Author Topic: Point Pressing Pop Quiz #2  (Read 7556 times)

0 Members and 1 Guest are viewing this topic.

Offline up2ng

  • Senior Member
  • *
  • Posts: 242
    • Awards
Point Pressing Pop Quiz #2
« on: July 04, 2013, 05:21:56 pm »
You are standing just to the right of the middle ladder on the 4th platform of a Level 5 rivet screen.  You have the top hammer in hand with 2700 remaining on the Bonus Timer.  You had grabbed the top hammer with 3000 remaining on the Bonus Timer and immediately smashed two firefoxes on your right.  They have now both respawned -- the first one respawned on the 3rd platform on the left side and is now climbing up to the 4th platform.  The second one just respawned on the 2nd platform on the left side.  Two firefoxes are both frozen on the top platform on the left side and the last firefox is bouncing around on the 4th platform on the left side.  The only remaining rivets which have not yet been unplugged are the top two rivets on the right side.  Because of how the board was played, only the Umbrella was grabbed -- both the Hat and the Purse remain on the screen.  Will you...

A) Run to the right, unplug the rivet, wait for your hammer to expire, then jump back across and climb up the inner ladder and "taunt" Kong for as long as possible before completing the screen?

B) Run to the right, unplug the rivet, wait for your hammer to expire, then quickly climb down to retrieve the Hat and climb back up, jump back across and climb up the inner ladder and "taunt" Kong for as long as possible before completing the screen?

C) Run to the right, unplug the rivet, wait for your hammer to expire, then quickly climb down to retrieve both the Hat and the Purse, climb up the middle section and "taunt" Kong for as long as possible before completing the screen?

D) Run to the left, stick your hammer just slightly over the gap and attempt to "Poach" ONE firefox before quickly running back to the right side of the screen (hopefully before this firefox respawns on the right side OR this firefox respawns but you are able to smash it again OR this firefox respawns on the bottom AND never comes into play), with the plan of [CHOOSE ONE]: finishing with plan A OR B OR C?

E) Run to the left as far as possible, try to smash as many firefoxes as possible before reversing back to the right, hoping to smash any and all respawners, with the plan of [CHOOSE ONE]: finishing with plan A OR B OR C?

F) Run to the right, and attempt to unplug the rivet by standing on it and reversing directions, with the plan of [CHOOSE ONE]: finishing with plan A OR B OR C OR D OR E?

G) Use this opportunity to take a much needed bathroom break?

Explain.

[BONUS] Does your answer change if there were 4700 remaining on the Bonus Timer?  3700?  1700?  700?  Why or why not?
Donkey Kong:  1,206,800  Kill Screen
Donkey Kong:  898,600     16-5
D2K:                 380,200     L=9
Donkey Kong Junior:  In Progress
Member for 11 Years DK 1.2M Point Scorer Wildcard Rematch Champion Blogger Former DK Level 1-1 World Record Holder Former DK No-Hammer World Record Holder DK 1.1M Point Scorer Individual Board Record Holder DK 1M Point Scorer Former DK World Record Holder - MAME DK Killscreener Twitch Streamer

Offline f_symbols

  • Spring Jumper
  • *
  • Posts: 833
  • wht u mean
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #1 on: July 04, 2013, 06:00:21 pm »
This is another great question, highly analytically.  Thanks for doing these.

I'd go with the D/B option due to the timer. Poach that one guy on the 4th girder and go get the hat, then immediately get to the top; the lack of bonus timer points means leaching wont be that profitable anyway (1000pts gained for 2500 bonus). 

That being said, if the bonus timer was greater my choice would be B, id go for the purse, and hump kongs foot after the single poach.   The payoff for humpery is high when the bonus timer is greater than 4000.

Had the timer been shorter, the choice would be E/A, I'd hellishly unleash my hammer on the left side in hopes of killing lots of firebeasts. If more then 1 of those trolls spawns on the right its time to head over with the last of the hammer; you don't want to have to clean those last 2 rivets with the potential death or slowdowns of 2 fireballs.

