Level Progress Diagrams – 2002ADO

[Soundblock]

6 hours ago, Linguica said:

I came across https://mermaidjs.github.io which has a live editor that you can use to make graphs like this:

 

 

The ancient Sumerian hieroglyphs are more on theme, c'mon.

 

Klaatu barada nikto!

Edited December 3, 2018 by Soundblock

[Wraith777]

On 12/3/2018 at 4:31 PM, Linguica said:

I came across https://mermaidjs.github.io which has a live editor that you can use to make graphs like this:

 

Would be really nice to see all this working vise-versa (mostly as a matter of experimenting with AI mapping)

 

On 12/3/2018 at 8:28 PM, fraggle said:

It would be great if you could aggregate all these into poster format; they look compact enough that I assume you could fit all four IWADs on one poster print.

 

Yes, and the best way to do it is.. making it in scalable vector graphics form:

https://www.dropbox.com/s/79ftcxzms6nppzr/DooM-LPD.pdf?dl=0

Now it's easy to make LPD's for the rest of wads) (well, excluding such a layout monstrosities as those found in Eternal DooM or some other BIG map megawads)

[kb1]

6 hours ago, Wraith777 said:

 

Would be really nice to see all this working vise-versa (mostly as a matter of experimenting with AI mapping)

 

 

Yes, and the best way to do it is.. making it in scalable vector graphics form:

https://www.dropbox.com/s/79ftcxzms6nppzr/DooM-LPD.pdf?dl=0

Now it's easy to make LPD's for the rest of wads) (well, excluding such a layout monstrosities as those found in Eternal DooM or some other BIG map megawads)

This is really something special! I have difficulties with spatial navigation, in general, so, for me, this is a very cool way to understand map flow in a Doom goal-related sense, without needing the actual map itself. Essentially, that's what this is, for me: A way to visualize level progression, highly distilled.

 

I wish there was some way to import one of these diagrams into Doom's automap, and show a pointer where DoomGuy currently is in the diagram, and maybe even grey-out areas that have already been visited. Do you think it's possible? (Just an idea)

 

I'm going to have to play through the IWADs again, while checking out the diagrams. I *do* think seeing the diagram on the automap, or via a different keystroke would be awesome. Great job - this was a lot of work, I can tell. Thanks for this.

[Wraith777]

Alright, I think it's time for another update.

Let's look at LPDs for good old WizardWorks double offer:

 

                               Perdition's Gate                                       and                                         Hell to Pay
                   

 

Final verdicts may look somewhat subjective since a lot of maps are pretty short and it's really hard to determine whenever an arena-based map may be counted as linear or nonlinear.

 

Perdition's Gate
Map01: nonlinear
Map02: linear
Map03: nonlinear
Map04: nonlinear
Map05: nonlinear
Map06: linear
Map07: linear
Map08: linear
Map09: linear
Map10: nonlinear
Map11: linear
Map12: linear
Map13: linear
Map14: linear
Map15: linear
Map16: linear
Map17: linear
Map18: linear
Map19: linear
Map20: linear
Map21: linear
Map22: linear
Map23: nonlinear
Map24: nonlinear
Map25: linear
Map26: nonlinear
Map27: nonlinear
Map28: nonlinear
Map29: linear
Map30: linear
Map31: linear
Map32: linear

 

Hell to Pay
Map01: linear
Map02: linear
Map03: linear
Map04: nonlinear
Map05: nonlinear
Map06: nonlinear
Map07: nonlinear
Map08: nonlinear
Map09: nonlinear
Map10: nonlinear
Map11: linear
Map12: nonlinear
Map13: nonlinear
Map14: nonlinear
Map15: linear
Map16: linear
Map17: nonlinear
Map18: linear
Map19: linear
Map20: linear
Map21: linear
Map22: linear
Map23: linear
Map24: linear
Map25: linear
Map26: linear
Map27: linear
Map28: linear
Map29: nonlinear
Map30: linear
Map31: nonlinear
Map32: nonlinear

[joepallai]

Outstanding work, can't wait to see what you do next.

[Wraith777]

17 hours ago, joepallai said:

Outstanding work, can't wait to see what you do next.

 

I fear just one thing - a day, when it's gonna be Eternal Doom...

