Duel40a (May 27, 2015)

[RoSKing]

What's the point of keeping around maps that never get played but besides a select few players

If you really want to go play the map, go host it yourself. These are suppose to be not only competitive means, but also to keep attracting players to come back. There's absolutely no benefit of removing maps that get played more than pobla, boo, and majority of the remaining list.

[Reaku]

We need more Dwango 1 in Duel32! I see the others, but why not a true classic?

[Catastrophe]

wait, so you remove SIPOOMA but not foxtrot? What sense does this make? SIPOOMA was played more than twice as much and played almost as much as some of the "standard" 1v1 maps. Even that piece of shit exodus got less plays than SIPOOMA.

What? Foxtrot is a really good map.

[AlexMax]

What's the point of keeping around maps that never get played but besides a select few players

Here are stats as of today:

Code: Select all

Map                   Pick PComp   PC%  Rand RComp   RC%
boo                     44    31   70%    60    54   90%
brit10 map07            35    29   82%    42    39   92%
brit11 map01            38    17   44%    61    55   90%
brit11 map05            17     9   52%    59    53   89%
class219 map01          89    74   83%    66    63   95%
doom2 map01             75    58   77%    65    54   83%
dwango5 map01          424   391   92%    46    44   95%
dwango5 map07          102    85   83%    55    48   87%
dwango6 map01           41    31   75%    55    49   89%
dweller2 map11         112   104   92%    49    43   87%
eondm1e map05          157   134   85%    67    64   95%
exec map03              52    39   75%    47    45   95%
exodus1on1              30    24   80%    62    57   91%
exposed                 53    46   86%    58    55   94%
eyedea                  46    37   80%    52    47   90%
foxtrot                 50    42   84%    58    51   87%
greenwar map20         272   259   95%    51    48   94%
greenwar map24          77    66   85%    61    54   88%
fuloser                 53    42   79%    63    59   93%
judas23_               655   609   92%    47    44   93%
king1                  320   298   93%    59    57   96%
lazarus1j map01         32    24   75%    51    50   98%
lazarus1j map04         70    64   91%    57    52   91%
moo2d                  165   150   90%    62    59   95%
oblivx                 164   148   90%    54    52   96%
pobla8                  51    46   90%    37    31   83%
ssl2                   496   458   92%    60    58   96%
styxdm01 map01         171   156   91%    46    43   93%
styxdm01 map02          55    42   76%    49    45   91%
uroburos                70    59   84%    57    55   96%
tedp map01              54    34   62%    46    39   84%
xenosx9 map04           53    37   69%    53    51   96%

shoot                  689   637   92%     0     0     -
bfgd5m1                 49    34   69%     0     0     -

Take a look at Doom2 MAP01 and Pobla8, both are "controversial" maps and both have an 83% random completion rate, the lowest of any of the maps. Does that mean that I should yank out both of them? Dwango6 MAP01 has 41 picks, which is near the very bottom. Should I yank Dwango6 MAP01 because of it? Brit11 MAP05? Lazarus MAP01?

Generally, I've removed and added maps using a combination of these stats and player feedback. But I'm not so sure that the status quo is maintainable - I hate retreading the same ground over and over again when I do a duel32 release. There are a couple of ways forward:

• Put more priority on the stats as a measure of popularity (picks) and quality (random completions). I'm afraid that this would homogenize the style of maps that are kept around. I'm probably not going to remove doom2 map01 anytime soon, but other controversial maps might not have that predigree to fall back on...what happens to them?
• Put more priority on player feedback as a measure of quality. This generally turns into "whoever yelled at AlexMax most recently/convincingly" - the decision to go with 32 maps instead of upping to 40 or 48 was the direct result of this, as well as some of the recent removals. Also, keep in mind that depending on who you ask even popular maps like greenwar map20 and uroburos can grate on players.
• Up the maplist to 40 or 48. The more maps are in the pack, the less likely I am to remove someone's controversial favorite, but more likely that maps that...just plain stink...stick around, the worse discover-ability is for new maps, the harder the pack is to maintain, et cetera.

I don't know the right answer, I'm just putting these issues out there. I think that having good maps should be a priority, and that bad maps should be able to be removed, but I also want the pack to have something for everybody, not just the majority or ones with easiest access to me via IRC.