If this were level 1 or 2, I would disregard paulines items; ain't no time for 300pt or 500pt items that cause a net loss in their acquisition.  Also, due to the slower bonus timer of levels 1 and 2, the preservation/creation of situations that lead to excessive monkey humping should be cultivated whenever possible :D
« Last Edit: July 04, 2013, 06:08:48 pm by f_symbols »
Member for 11 Years Former DK Remix World Record Holder DK Remix Killscreener DK 1M Point Scorer Blogger DK Killscreener Twitch Streamer

Offline ChrisP

  • Spring Jumper
  • *
  • Posts: 1763
  • I'm going to jump next to your leg.
    • Donkey Blog
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #2 on: July 04, 2013, 07:31:20 pm »
Just as in life.
http://donkeykongblog.blogspot.com

4 Quarters :-* - 800K Avg. Per Qtr. :o - No Restarts 8) - No Proof :'(

7/26/2013   Coin 35,946   710,800   18-1
7/28/2013   Coin 35,947   903,700   22-1
8/16/2013   Coin 35,948   694,100   17-6
8/17/2013   Coin 35,949   893,100   22-1

3,201,700: the $1 World Record?
Member for 11 Years DK Masters - Rank D DKJR Killscreener Blogger Twitch Streamer DK Killscreener CK Killscreener

Offline Milehighdt

  • Senior Member
  • *
  • Posts: 243
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #3 on: July 04, 2013, 09:27:43 pm »
I think the gain of grabbing the lower prizes is less than the gain from taunting kong longer so A is better than B or C.  If thats true(fairly certain it is) then F-A is better than A since you get an extra timer tick from not having to jump the gap. If your just waiting for the hammer to end to do F-A then you should try for an extra smash. With 6 timer ticks left you have time to pause slightly on the left side and still have the hammer for the full trip back so a conservative E is better than D. My answer is E with a F-A finish. G ? you need to wait for a level you can walk/run away with some expectation of jumpman living through your departure.

If the timer was at 700 I would play F-A so as to not have to deal with a spawn, climb firefox that might make completing the level imposable. 
DK (Arcade) 1,129,700
DK No-hammer (Arcade) 768800 Killscreen
DK No-hammer (mame japan set)  524600 level 21-4
Member for 11 Years DKJR Killscreener Individual Board Record Holder DK No-Hammer Killscreener Former CK World Record Holder - MAME DK 1.1M Point Scorer DK 1M Point Scorer Former DK No-Hammer World Record Holder DK Killscreener Twitch Streamer CK Killscreener

Offline stella_blue

  • High Score Moderator
  • Spring Jumper
  • *
  • Posts: 1529
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #4 on: July 09, 2013, 06:16:32 am »

I've attempted to create a save state to match Dean's specifications as closely as possible.  The general idea is that a readily available "study aid" might encourage more active participation in this thread.

I'm pretty happy with how this turned out.  It's not exact, but it's very close.  Here are the details:
  • The current stage is the Level 5-6 rivets
  • You are standing on the 4th platform, to the right of the middle ladder, facing left
  • You grabbed the top hammer with 3000 remaining on the bonus timer
  • The bonus timer has just ticked down to 2700
  • The top hammer will expire less than 1/10 of a second after the timer reaches 2100
  • All 5 fireballs are trapped on the left side of the structure (2 on the top platform, and 1 each on the 2nd, 3rd, and 4th platforms)
  • The 2 top right rivets are the only ones not yet cleared
  • The umbrella has been taken, but the hat and purse remain to be collected

Here is a screen capture (75% of original size), just as the save state kicks in:




What isn't obvious from this screenshot are the 2 differences between the save state and Dean's original description.  Both concern fireball behavior:
  • Only 1 of the 2 fireballs on the top platform is frozen
  • The fireball on the 3rd platform is the other freezer, with little intention of climbing to the 4th platform
Additionally, I set this up with a beginning score of 172,000 points.  I wanted an even multiple of 1000, to make it easier for players to evaluate the alternate methods of completing the stage.

Good luck!

DK High Score 
1,064,500
  (L22-1)
DK Level 1-1 
13,400
DK No Hammer 
535,300
  (L15-2)
Member for 11 Years Most Time Online DK 1M Point Scorer DK Killscreener Blogger Twitch Streamer

Offline mikegmi2

  • Elite Member
  • *
  • Posts: 431
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #5 on: July 09, 2013, 06:55:47 am »
I would go with E/A for max points.

