Point System Discussion

[WingsWithoutBird]

It's boring if you go slow and on full wallpogo maps, demo maps aren't just wallpogo as someone already said. And yes, some wallpogo is hard you know it is (im not talking of mireals which then I agree just boring shit), im talking more like more techinque wallpogos or when you go fast (or at least try to).
Second of all, I actually have completed non wallpogo t6's and also have non wallpogo map tt's, please make sure to check things before saying it here, ty (even tho they are the majority, but I never hided that).
You also have bad soldier map tts, go get them, easy points no? Just because tt's are bad doesnt mean they are easy to get (surely easier than overgrinded maps, yes).
At last, I define wallpogo maps maps like, jump_storm, mireals, desa, etc I dont think you can qualify maps like jump_cake, daylight, ugly, hardware, speed, etc that actually have only 1/2 wallpogo jumps and they are still demo maps.
As always, my opinion and my thoughts, ty

PS: lets get things straight here, my complains arent about maps that, I consider, as wallpogo maps (and if you see I ran them few times for easy tt), but more about maps like ugly, cake, hardware, maps that have few wallpogo and that I enjoy playing (for some reason).

edit: I meant 1 or 2 with 1/2 and not half

[VAVLIE]

You're generalizing maps that use wallpogo as wallpogo maps, yet not doing quite the same generalization to syncs (and other techniques), how many maps can you think of that have a vanilla quad/quint? How many include syncs in most of their jumps?

Pretty sure I treated them both the same way: maps that mostly revolve around syncs vs maps that mostly revolve around wallpogo. Do the exercise, go on https://tempus.xyz/maps and count the number of maps that mostly have sync levels. Then do the same for maps that mostly have wallpogo levels. The numbers just don't match in the most ridiculous way. Do keep in mind that I'm not using this to say "death to wallpogo", heck I'm not even sure I understand what people are arguing about here, although I'm seeing from some jumpers some insightful remarks on the situation. I'm only pointing it out to show that saying "well there are maps that are mostly about bounces" is in my opinion a weak argument if the purpose is to be compared to wallpogo, and I think it was well understood. Cheers.

[yaboi]

I notice there are many conflicting views regarding how points should be allocated to demo maps that mainly consist of boring wallpogo for soldier. As a demo main, I don't care what the immediate solution is, but I do have suggestions for the future (although this would require the co-operation of mappers).

To anyone who intends to make a demo map, PLEASE try to tweak the map so every level isn't just wallpogo for soldier.
There are many ways to go about this:

• Remove some or all of the teles for soldier
• Add platforms that only soldier can stand on or make them roller for demo
• Create alternate pathways that only soldier can access
• etc.

Some maps already implement these.
[jump_compound - level 10]
The vert is impossible as soldier so there is a separate path that is no-det for demo. Why not implement this even when a jump IS possible, that way lessening the amount of wallpogo.
[jump_arugula]
There are small platforms scattered across the map that soldiers can use but demos can't. Why don't map makers use these more often?
[jump_neptune]
No teles at all but more fun to speedrun than a map like jump_mireal. You might not like the fact that it yields less points, but after the changes to the point system all maps will give the same WR/tt points regardless of tier so that won't matter. There are so many demo maps that would be so fun to run as soldier if some or all teles were removed, for instance jump_shiver and jump_mohr.

jump_fastnfar is a good example. The map is made for soldier but the creator intentionally made it airpogo-only for demo. Why don't we do the reverse? Imagine modifying some of the levels in jump_storm so that there is a class-restricted tele for demo, but soldiers can go below the tele to maybe perform back-to-back syncs. This obviously requires more effort, but imagine how refreshing it would be to encounter a non-wallpogo jump. If we're dealing with a straight-line airpogo level, it could be as simple as adding 40-50 skip platforms. Even though this is somewhat lame, it's still better than monotonous wallpogo. Imagine changing lv.4 of jump_mireal3 so that the platform at the bottom could be stood on by soldiers. This changes the level from wallpogo to a sync into wallshots. Much less boring. Look at the last level of jump_carrot. What if soldiers could slide down the slope, hit an auto-eb or synced rampshot, and THEN go into wallpogo? Much more interesting.

It's obviously too late to make such changes but it's something to note for future mappers.

[tyjle]

technically it's better design to make your jumps based around both classes, but I don't think someone making a map for a specific class should have to design it with the other in mind. I've only made soldier maps and I never once think about how a demoman would jump on the map, and I know plenty others have done the same.

imo the completion points from a tedious wallpogo map are probably deserved. Anyone who doesn't have the time or patience to beat a map like that should stop whining about their virtual points

[five]

Heya, I'm new >_< bit out of place/late now but wanted to reply to the original topic

I got a 399th/399 demo completion on a soldier map today & must confess I thought to myself "well thank fuck I can forget about this map now"
It'd be nice to have more incentive to improve my own time, even if it means less of a reward for my snail runs.

