Map size vs Quality vs Adoption

Terr

Cranky Coder
aa
Jul 31, 2009
1,590
410
If you're trying to popularize a map, do you think it's better to focus on a small file size early on, so people come to like the basics (and are more-willing to download the next, larger, nicer release) or is it better to do a really nice compile to make a good first-impression?
 

Inacio

L1: Registered
Jun 28, 2009
27
1
I think it's better to do the map as good as you can, even if it's an early release.

Though I once made a map in 15 minutes for me and my friends to play on, it was small, ugly and, well, not much of a map.

People played it for more than two hours. :D
 

J4CK8

L11: Posh Member
Mar 4, 2009
820
243
Yeah, as long the map plays well and is balanced, I don't mind downloading.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
<30mb size doesnt even matter. Between 30 and 70mb you should have some decent multi stage (or a long - 5CP linear, 4 cp PL - single stage) map at least (no arena or koth). People might be too lazy to download it. 70mb is however quite large already and people then will complain about the download time. Anything 100mb+ is just too much for most.

For TC maps there however can be an exception on that, those maps are huge and often you cant prevent a large filesize because of that. Dont make the filesize harm you in that case but to try to lower it as much as possible.

If the quality is less good the filesize people are willing to download gets lower. Early on however people dont know your map yet and so the filesize should me smaller (it doesnt matter how good your map is - as people dont know it that doesnt make a diffirence - and a first bad impression often keeps bad).
 

Icarus

aa
Sep 10, 2008
2,245
1,210
Normally, I consider 70MB to be the upper limit regarding mapsize. That's pretty generous. Anything higher gets scary.
 

Radaka

L420: High Member
May 24, 2009
491
244
Generally when you release your map it needs to stand out somehow. If you release a generic dev-textured only map it probably won't get as much attention as one that at least has texturing and preliminary detailing. I still recommend at least a few dev textured tests before moving into texturing though just to make sure the layout somewhat makes sense.

Just don't release your dev textured map to the general public and expect people to think it's something revolutionary.

---My opinion on the topic---
Filesize doesn't usually matter for the first release of a map (it shouldn't be big if it's a first release anyway).
Generally I would lean towards quality, which I'll be following in my next release.
 

Caliostro

L6: Sharp Member
Jul 6, 2009
261
110
I've recently came to the realization people are much like magpies. They can't help themselves when they see shiny things.

If your map looks good and plays decently, people will be all over it like a fat man over pancakes. Conversely, if your map plays pretty good, but only looks decently, it's generally a gamble (or matter of "social status") whether or not people are going to give your map a second look.

See, while people complain a lot about "map sizes!!!1OMG!", truth is nobody cares. They complain because they see the number, not because of how long it takes to download. "100?!? OMG!". See, as stated several times before, unless you're running your game from an old toaster plugged to a dial up modem with a rusty copper wire, it doesn't particularly matter if your map is 30, 50 or 100 mbs. The time it takes nowadays to download files that large is laughable. Shmitz's Meridian was criticized for having 80 mbs! OMG! It takes me around 1 minute at most to download. And that's being generous.

Off course, you do have to justify your map's size though. Waiting, even if it's only an extra 20 seconds, to download a map that's completely dev textured isn't very amusing... Specially considering that if your map is that big now, imagine once it's done...

The difference between 50 and 80 mbs might not be much, but 200 for a perfectly average map is a bit of a stretch...
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
See, while people complain a lot about "map sizes!!!1OMG!", truth is nobody cares. They complain because they see the number, not because of how long it takes to download. "100?!? OMG!". See, as stated several times before, unless you're running your game from an old toaster plugged to a dial up modem with a rusty copper wire, it doesn't particularly matter if your map is 30, 50 or 100 mbs. The time it takes nowadays to download files that large is laughable. Shmitz's Meridian was criticized for having 80 mbs! OMG! It takes me around 1 minute at most to download. And that's being generous.
Many times the server you download the map from has a poor upload for the maps so you could get 100kb at most... with a 60mb map it quicky gets 10 minutes. On a server with a good upload (and they do exist - often that are the most popular servers with custom maps) it indeed doesn't matter. But the problem is that even on the small servers your map needs to be popular.

Btw, im talking about bspzipped files. many times maps are 80mb and after being bsp zipped turn out 40mb. so thats like 4 minutes at 100kbs. Which is a normal time. But what if a map is 100mb bspzipped. At 100kbs thats 14 minutes. people dont like to wait that long... Only if your map has a reputation of having good gameplay already when it was just 30mb people are willingly to wait for it. And when you already ahve the reputation they sometimes are willingly to download allmost any size.

However, on IR servers the map must play well. Hoodoo had the problem of not being balanced well on those servers and because of that people werent willingly of downloading it. Valve at least solved that issue, but people still find it a quite bad map (even though its just stage 1 that needs a change to be more IR friendly).

Most important remains, dont be scared if your map gets a large filesize. All you have to do is ensure you get the playtests needed to build up the maps reputation. Once people start to know about it they will download it more likely.
 

Garner

L4: Comfortable Member
Aug 16, 2009
154
38
Agreed with alias

Also, unfortunately most servers which are'nt a large gaming community or dont have admins who know what their doing tend to upload X amount of decent maps, have cheap / poor servers with bad upload/download speeds and most importantly no sv_downloadurl

I think to popularise your map across the word of gaming communities, is to sign to to their forums and provide links, description and a request to get your map onto their rotation. Im personally tied to Fragmasters.co.uk here in the UK, and i have gotton maps like Swift and Outback onto our rotation. The UK has 4 great communities with TF2 servers. They are Fragmasters.co.uk, UKCS.net, ukmandown.co.uk, and viaclan.com so it may be in mappers interest to promote your maps on those communities to get a wider feedback and the possibility of your map being put on the rotations.

Just my 2cents :D
 

Caliostro

L6: Sharp Member
Jul 6, 2009
261
110
However, on IR servers the map must play well. Hoodoo had the problem of not being balanced well on those servers and because of that people werent willingly of downloading it. Valve at least solved that issue, but people still find it a quite bad map (even though its just stage 1 that needs a change to be more IR friendly).

While I do agree with the rest of what you said (and keeping in mind that, rarely will alphas reach any problematic size like 50 mbs as bz2) , I must, again, stress the fact that you can't, and should not under any circumstances, balance maps for IR servers, unless you intend on making it IR only (and consequently get it shafted along with all the "silly/stupid" maps like Billiard and mario kart. Instant respawn is game breaking by sheer concept: permanent defense. In order to balance a map for IR maps you need to make it brutally unfair for non-IR maps.


See, in normal settings you need to use the buffer time between respawns to complete the given objective (push the cart, cap a point, steal the flag, whatever). If you remove the respawn time, it becomes nigh impossible for the attacking team to ever win unless they're downright annihilating the enemy (e.g: massive spawn camping or outnumbering the enemy). If you want the attacking team to have any chance of winning you need to either force that buffer back by, for instances, creating a longer distance between respawn and objective, or give the attacking team a great enough advantage (like advantage points) to offset the fact that they need to deal with permanent defense.

Once you do this, you completely ruin the balance for any other server that also has respawn times. It's just pure action and reaction.

That said, IR is already a dumbed down version of normal respawn times, favoring spam class over the others, so I fully recommend mappers that don't intend to exclusively work for IR servers to ignore them entirely (as far as map making goes at least).

Garner: You just described terrible servers. If an admin doesn't know what they're doing odds are it's a shitty server anyways, since there's enough easily accessible online documentation to teach them.