[Wraith777]

I just thought it'll be interesting to look at LPDs of TnT Evilution's sibling -

                                  Icarus: Alien Vanguard

Map01: linear
Map02: nonlinear
Map03: nonlinear
Map04: nonlinear
Map05: linear
Map06: linear
Map07: linear
Map08: nonlinear
Map09: nonlinear
Map10: nonlinear
Map11: linear
Map12: nonlinear
Map13: nonlinear
Map14: nonlinear
Map15: nonlinear
Map16: linear
Map17: nonlinear
Map18: nonlinear
Map19: linear
Map20: linear
Map21: nonlinear
Map22: linear
Map23: linear
Map24: nonlinear
Map25: linear
Map26: linear
Map27: linear
Map28: nonlinear
Map29: nonlinear
Map30: linear
Map31: nonlinear
Map32: linear

Some comments on this one:

- it's all about getting the red key in the end of every map (except map30)
- more collector maps this time

- nonlinear maps are in majority here

- it was really hard to determine player's progression without replaying a lot of maps since they have extensive mechanics for map morphing from SP to MP (and that changes a lot)

- I was really shocked to realize that map11 is actually pretty linear and straightforward (unlike I always remember it being wide and expansive)

 

[Grazza]

If you take suicide exits into account, many of the levels become a lot less linear, as this removes the need at least for the red key. Or, if you assume the player will always use the shortest route to the exit, a lot more linear, I suppose, but then that is true for any map.

 

And your penultimate point is pertinent too, since MP areas are relevant to SP play in a number of cases (14 and 28 are ones that spring to mind).

 

[Wraith777]

I should've stated somewhere that LPDs count only official pathways intended by level author. Therefore, they don't take into account any exploitable bugs and glitches (grabs, suicide exits, glide etc) that create new pathways to progress thru level. But, if a specific trick/exploit/bug is required to beat the level (as intended by author) this counts as official pathway on LPD. Anyway, it's always possible to extend LPD classification by adding a new class of LPDs that includes such exploits, so that we get normal and extended LPDs.

 

I always assume the shortest way as the main one, but if the alternative longer ways provide significantly different experience (like granting a new weapon, ammo/armor, making it easier to take out monsters from the distance) then they are shown as parallel paths. If both paths are significantly long or there are series of them it definitely makes level more nonlinear.

[Grazza]

1: How do you even know what was intended by the author? Some maps (e.g. in Ksutra) deliberately leave a bunch of possible shortcut ideas, and allow players to decide whether they are useful and if so how to use them.

2: Some tricks/exploits have become so standardly used that excluding them seems to make little sense. Many early levels didn't even take straferunning into account. Likewise, some early authors assumed people knew about AVJs while others didn't. The red key being skippable in many levels of Icarus is obvious and well-known nowadays.

3: Whatever the rationale, seeing, e.g., Icarus map22 described as linear just seems... wrong.

Edited July 10, 2019 by Grazza

[Wraith777]

1) I don't know exactly, I just assumed that there may be maps that execute such exploits as obligatory to perform in order to complete them.

 

2) The fact that tricks/exploits have become standard doesn't mean they are official. Doom is an officially released title with it's rules intended for general public. Original game designers didn't acknowledge those tricks, so their use is not intentional. Again, they may become official when they are made unavoidable by author (IF there are such maps, I don't know, never was interested in such things). For example, in my maps I always try to exclude situations when tricks/exploits can be performed because they are not intended here, however, if after release someone still finds a way to use one of those tricks I don't have anything against it and won't rush to fix it (after all that's the point of everything - to find something hidden and unintended, it adds some sportive interest). That's why I count all red keys in Icarus - because its the rule of the game that you're supposed to find the key and use it (and I don't think Icarus authors acknowledged suicide exits and intentionally made them passable that way).

 

Every classification has it's rules. In this case making an LPD follows layout, mechanics and author intentions of original game.

For map classification regarding player progression (linear/nonlinear) using LPD rules exploitable bugs, tricks and glitches are not taken into account
If someone wants to create alternative LPDs of extended type (with exploits included) they are welcome to try it, no problem.

 

3) Again, this classification has its rules that need to be understand.

In Map22 player may choose of 2 main progression sequences (which are not options):

1)

- go all way to the north until closed yellow door,

- after that go back to graveyard for yellow key

- return to yellow door and lift down red key

- go for red key into tomb

- go to red door and exit

 

OR

2)

- go to graveyard and take yellow key

- go to yellow door and lift down red key

- go for red key into tomb

- go to red door and exit

 

In both cases player can go to tomb with red key first (that makes 4 possible choices), but his progression doesn't change because in all 4 cases key points are achieved in the same order (yellow key > yellow door > red key > red door) and there is only one way to achieve key point every time (except some small forks shown on LPD). Also in case 1 you have a lot of unnecessary backtrack (which makes it different from collector maps).