Changing to the next map in the rotation & seeing so many of the same names in the top times, so far from my own timer is intimidating as a newbie
because I feel like my points/rank progression will plateau once I get completions on all the accessible maps

I think players could be awarded completion points as a percentage based on the speed of their run relative to other jumpers.

For example atm if I get a 4000th/4000 t3 completion with 2h51m I would receive the same points as jumpers that made the effort to rerun the map enough times to get a neat 8min ~800th/4000.
Lets say this would have been 50 pts for both of us.

If points were rewarded as a rounded percentage then I would receive 1% of the completion points (0.5 pts) for my pitiful showing & a player at 800th/4000 would recieve 80% (40 pts)

We could then have a map specific completion goal time (lets say 6m30 for this example) below which the full 100% of points are awarded.
You could also soft-nerf wallpogoing demomaps here by not specifying a soldier time for demo maps. I personally love wallpogo but understand the aversion to 'grind & forget' approaches being up next to consistent record breakers in the leaderboards

Percentile tiers (5%/10%/15% etc) could be used to avoid awarding fractions of points, or maybe today's 10 points could become 100.

Profile-wise a completion could still be considered a completion, preserving nice %completions but pulling back the points a bit to be more indicative of speed/consistency?

Probs making a right muppet of myself but I mean well =(

At any rate I'm a big fan of tempus in its current state & very grateful you guys exist mm! sorry for wall of text

[tyjle]

Heya, I'm new >_< bit out of place/late now but wanted to reply to the original topic

I got a 399th/399 demo completion on a soldier map today & must confess I thought to myself "well thank fuck I can forget about this map now"
It'd be nice to have more incentive to improve my own time, even if it means less of a reward for my snail runs.

Changing to the next map in the rotation & seeing so many of the same names in the top times, so far from my own timer is intimidating as a newbie
because I feel like my points/rank progression will plateau once I get completions on all the accessible maps

I think players could be awarded completion points as a percentage based on the speed of their run relative to other jumpers.

For example atm if I get a 4000th/4000 t3 completion with 2h51m I would receive the same points as jumpers that made the effort to rerun the map enough times to get a neat 8min ~800th/4000.
Lets say this would have been 50 pts for both of us.

If points were rewarded as a rounded percentage then I would receive 1% of the completion points (0.5 pts) for my pitiful showing & a player at 800th/4000 would recieve 80% (40 pts)

We could then have a map specific completion goal time (lets say 6m30 for this example) below which the full 100% of points are awarded.
You could also soft-nerf wallpogoing demomaps here by not specifying a soldier time for demo maps. I personally love wallpogo but understand the aversion to 'grind & forget' approaches being up next to consistent record breakers in the leaderboards

Percentile tiers (5%/10%/15% etc) could be used to avoid awarding fractions of points, or maybe today's 10 points could become 100.

Profile-wise a completion could still be considered a completion, preserving nice %completions but pulling back the points a bit to be more indicative of speed/consistency?

Probs making a right muppet of myself but I mean well =(

At any rate I'm a big fan of tempus in its current state & very grateful you guys exist mm! sorry for wall of text

I think that's a pretty neat idea, but I think that would give a lot less incentive for new jumpers to complete maps because they know that they're not going to get anything substantial for it.

[Oatmeal]

Doesn't the surf equivalent of Tempus do that? You get points for being in a certain percentile of map completions. You get loads of points if you are in the hightest 5th percentile on an easy map for example, which gives people an incentive to run easy maps even though there are thousands of completions on that map already and its rather unlikely that you will get into the top ten without grinding forever.

[Superchuck]

Doesn't the surf equivalent of Tempus do that? You get points for being in a certain percentile of map completions. You get loads of points if you are in the hightest 5th percentile on an easy map for example, which gives people an incentive to run easy maps even though there are thousands of completions on that map already and its rather unlikely that you will get into the top ten without grinding forever.

Yes, and the more completions a map has, the more points you get from being in the top percentiles.

[WingsWithoutBird]

Didn't know surf system was that good, that seems pretty decent.
Where can I find more information about how it works?

[WingsWithoutBird]

Merged to consolidate threads

This has already tons of topics but they are all way too messy and because I think this would solve the completion problem I wanted to make sure that everyone reads this.

(This was mentioned by someone, somewhere on some topic)

Someone mentioned this system where you would get more points depending on how close you are to wr. I kept thinking on how to do this but the only thing I came up with was a sytem where the given points for completion would depend on the location of your time between the wr and the last time.

It would still give more points to wr and to tts separetly as it is now.

I made a simple one, just super basic math and basic programming (just to get things done and test)
 http://prntscr.com/jxurno

It basically needs the wr time and the slowest time, then it creates a simple function where it gets the points for a given time.