[AlexMax]

To take this picture I pressed against the Skip (currently Random) warp and backed away really quickly, to show the two warps line up. Either player can accept the random map on their own, it takes both players to skip, this is to keep stalling to a minimum. Yes, I know the name plate is out of alignment.

There's also another idea floating around in my head called Pick 4. It would select 4 random maps from the 32 and allow players to pick from those 4. Two problems:

• Twice the potential places to stall picking.
• I seriously have nowhere on the map to put the warp to Pick 4. It's not a substitute for Random, so it can't replace that.

[AlexMax]

Duel32e is in final testing on FUNCRUSHER and will be available for general release within a day or two.

- The START map has been completely overhauled. Brought back the skybox from older releases and gave the insides a good once-over.
- EONDM1E MAP05 is upgraded to EONDM1F MAP05
- TEDP MAP01 is gone, replaced with EONDM1F MAP14
- RANDOM will now bring you to the Random Dimension. If either player is happy with the randomly picked map, you can play it, or if both players would rather skip to another map that can be done as well. You can see a picture of the completed working Random Dimension here.
- BFGD5M1 has been updated and is now stocked with infinicells, which never vanish and always bring your cell count to 40 and 200 respectively.
- A replaceable 256x128 texture called _AD_256. Run a server? You can replace this 256x128 texture with whatever you like via a PWAD.

EDIT: Duel32e has been released. If you recorded any demos with test3, that is identical to the final product. If you recorded any demos with test1/2 toss them.

[AlexMax]

Happy New Year! Here's a list of everybody's win count since I started taking consistent logs a few years ago:

http://files.funcrusher.net/static/duel_wins_2013.txt

Congratulations to player for being the undisputed champion of duel32!

Generated with:

Code: Select all

cut -d ' ' -f 2- /opt/zandronum/log/duel32-*.log | grep -v '^CHAT' | grep 'wins!$' | tr '[A-Z]' '[a-z]' | sort | uniq -c | sort -nr

[Combinebobnt]

Damm player is pretty good

[IdeIdoom]

Fuck me, this is old:
1 ide idoom[dn] wins!

[Konar6]

Thanks for coming with another stats idea, AlexMax!

Happy New Year! Here's a list of everybody's win count since I started taking consistent logs a few years ago:

Ditto for GV:

http://sickedwick.net/misc/uk.txt
http://sickedwick.net/misc/fr.txt
http://sickedwick.net/misc/tx.txt

Looks like we have some more interesting winner names here ;)
Congrats Skinlainen, IdeIdoom and Strych.

[Kara Kurt]

Ew, NRM on top!

Damm player is pretty good

I don't know if the fact that stats of all players called and playing as player were all collected together.

[one_Two]

Ew, NRM on top!

Damm player is pretty good

I don't know if the fact that stats of all players called and playing as player were all collected together.

Yeah obviously.

[AlexMax]

I'm still taking community map requests for a potential Duel40/Duel48. So far, though, the only things that are for certain is that I am replacing lazarus1j map01 with lazarus1j map16 and bringing back greenwar map16 and sipooma due to repeated requests.

Here are a the community submissions I have so far that I'm seriously entertaining:

• coll_n_devi map02 by Collision

A short list so far, so get to submitting. Note that you're not limited to just new submissions, if you have a map as part of a previously released pack that you would like to get player scrutiny on, that's fine too. However, keep in mind that these map submissions are still for oldschool doom - no jumping or freelook should be required, and it should play well with oldschool dmflags, and take it easy with the ZDoom features.

[jwaffe]

I'd like opinions on FastDM MAP07 and FastDM MAP11 as OS duel maps. They're hexen format but they have no thing bridges and could work as at least Boom compatible maps, if not vanilla doom.

EDIT - only thing keeping me from saying they'd work in vanilla doom (if converted to Doom format) is possible visplane/other vanilla limitation issues.

Also just wanted to say I liked Lazarus MAP01 personally, and I'm disappointed to see it go.