Bonus - Doesn't matter because you are chosing E with hopes of getting 500/800 smashes on any/all of the firefoxes on the left side, and also resmashing for 500s/800s after they respawn on the right side.  I suppose you would have to net a profit of at least 600 points from the firefox smashes to beat out simply taunting Kong for points for the remaining 2200 or so ticks of the bonus timer (after hammer runs out and you climp up).

Taking max risk for max reward, E/A would be my answer.
Donkey Kong - 1,076,000 (arcade, KS)
Member for 11 Years DK 1M Point Scorer DK Killscreener Twitch Streamer

Offline up2ng

  • Senior Member
  • *
  • Posts: 242
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #6 on: July 15, 2013, 02:07:49 pm »
Ok, I thought we'd have more interest in this one but I'll go ahead and post an answer now.  The answer will likely be more vague this time around.

First, option G is probably the worst option.  You have less than a 30 second break real time in this situation so you'd have to have a VERY efficient bathroom break to make it back before dying.  However, if NOT taking a bathroom break will almost definitely result in a premature game over, this becomes a legitimate choice.

If we eliminate choice G, we have two main options -- smash firefoxes or don't smash firefoxes, and then for each option we decide precisely HOW we will go about doing so with the most efficiency.

Based on my quick save state analysis, the choices between A, B, C and F are all extremely close!  It turns out that choices A and B are pretty much identical!  So, you can go and get the hat, or not, and it won't cost you any points either way!

Believe it or not, choice C -- getting BOTH prizes -- only loses 100 points as compared to choices A and B!  This is much closer than most people think!  But still an inferior choice in this situation.

(Interestingly, if the situation was slightly different and there was a firefox directly above you "guarding" the final rivet in the top right corner, it "might" be profitable in the long run to run down and grab one or both prizes (depending on how the Firefox responds) before climbing back up for the final rivet.  Any time you can do SOMETHING rather than NOTHING in a situation where you are currently being delayed, it's often the better choice.  It's not a clearly better choice in this case, but something to think about anyway.)

That brings us to choice F.  Choice F (followed by choice A) is actually better than choice A by about 100 points IF you are able to execute unplugging the rivet successfully before the hammer expires.  With this much time remaining in your hammer cycle, I think it's worth trying -- just be mindful of whether or not you have unplugged it by running too far, in which case you'll have to be careful not to turn around and fall down the hole.

With choice F followed by choice A, you can gain about 2800 points from leeching Kong.  This is based on the hammer running out at 2100, then immediately climbing the ladder and beginning to taunt Kong when the timer is just ticking down from 2000 to 1900.  At a rate of 1000 points per 800 Bonus time, you can gain about 2400 points here plus another 400 points at 000, for a total of 2800 points.

So, the question now is, will you gain MORE points than this by smashing Firefoxes?  The answer now becomes a bit fuzzy and vague because we will now have to derive our answer on long term expectation for points gained based on probabilities and results which can only be loosely estimated.  For example, each Firefox smashed has an expected value of about 429 points based on current knowledge, but COULD be as low as 300 or as high as 800 points.  If we choose option E, how many firefoxes are we LIKELY to smash (and with what sort of distribution)?

Keep in mind, in the given situation we still have about 2/3 of our hammer cycle remaining so there should be plenty of time to get multiple smashes if we choose.  In general, I think the more time you have remaining in your hammer cycle, the more willing you should be to try to go get extra smashes -- especially if you have enough time to reach across the gap for smashes AND run all the way back to the right side WITH the hammer in hand to try to smash the respawners.  (It takes approximately 2 Bonus seconds to run from a position where you are reaching across one gap to the same position on the opposite side of the screen where you are reaching across the other gap.)

Another factor to keep in mind is the Timer value itself which translates directly into a number of taunting points if all firefoxes are trapped.  In general, I think that the greater the Timer, the LESS willing you should be to try to go get extra smashes.  This is because if you are unable to smash or trap the respawners you are likely going to have to sacrifice a significant amount of taunting points in order to finish the screen alive.  The more taunting points that you are putting at risk, the less value additional firefox smash points have.  In addition, if you are in a situation where you are very close to running out of time, you should be LESS willing to try for extra smashes because any extra delay caused by respawners could cause your Timer to run out before you can finish the screen.