That's why map22 is classified as overall linear.

This classification is still under development, I need to write down all main rules here with detail.

 

Edited July 10, 2019 by Wraith777

[Diabolución]

Quick question, just by curiosity: Do you plan to analyze in the future the rest of official releases (No Rest for the Living and Master Levels for Doom II)? And certain semi-official releases, such as Betray, Sewers and Sigil? And, perhaps, The Lost Episodes of Doom?

[Wraith777]

No Rest for the Living and Sigil are definitely on my list. The Lost Episodes of Doom.. may be. Overall, I'm interested in covering complete episode/megawad releases first. This goes both for classic (Memento Mori1,2, Requiem, 2002 A Doom Odyssey, Alien Vendetta, Heroes...) and modern (Base Ganymede, Concerned, No End in Sight, Interception, Jenesis, Estranged...) releases I've played before.

[Wraith777]

Ok, time research some true user-made classics:

                                      Memento Mori

Map01: linear
Map02: linear
Map03: linear
Map04: linear
Map05: linear
Map06: linear
Map07: linear
Map08: linear
Map09: nonlinear
Map10: nonlinear
Map11: nonlinear
Map12: nonlinear
Map13: nonlinear
Map14: linear
Map15: nonlinear
Map16: linear
Map17: linear
Map18: nonlinear
Map19: linear
Map20: nonlinear
Map21: nonlinear
Map22: nonlinear
Map23: nonlinear
Map24: nonlinear
Map25: nonlinear
Map26: linear
Map27: nonlinear
Map28: nonlinear
Map29: nonlinear
Map30: nonlinear
Map31: linear
Map32: nonlinear

 

Some notes:
- like in Icarus, a lot of nonlinear stuff here (more than in commercial wads);

- first episode (up to map08) is totally linear, but in no way you can call it straightforward, no sir!

- statement above means it was really hard to track down progression in a lot of larger maps that ended being linear;

- no matter how many ways out you have (1, 2 or more), only their quality matters (and in MM it's there - all way along down to hell)

 

[RonnieJamesDiner]

This is awesome. I was really curious how the mapsets broke down in their percentages between linear and nonlinear maps (don't ask me why), and the results are sort of interesting. In case anyone else is curious --

 

Ultimate Doom

Linear: 42%

Nonlinear: 53%

Arena: 5%

 

Doom II

Linear: 59%

Nonlinear: 41%

 

TNT: Evilution

Linear: 63%

Nonlinear: 37%

 

The Plutonia Experiment

Linear: 63%

Nonlinear: 37%

 

Perdition's Gate

Linear: 69%

Nonlinear: 31%

 

Hell To Pay

Linear: 56%

Nonlinear: 44%

 

Icarus: Alien Vanguard

Linear: 47%

Nonlinear: 53%

 

Memento Mori

Linear: 44%

Nonlinear: 56%

 

Memento Mori 2

Linear: 63%

Nonlinear: 37%

 

I thought it was kind of fascinating that, right out of the gate, Doom 1 set a precedent toward a nonlinear majority (which could've set a standard), but got reversed pretty much immediately with the release of Doom II, favoring linearity (though it's worth noting that 6 of the 19 nonlinear maps from Ultimate Doom were in episode 4). And from what you've provided so far, it was only the major non-commercial releases which went back to that initial nonlinear majority.

 

I also love how TNT and Plutonia happened to be identical in their number of linear/nonlinear maps.

 

I really admire these graphs, and I've had a lot of fun thinking about this. It's also helped me to see the subject of nonlinearity in a different light. I think there's a lot more to be said about the qualities (good and bad) of a mapset than it's non/linearity. I often see a lot of "these maps are too linear" type of criticism for WADS, and while this may have its place in feedback, I don't really think it's as big of an issue -- or, as critical or relevant of an issue -- as it's made out to be. Looking, for example, at a mapset like TNT, with 20 of it's 32 maps being linear, it's a really strong set that I don't feel is hampered at all by it's linearity (in fact, many of the popularly disliked maps in TNT are from the small pool of the 12 nonlinear ones).