[Springy]

I hate pimping my own shit but I would quite like an opinion on these two to see if any of these are worthy? Would be nice if they got some play time or too at least see if these are worthy. http://www.doomworld.com/idgames/index.php?id=17333 http://www.doomworld.com/idgames/index.php?id=17377

[Danzoa]

So many of my aliases in those winlists lol

[Combinebobnt]

Here is an older duel map I made for sfm-duel with some patching up that you can use if you want. Title is "Corridor 9"

[jwaffe]

Here is an older duel map I made for sfm-duel with some patching up that you can use if you want. Title is "Corridor 9"

I support that suggestion, I also recommend MAP22 of that wad, "Fissured Cloister" by Argentum,

[AlexMax]

I'm going to have to release Duel32f soon - I found a bug in the random picker which prevents the last map from being shown as a random map (xenos in this case). However, thanks to additional logging features added to duel32e, I can get a much better idea of what maps people don't like based on how often they're skipped in Random. Here's the duel32e stats so far:

Code: Select all

Map                   Pick PComp   PC%  Rand RComp   RC%  Skip    S% Blind BComp   BC%
brit11 map01             6     6  100%    14    13   92%    10   41%     5     4   80%
greenwar map24          38    32   84%     8     8  100%     3   27%     3     3  100%
dwango6 map01           19    10   52%     3     3  100%    12   80%     4     1   25%
brit11 map05             5     4   80%     8     7   87%     7   46%     6     3   50%
exposed                 47    42   89%    12    12  100%     6   33%     9     9  100%
bfgd5m1                 18    13   72%     0     0     -     0     -     0     0     -
dweller2 map11          46    40   86%    10    10  100%    12   54%     7     6   85%
dwango5 map07           68    55   80%     5     5  100%    10   66%     8     7   87%
dwango5 map01          174   149   85%     8     7   87%     4   33%     7     6   85%
exodus1on1              11    10   90%    11     9   81%    12   52%     9     8   88%
greenwar map20         141   133   94%    20    17   85%     1    4%     6     4   66%
styxdm01 map02           9     6   66%     9     9  100%     7   43%     8     8  100%
eondm1f map05           82    74   90%    11    11  100%     8   42%     9     7   77%
xenosx9 map04           29    21   72%     0     0     -     0     -     5     4   80%
brit10 map07            10     9   90%     9     9  100%    10   52%     7     6   85%
doom2 map01             42    35   83%     7     7  100%    10   58%     4     3   75%
fuloser                 32    31   96%     9     9  100%     4   30%    10     5   50%
judas23_               315   293   93%    17    16   94%     5   22%     6     5   83%
ssl2                   246   230   93%     5     5  100%     5   50%     3     3  100%
class219 map01          65    61   93%    23    20   86%     5   17%     7     7  100%
moo2d                  115   107   93%    17    16   94%     1    5%    15    13   86%
oblivx                  43    39   90%    14    13   92%     0    0%    10     9   90%
exec map03              15    11   73%    12    12  100%    10   45%     5     5  100%
uroburos                49    48   97%    11    11  100%    12   52%     6     6  100%
styxdm01 map01          59    56   94%    18    18  100%     5   21%     8     5   62%
king1                  201   179   89%    19    19  100%     5   20%     8     8  100%
foxtrot                 33    23   69%     8     8  100%     9   52%     5     5  100%
shoot                  252   205   81%     0     0     -     0     -     0     0     -
eyedea                  33    29   87%    10     8   80%     3   23%     5     3   60%
pobla8                  45    38   84%    14    13   92%     6   30%     8     7   87%
lazarus1j map01         16    11   68%    10     9   90%     8   44%    11     9   81%
lazarus1j map04         66    57   86%    14    12   85%     5   26%     6     5   83%
eondm1f map14           37    32   86%     6     5   83%    10   62%     4     3   75%
boo                     65    24   36%    11    10   90%    22   66%    13    12   92%

Based on these numbers there's at least five or six maps that I should remove right off the bat based on these stats alone. However, I'd like to get the bugfix out soon, so who knows...

And no, boo does not actually have a 32% completion rate. That's part of the bug. :)

EDIT: Found and fixed the bug. As I suspected, an off-by-one in the random picker meant that Xenos was never randomly picked and half the time that "boo" was shown as a random pick, you couldn't actually use the map portal.

[one_Two]

I noticed just one person can select the rand pick to choose it?