design decision to use RNG, however it is neither a bug nor a flaw as it is an inherent part of any system.

The question I had; is the issue detrimental to gameplay and does the issue occur often enough to be worth any correction effort?

IMHO, It is not detrimental at all (just weird) and based on a sample size of one, it does not need addressing.

If as I suggested above, it was much more common then a case could be made to check for it/correct it.

Edited July 4 by 8_Hussars (see edit history)

[theFlu]

8 minutes ago, 8_Hussars said:

Yes, its a design decision, however it is neither a bug nor a flaw it is an inherent part of any system.

Rare RNG events are inherent, bugs caused rarely are still bugs, "poor design outcomes" occurring rarely are still "poor design outcomes".

 

Whether this is "acceptable design-by-RNG" can only really be decided by TFP, but Not reporting on some weirdness isn't going to help them. Reporting on it might.

[Laz Man]

6 hours ago, 8_Hussars said:

To be fair, it is RWG and random outliers will happen.  It appears this issue is identified, based on a sample size of 1.  Perhaps, give it 20-30 maps to get a feel if the gateway tiles are too densely used or if this map contains an outlier.

 

The birds eye looks a bit weird but having more gateways for the big city creates a denser highway vibe around it which gives it a distinct feel that is different to the smaller settlements.

 

I'm sure rwg will continue to get more improvements over time.  

 

 

[8_Hussars]

12 minutes ago, theFlu said:

Rare RNG events are inherent, bugs caused rarely are still bugs, "poor design outcomes" occurring rarely are still "poor design outcomes".

I refer back to the standard normal distribution, and the occurrence of 3 and 4 Std Deviation outcomes where they are not considered bugs or poor design outcomes, just part of the system (random or not).

Its a point on which we agree to disagree.

 

[theFlu]

2 minutes ago, 8_Hussars said:

where they are not considered bugs or poor design outcomes, just part of the system (random or not).

Who/where/what is the judge of "this is rare enough not to be a True Scotsm.. umm, Bug(tm)"?  You're running on some industry standard (on some type of event) and framing it like it's a universal rule?

 

Bugs can be chosen to be ignored due to a combination of rarity and low severity of the outcome, but the decision should be made by the party responsible - for artistic endeavors such as this, it's literally just a matter of the creator's opinion.

[8_Hussars]

22 minutes ago, theFlu said:

Whether this is "acceptable design-by-RNG" can only really be decided by TFP, but Not reporting on some weirdness isn't going to help them. Reporting on it might.

I am not saying don't look at it, I am saying look at the data and see if its a problem.

Decisions based on little or no data and CYA statements have cost society greatly.  There no documented cases of cell phones starting fires while refuelling vehicles.   It was originally recommended by the phone manufacturers as the phones are not intrinsicly safe (CYA).  Yet, the "myth" it has impacted refuelling and payment schemes for years.  There is a risk but its 1 in 10 billion.

The original Dayton airbag incident, sent the fire services on a 30 year diversion, over worrying about SRS.  SRS is a legitimate risk and knowledge is power but extrications were delayed that likely cost people lives due to the acceptance of a one off incident as indicative of the situation in general.       

[theFlu]

Hmm, I wouldn't compare "gas station operations" and "misfiring airbags" with "spending a couple of hours on tweaking RWG so that a 100 ppl won't see a weird cluster of POIs in their game". The formers play with potential risks, here, whether something is changed or not, nothing is changing..

 

And if "wait for more data" was your original intent, the framing of things as "neither a bug nor a flaw" and such, well, they seem a little confusing. But with RWG here, they could also just "look for more data" rather reasonably, they'll prolly have 100s of ready-made maps littering their hard drives, could just visually inspect the distributions. And/or tell the people working on RWG about the issue, and they'll semi-automatically start spotting it if it happens.

[zztong]

On 7/4/2024 at 4:17 AM, Riamus said:

Yeah, but this?  Lol!

 

I see what you mean. Fewer Gateway T's is perhaps best when there's a limited selection of those Tiles though in my mind this just screams for more Gateway T blocks and the heavy use of Parts on those Tiles.