 

The lesson I'm taking away from this is that linearity can be easily overshadowed by the quality of level design -- from the type of enemy encounters, to the atmosphere or tone of the level, and much more beyond that. I think a really cleverly designed linear map can be just as (or perhaps even more) replayable than a nonlinear one, if the time and effort is put into it.

 

That said, it looks like the winning combination is in striking a strong balance between linear and nonlinear. Which makes sense -- breaking up repetition I think is key to making a mapset less predictable, and thus more enticing to experience the whole way through. And I think Ultimate Doom and Icarus did the most admirable job of this. Looking at both of their lists, not only are they straddling relatively close to the 50/50 margin for linear/nonlinear maps, but they also tend to distribute linear and nonlinear maps very evenly across the length of their sets, rather than having series of linear and nonlinear clumps.

 

I hope you continue this project, I'm really excited and looking forward to seeing more WADS (especially community-made ones). If you're at all taking requests, I'd love to see the Scythe series, and Eviternity.  

 

Edited October 22, 2019 by RonnieJamesDiner

[LadyVader1138]

This is just beautiful.

[Wraith777]

And.. here we are, at the next stop:

                                      Memento Mori 2

Map01: linear
Map02: linear
Map03: linear
Map04: nonlinear
Map05: linear
Map06: linear
Map07: linear
Map08: linear
Map09: linear
Map10: linear
Map11: linear
Map12: linear
Map13: linear
Map14: nonlinear
Map15: linear
Map16: linear
Map17: nonlinear
Map18: linear
Map19: nonlinear
Map20: nonlinear
Map21: linear
Map22: nonlinear
Map23: linear
Map24: nonlinear
Map25: linear
Map26: linear
Map27: nonlinear
Map28: nonlinear
Map29: nonlinear
Map30: nonlinear
Map31: nonlinear
Map32: linear

Notes:
- much more linear stuff this time;
- almost all nonlinear maps here are of collector type (comparing to other megawads - the biggest amount in one wad so far);

- a lot of maps here really hard to track down (especially map03 and map27 - those were just hell)

Edited October 24, 2019 by Wraith777
LPD corrected

[joepallai]

Another fabulous bit of research Wraith777.    I'm curious as to whether maps become more linear as the game got older and gameplay styles were solidified.  (my gut instinct is that's what happened).     

 

Any chance of doing some singles like UAC_Dead; Crossing Acheron, etc?  

 

I look forward to whatever else you do with this.  

  

[SOSU]

I dunno mate to me these diagrams look more like modern art but alright :)

[Xyzzу]

They're all trees in an abstract forest!

[vdgg]

I don't feel competent enough to absorb and analyze most of what you did, but still I would like to mention that MAP15 feels diffent to me: there should be a fork before the blue key. The window jump at the start is intended (link)

[Wraith777]

Well, emm.. Eah, it really is.. but only because author himself mentioned this. Otherwise, nothing really hinted at it being official: there is no secret flag on it, no hints like different textures or lit arrows pointing there. Also, all other windows and fences on this map are without impassible flag, otherwise no flags only on those windows could point on it being intentional...
Corrected, but anyway, that fork is still too small to affect final verdict for this map.

 

Overall, sometimes it is really difficult to make difference between such secret tricks and just exploits of the map. Usually I rely on mapping technique analysis, map reviews and written walkthroughs (by the way, this secret is not mentioned at all on Doom Wiki page of MM2 MAP15).

[Capellan]

The window route in MAP15 not a secret.  Just an option for progression. :)

 

I feel like the diagram is fine though, because even if you hop out that way, you have to go back through the interior zones in any case.  It's not like the end where in theory you only have to do one of the two corridors to the exit courtyard: you have to do every room between the start and the blue door, you just get some choice about what order you do them in (straight out the window for the BK, opening courtyard and then truly outside to the BK area, or clear out everything "inside" before hitting the BK area).

 

edit to add: what I really want to see is a diagram for Spectrum e3m4.  It probably not as complicated to draw as I think, but it's one of my more deliberately "give the player options" maps.

Edited October 25, 2019 by Capellan

[Grazza]

Funny, I referred to this window jump, albeit obliquely, in an old post.

[Wraith777]

On 10/25/2019 at 7:48 AM, Capellan said:

edit to add: what I really want to see is a diagram for Spectrum e3m4.  It probably not as complicated to draw as I think, but it's one of my more deliberately "give the player options" maps. 