This leads us directly to this quiz's Bonus questions.  Based on the above logic, I think that it might be more correct to pass on extra smashes if the Timer were at 4700 or maybe even 3700 points AND at 700 points.  If the Timer is at 1700 or 2700, like in our current situation, it "might" be better to go get some extra smashes, depending on the situation.

Next, we have the position of the Firefoxes.  It's important to understand that you can NEVER get a "really big" hammer in this situation because of the two freezers.  Because of the location and current action of the two freezers (sitting frozen on the top girder) you will NOT be able to smash them.  That only leaves 3 other firefoxes in play.  Fortunately, in this example, both of the firefoxes that we have just smashed have just about immediately respawned on the left side (and are not bottom spawners!)  In general, if you have firefoxes which have not yet respawned or have respawned on the bottom, you should be MUCH less willing to cross over the half way line en route to going after more smashes.  But in this case, we still have 2/3 of our hammer cycle remaining and do not have to worry about triggering a bad respawn.  Next, it is pretty unlikely (although still possible) to be able to smash the one which just respawned down on the 2nd platform.  The one that we probably have the best chance of smashing is actually the one that is now climbing the ladder up to the 4th platform since the one which is already located on the 4th platform may end up fleeing up to the 5th platform before we can get to it (this happens VERY often, more than it "should" and I would love to see an explanation for this via code analysis).

For the reasons described above, I believe that in this example it would be a mistake to to choose option E, which is basically trying for a "really big" hammer, mostly via left side smashes.  I think that the most likely result here would be one fairly quick smash, and then you'll be waiting around and waiting around, frustrated that you can't reach the freezers, and then sometimes getting a second smash just before the hammer expires.  On those times where you get the late second smash (say, 50% of the time), you are VERY likely to get two respawners on the right side which you cannot smash.  This will often lead to delays, possible death, and often will destroy most, if not all of your taunting points (although you will often still result in at least some Bonus points).  The two smashes are only worth 858 points.  You are losing 200 points in the time it takes you to run back across the screen without the hammer.  You'll now need to unplug the right side rivet and climb up into taunting position without any delays (you'll be delayed pretty often) and most likely by running across and jumping back (-100 points) and then you'll have to GAIN at least 400 points through taunting on average to make this the best option.  This is because earlier with option F we scored 2800 points off of the last 1900 Bonus for a GAIN of 900 points -- in this case we have a GAIN of 858 (from smashes) minus 200 minus 100 (inefficiencies) which means we are now 342 points behind choice F.  So, we are left with around 1600 on the clock when we begin taunting and must GAIN 400 points.  With this much time on the clock we only gain about 400 points by the time we hit 000.  So, we'll have to be finishing this screen somewhere in the 000 range ON AVERAGE, WITH two firefoxes on the right side of the screen for this to be a superior choice.  Not gonna happen.  Next, another 25% of the time we might only get ONE smash.  This yields 429 points -- but at least we now only have ONE respawner on the right side getting in the way.  So, the likelihood of delays and death is significantly lessened.  There is also the possibility of trapping that firefox in some situations.  (But remember, by trapping the firefox, we must eventually climb back DOWN for the last rivet, losing about 300 points).  But, there is still the 200 point loss for running back across the screen and another 100 points for getting into position.  We're now at a GAIN of just 129 points and must GAIN another 800 points with only 1600 on the timer for the smallest of edges over choice F -- if we max it out with 4 jumps at 000 we can expect exactly 800 points gained -- and with a firefox in play on the right side we can expect to lose many of these taunting points some portion of the time -- again, inferior.  Perhaps the last 25% of the time we might get a favorable situation -- like getting two quick smashes, being able to run back across early, and possibly getting one or two respawn smashes before finishing.  So, some small percentage of the time we'll clearly end up with more points by going for the "really big" hammer -- but in my opinion, in this situation you are losing points in the long run.

