Point Pressing Pop Quiz #4

[xelnia]

Good stuff Dean. I had planned on answering the spring section of your question with frame-perfect answers, but I never quite finished. Anyway, I have a couple of comments that you and others might find interesting. I've been collecting information and making graphics in the hope that someday I can start a comprehensive "spring theory" thread in the Advanced Strategy forum...but I need to be an advanced player first. 

1) I don't know the location of every spring's ladder clearance (something I plan on looking at now), but I DO know that there is a "safe zone" near the top of Pauline's ladder where a spring will NEVER hit you regarding of length or scrunchiness (which I'm not sure matters). If you look at Jumpman's Y-position on the ladder in the debugger you'll see that Kong's platform has a hex value of 50. The top of Pauline's ladder is 30. I'm not sure why going up counts down, but that's the way it is. Once you reach a Y-value of 38 you're safe. So, 50 minus 38 converted to decimal is 24 pixels of danger (and 8 pixels of safety) on a ladder that is 32 pixels tall. Jumpman doesn't climb ladders one pixel at a time however; I believe it's 2 pixels for every 5 frames.

2) Each of the two safe spots on Kong's platform are 13 pixels wide. The first safe spot at the top of the short ladder occupies X-values A7-B4 (hex again). The center of the short ladder is at B3. The second safe spot occupies X-values 75-82. These safe spots are obviously the areas where will no spring will every hit you and, as you pointed out, can shift depending on length of each spring. The safe area for leeching the right-side fireball has X-values of D5-D9 (4 pixels!). At DA you can hit the fireball while jumping, and DB is falling off the right edge. The left edge (D5) can obviously shift depending on spring length, which Dean often demonstrates by edging to the left to leech a spring.

Also, the 13-pixel-wide safe spots leads to me believe that the horizontal dimension of a spring's hitbox might be 6 pixels.

3) The center of Pauline's ladder has an X-value of 93. However, as Dean pointed out in a previous post, you can "corner" up the ladder by grabbing it before you reach that center position. You can grab the LEFT side of the ladder at X-position 90, and the RIGHT side of the ladder at X-position 97. In regards to D2K, you can actually stand at X-position 90 on L14 springs for the duration of the screen, however you can only make it to Y-position 3A instead of the safe 38. The L14 spring on D2K is the second longest of the 16 springs, or #14 in Dean's 0-15 system.

4) Each of the 16 springs lands one pixel farther than the one previous. So, a 0 spring lands at X-value 28, a 1 spring at 29, etc. till you get to the 16th spring which lands at X-value 37 (again hex values here). Every spring moves horizontally 2 pixels for every frame of animation. When a spring lands on Kong's platform it actually remains on the platform for 2 frames, and slides to right 2 pixels in the second frame. This means that's possible to "see" a 0 spring as 2 spring, or 13 spring as 15 spring, etc. depending on which frame you're actually "seeing" the spring.

5) I'm not entirely convinced that scrunchiness matters, but I've seen some weird values in the MAME debugger that I need to look at more closely. A spring will always "land" at Y-value 50...that's Kong's platform. It will stay at 50 for two frames while it's X-position shifts. In it's animation, the next Y-value on the upswing of it's arc will always be 4D. However, I've seen springs that visually land on the platform but have a Y-value of 4D, followed one frame later by Y-value of 50, with an X-value shift of 4 pixels between the two frames instead of 2....with both springs appearing to sit directly on the platform.

Lots of stuff to work out. 

[up2ng]

Wow, excellent post!  This is great information which should probably be moved or copied into its own thread for reference.

I'm not totally sure if "scrunchiness" comes into play on this stuff either, but my instinct is that it actually does.  One easy thing that someone could try is to simply frame advance an elevator screen in MAME and watch how the springs move in super slow motion (I feel like I've done this at some point but don't remember what I found).  My guess is that the scrunchiness of a spring does not oscillate at the same rate as they bounce, so you can have some funny looking springs in the air -- but the key is that perhaps some of them are "closed" and some of them are "open" when crossing Pauline's ladder -- and IF closed and open springs actually have different vertical hitbox dimensions (a big IF), this could become a small variable in whether or not Jumpman makes it up the ladder.

One example -- in the past, I've used a save state to stand in the spot exactly where you are describing on the D2K kill screen, so that all that is required is to push UP to climb the ladder.  With a save state, obviously all enemy behavior is repeatable.  I would climb at the first available spring over and over again and would get hit in the leg in the same spot over and over.  BUT, I found a specific spring (say, the 4th spring into the save state for example) where I could make it one pixel further -- Jumpman's animation was different when hit and the spring appeared to just barely hit Jumpman in the right foot, making it seem like he ALMOST makes it -- but this only was possible on that specific spring, and as you know in D2K they all have the same "length".  The only thing that I felt might make a difference in that situation was that this specific spring had a more favorable scrunchiness when crossing the ladder.

Anyways, it would be interesting if that aspect of it were solved.  But, great info on the rest of it! 

[xelnia]

Dean, I'll try to put together all my spring info into a readable format and make another forum post about it.