Lets take a few examples and take a look at the results:

Spoiler (click to show/hide)

map: jump_rush
tier: 1 (10 completions points)

#1 - Boshy - 47.32 - 10 points
#2 - ungato - 47.59 - 9.99932 points
#3 - soup - 48.28 - 9.99932 points
.
#10 - guyyst - 58.02 - 9.97301 points
.
#5808 - Mouse* - 1:19:15.79  - 1 point


map: jump_soar
tier: 5 (100 completions points)

#1 - kaptain - 4:24.22 - 100 points
#2 - ungato - 5:04.89 - 99.8004 points
#3 - brockington - 5:06.52 - 99.7924 points
.
#10 - z1- 5:45.11 - 99.6031 points
.
#260 - Wÿm - 5:40:38.75 - 1 point 

map: jump_homie_beta
tier: 6 (200 completions points)

#1 - Boshy - 16:00.00 [perfection]- 200 points
#2 - Larry - 18:53.08 - 195.981 points
#3 - Scoot - 52:06.13 - 149.705 points
.
#10 - kaptain - 2:38:50.63 - 1 point

Of course when new wr is set, it needs to start using the new wr as reference and same thing for last time.
With this system points would be changing on epic world records.

here is the simple code I made - http://prntscr.com/jxv8me
formula here - http://prntscr.com/jxv9dt

If I missed something, please tell me and I try to fix.

Wings

Spoiler (click to show/hide)

I should be studying.

[MrHappyCamper]

Merged to consolidate threads

Spoiler (click to show/hide)

This has already tons of topics but they are all way too messy and because I think this would solve the completion problem I wanted to make sure that everyone reads this.

(This was mentioned by someone, somewhere on some topic)

Someone mentioned this system where you would get more points depending on how close you are to wr. I kept thinking on how to do this but the only thing I came up with was a sytem where the given points for completion would depend on the location of your time between the wr and the last time.

It would still give more points to wr and to tts separetly as it is now.

I made a simple one, just super basic math and basic programming (just to get things done and test)
 http://prntscr.com/jxurno

It basically needs the wr time and the slowest time, then it creates a simple function where it gets the points for a given time.

Lets take a few examples and take a look at the results:

Spoiler (click to show/hide)

map: jump_rush
tier: 1 (10 completions points)

#1 - Boshy - 47.32 - 10 points
#2 - ungato - 47.59 - 9.99932 points
#3 - soup - 48.28 - 9.99932 points
.
#10 - guyyst - 58.02 - 9.97301 points
.
#5808 - Mouse* - 1:19:15.79  - 1 point


map: jump_soar
tier: 5 (100 completions points)

#1 - kaptain - 4:24.22 - 100 points
#2 - ungato - 5:04.89 - 99.8004 points
#3 - brockington - 5:06.52 - 99.7924 points
.
#10 - z1- 5:45.11 - 99.6031 points
.
#260 - Wÿm - 5:40:38.75 - 1 point 

map: jump_homie_beta
tier: 6 (200 completions points)

#1 - Boshy - 16:00.00 [perfection]- 200 points
#2 - Larry - 18:53.08 - 195.981 points
#3 - Scoot - 52:06.13 - 149.705 points
.
#10 - kaptain - 2:38:50.63 - 1 point

Of course when new wr is set, it needs to start using the new wr as reference and same thing for last time.
With this system points would be changing on epic world records.

here is the simple code I made - http://prntscr.com/jxv8me
formula here - http://prntscr.com/jxv9dt

If I missed something, please tell me and I try to fix.

Wings

Spoiler (click to show/hide)

I should be studying.

I don't mean to burst your bubble, but completion points is static by definition, if you are talking about points that change with run time, that's tt/whatever other points, but that's just semantics. And also your equation has a divide by zero if there is only 1 run so you would need to add a little exception into your code.

Your equation is alright, it does what the plot shows, I simplified it so it's easier to understand:

It makes sure to give the last time at least 1 point, but the issue I see with this is that the amount of points you get depends on the last time, so if someone were to do a really long time on the map (as L goes to infinity) then the amount of points you get increases (the fraction approaches 1). This issue is fundamental to a system that takes into account how close your time is to wr time since it has to make that calculation relative to something, that something being the average or slowest time which can be skewed by very slow times. So I do not think it can work. Instead, using percentiles is the way to go without taking into account the actual time.

[MrHappyCamper]

Nvm I'm an idiot you can just make a nonlinear equation without taking into account the last time:

which will give you the distribution:

I kind of like this method, it gives you large increases in points when you are close to wr, like tt does, and if you want to make it even more of a spike you can just square your time.

[yaboi]

That formula wouldn't work and even if it were adjusted I don't think it's a good measurement for points. Taking into account the slowest time is a bad idea for the reasons MrHappyCamper stated. Using a normal distribution or comparing your time to a weighted average could work though. A really slow last place time won't skew calculations in this case.

[Starkie]

under this system you get punished for running a map with a really good WR, doesn't seem to make much sense to me

[Ladyboog]

this thread is epic