 

You mean like this?

Something I'd like to note:
E3M2 - just can't run into that window to get radsuit; should I just... jump there??

E3M4 - now here we have 2 LPD variations: #1 (linearized) is easier to track down (no need to decipher all those coloured signs) and #2 (backtrack) is more accurate (no doubled parts, all map logic is shown).

E3M5, E3M6 - actually it's nice to have some steady linear action after (and before) such labyrinths as M4 and M7.

Overall, a nice classic action episode, some really big hordes there (and without a single imp), although not leaning towards slaughter (which I don't like that much); the new boss is pretty surprising. Only thing that bothers are secrets: they could use a little bit more hints (some of them are especially cryptic).

[Capellan]

1 hour ago, Wraith777 said:

You mean like this?

 

Very cool.  Totally didn't expect you to do the whole episode!

 

Is that a tiny skull and crossbones on map3?  It is actually possible to survive that fight :)

 

Rad suit on map2 - assuming it's the secret one on the computer bank, you can run through the window behind it.

 

I'm surprised map8 looks as complex as it does in the diagram.

[Gez]

How about doing other games? Like Heretic, Hacx, and Chex Quest?

 

Hexen and Strife may be more challenging to do, based on how you want to handle hubs. Either strictly map-by-map, abstracting away the need to visit other maps to progress before re-entering a map to complete it (which will usually not be "reaching the exit" but "reaching the last item/switch needed to continue progressing elsewhere") or "merging" all the maps of a hub into a megamap.

[P41R47]

7 hours ago, Gez said:

How about doing other games? Like Heretic, Hacx, and Chex Quest?

 

Hexen and Strife may be more challenging to do, based on how you want to handle hubs. Either strictly map-by-map, abstracting away the need to visit other maps to progress before re-entering a map to complete it (which will usually not be "reaching the exit" but "reaching the last item/switch needed to continue progressing elsewhere") or "merging" all the maps of a hub into a megamap.

I thinks that, aside from being different areas, hubs are part of a very long and interconected map in respect to progression. That way, i think they can be interpreted for this kind of diagrams. So, for example, Hexen first Hub, at first just a few doors are open, so linear, then it opens different parts of the differents areas, so non-linear at this point, but in the end, they all need to be solved to reach the end, so maybe linear with branching tackled areas? Or something between collector and key independent? as there are many optional areas, in later levels at least.
 If not, its grid or sandbox, no other choice for sure.
Anyway, thanks Wraith for this amazing analitical work. It surely take time

[Wraith777]

On 12/4/2019 at 12:38 PM, Gez said:

How about doing other games? Like Heretic, Hacx, and Chex Quest?

 

Hexen and Strife may be more challenging to do, based on how you want to handle hubs. Either strictly map-by-map, abstracting away the need to visit other maps to progress before re-entering a map to complete it (which will usually not be "reaching the exit" but "reaching the last item/switch needed to continue progressing elsewhere") or "merging" all the maps of a hub into a megamap.

 

Hexen and Strife will definitely be challenging to classify within usual LPD format. I may need... some help there actually...

Meanwhile - here is

                                             Requiem

Map01: linear
Map02: linear
Map03: linear
Map04: linear
Map05: linear
Map06: linear
Map07: nonlinear
Map08: linear
Map09: nonlinear
Map10: linear
Map11: linear
Map12: nonlinear
Map13: linear
Map14: linear
Map15: nonlinear
Map16: linear
Map17: nonlinear
Map18: linear
Map19: nonlinear
Map20: nonlinear
Map21: linear
Map22: nonlinear
Map23: linear
Map24: linear
Map25: linear
Map26: linear
Map27: nonlinear
Map28: linear
Map29: nonlinear
Map30: linear
Map31: linear
Map32: linear

 

- Even more linear maps

- Lot of conceptual maps (which tend to be linear)

- About map23 - this one rose a serious question for me: could a sandbox map actually count as linear? The white brick city here seems to be a sandbox (and it actually is), but in the end you still need to enter all buildings in a strict order. So the outer area is reduced to simply being a big room.

Also, a map size seriously matters when level progress is defined: just reduce the size of outer area and it becomes clear, that it's one single non divided space with only one way out.

This needs to be developed further, but overall, it seems that large maps which have all attributes of grid or sandbox on bigger scale may end up being still linear.

Edited December 12, 2019 by Wraith777

[SOSU]

Aww man a classic megawad :D