I imagine the DK code has something to say about sprite hitbox dimensions. After all, it's always checking for sprite collision, right? So I imagine it must check that each sprite is at whatever x,y value and is compares that against some specific pixel offset from that x,y value. Closed springs are definitely wider than open springs, and open springs are definitely taller than closed springs...visually, at least.

[Bliss1083]

I've been trying to perfect my right of the ladder climb to eliminate wasting time going to the yellow safe zone I noticed that two short springs having the second spring around half way on the yellow platform to continue up is safest. I finish around 6800 with top shelf and no prize collecting. I also get around 6500 with a top prize usually. If you get up soon enough it seems to have more short springs then around 5500 left the longer springs kick in more. Everything has it's own pattern though so if I see long springs from the begining I go for two prizes and if short springs are happening more I go for top prize and a right of the elevator climb with 6500 left usually. 7300k has been my best but more can be possible for the right person. This only goes for levels 4 and up. Keep your regular pressing for the level 2 and 3.

[Bliss1083]

Ok I think if the first spring is short enough I could successfully climb up the ladder regardless of the next springs landing point. I just did it on a really long spring and a short one.

[up2ng]

Hmm, if you're talking about Level 4+ this makes no sense.  You are farther away from the ladder and you get a later jump when approaching from the right than from the left and when you approach from the left there are springs which cannot be passed (the longest ones).  There are likely even more springs that cannot be passed from the right side.

[Bliss1083]

I'll try and make a video. The really short springs land right at the foot of the ladder allowing for a really good jump on the spring and had a really long spring not hit me. Maybe lucky but two really short springs seem to kill me more but still can make it. I've finished an elevator today at 6700 on the ticker and grabbed the top left prize for 7500k finish. I haven't mastered it but am getting pretty comfortable and will stream or record it tonight if I have time.

[marinomitch13]

Yeah, this the sort of thing I pointed out right away, but I'm not sure it works of any spring following the shortest one, which is why I only made an example of the '0, 2' case here:

I'm not 100% sure, but doesn't the '0, 2' order of the springs actually allow you to approach and climb the ladder from the right!? I know a '0,0' progression certainly would, but I'm not sure about the '2'.

Also, if I remember correctly, this facet was un-utilized in Dean's eventual solution for this case either.

[up2ng]

I'd definitely be interested in seeing a video of that example.

But, using our nomenclature (little did I know that it would become "standardized" when I was writing it, but it helps that the code is now available and it's confirmed that there are indeed 16 different spring locations) if you were approaching from the right on a 0, 15 combination I'm 100% sure that you'd have less of a chance of survival than if you were to approach from the left on a 15, 15 combination.  Remember with the 15, 15 combination, you can literally stand ON the ladder and do nothing but press UP immediately after the longest spring goes over your head and you STILL can't make it past a 14 or 15.  You definitely cannot get a BETTER jump than that by approaching from the right no matter how short the spring is.

But, anyways, the point is still well taken that there are likely a larger number of spring combinations that can be safely approached from the right than most people think and that it might be worthwhile to look into adding this into a point presser's arsenal.

Think about how efficient this case would be -- you see a very short spring and begin approaching from the right with the intention of climbing from the right.  You might even get a foot on the bottom rung and you'd really only need to retreat if there was a very long spring coming -- this retreat only needs to be about a half-step away from the ladder and you could immediately turn and climb up again from the LEFT -- basically you'd NEVER have to run all the way to the yellow portion of the "safe spot"!  This would take a lot of practice but I could see it potentially gaining an average of 200 - 300 points per level.

[marinomitch13]

Think about how efficient this case would be -- you see a very short spring and begin approaching from the right with the intention of climbing from the right.  You might even get a foot on the bottom rung and you'd really only need to retreat if there was a very long spring coming -- this retreat only needs to be about a half-step away from the ladder and you could immediately turn and climb up again from the LEFT -- basically you'd NEVER have to run all the way to the yellow portion of the "safe spot"!  This would take a lot of practice but I could see it potentially gaining an average of 200 - 300 points per level.

This is one of the things I've had in the back of my mind for a while to have you try to incorporate into you game. You seem to be maybe the only person whose game would really require that sort of effort to completely overhaul your normal Springs gameplay to get a few extra 100 points. 

"Because for you to get a single point further in an old game, from the early eighties, meant a tremendous increase of skill!" -Walter Day

[Bliss1083]

Yeah Mitch, dean would be the best at this. I will stream a video Monday night when I get some time. I would say I had a short 0 spring and practically just climbed straight up the ladder. After I'd say the next spring was solid on the yellow platform. I wouldn't have believed it either if I didn't do it. Might not been the longest spring but was def. Really long.

[marinomitch13]

About 1.5 years I messed around with this and had a couple 400k+ games using only this method, so I'm quite familiar with it. I just stopped using it, since I figured it was ultimately safer and not any significant loss of going from the left given my pace at the time.

[Bliss1083]

Yeah. I think that you can definitely gain some points on it immediately and if it's not there then go back to the yellow platform for the retreat and set up for the classic climb. Makes sense since you're waiting on a shorter spring to get to the safe zone anyways. Might as well try and climb and gain what you can point wise.