This brings us to option D -- trying for ONE quick firefox "poach".  With 2/3 of our hammer cycle remaining it is possible to run to the left, get a quick poach and run all the way back to the right and unplug the rivet before the hammer expires, losing almost no points to inefficiency.  In addition, there is some (relatively small) percentage of the time where the respawn is delayed and you can run back across the half way line in time for the respawn to occur on the left side.  By positioning yourself so that you are just barely reaching across the gap you can give yourself just a little extra chance of this happening.  In this case, you can be in taunting position with perhaps 100 points of inefficiency as compared to option F but with a gain of 429 points from the smash -- a clear 300 point victory.  Let's say this might occur 15% of the time.  Next, we might not be able to reach any firefoxes right away from this position and we might decide to abandon the smashes and just get back into position with a 100 point inefficiency -- a clear 100 point defeat, perhaps 10% of the time.  Next, we might poach one firefox, run back across and smash the respawner, again with only 100 points of inefficiency -- now we're talking about a clear 700 point victory another 15% of the time.  Next, another 10% of the time the firefox may end up being a bottom spawner and never climb up in time to cause any delays or lost taunting points.  Another clear 300 point victory.  Perhaps another 10% of the time we get lucky and "poach" TWO firefoxes which are sitting right on top of each other -- this is just as good and is often better than the best scenario for option E discussed above -- you'll often get at least one of the respawns delayed and sent over to the left side in this situation and you can also often smash one or both respawners for big points.  Many things could happen in that scenario but it's generally a positive outcome.  Lastly, (the remaining 40% of the time) we'll be unable to smash the respawner and it will eventually climb up to interrupt our taunting.  By how much?  It will vary widely but sometimes it will steal most of our taunting gains.  Again, this is where the Timer comes into play -- we don't want this happening if we began this process with 4700 Bonus since we are now losing a pretty big chunk of taunting points perhaps 40% of the time or so -- with this much Bonus it might be better to pass on the poach and just finish the screen with option F.  But, in our current example, if we average this situation out where we are looking for a 900 point taunting GAIN but have anywhere between 0 and 900 points stolen from us 40% of the time, we can say that during this 40% we have about 450 points stolen.  We gained 429 from the poach, lost about 100 points in inefficiency and had 450 points stolen for just a slight loss of about 100 points.  When you look at it in the long run, in this specific situation, I believe poaching ONE firefox is profitable.

Lastly, we could consider option F followed by option E or option D.  It's an interesting sequence, especially if we are fast enough with unplugging the rivet to conserve at least half of our hammer cycle.  I actually do believe that there are situations where this would be the most profitable choice -- most likely that would involve situations where you have a strong chance at a "really big" hammer combined with a relatively low Timer.  In some ways, the unplugged rivet will "protect" you against respawners coming to get you after your hammer has expired if we just hang out smashing firefoxes on the left side.  But, at the same time, a firefox on the 4th platform on the right side in this setup has only one place it can go -- UP.  You could end up with one or more firefoxes climbing up and guarding the last rivet at the top right, causing risk and significant delays.  Still, like I said, this might be worth the risk in certain situations, but I don't think this is one of them.  Even in the case of combining it with poaching a single firefox, I think it's inferior since you have significantly less chances of smashing the respawn and it might be more likely to climb up and guard the last rivet.

In conclusion, in this particular setup, my opinion is that option D is the best, and if the Timer were different as per the bonus questions, the best option will sometimes become option F as discussed.
Donkey Kong:  1,206,800  Kill Screen
Donkey Kong:  898,600     16-5
D2K:                 380,200     L=9
Donkey Kong Junior:  In Progress
Member for 11 Years DK 1.2M Point Scorer Wildcard Rematch Champion Blogger Former DK Level 1-1 World Record Holder Former DK No-Hammer World Record Holder DK 1.1M Point Scorer Individual Board Record Holder DK 1M Point Scorer Former DK World Record Holder - MAME DK Killscreener Twitch Streamer

Offline mikegmi2

  • Elite Member
  • *
  • Posts: 431
    • Awards
Re: Point Pressing Pop Quiz #2
« Reply #7 on: July 16, 2013, 08:11:43 am »
These are fun Dean thanks for doing the write ups and posting the quizzes.

Yea I dropped the ball on the 2 freezers...oops.
Donkey Kong - 1,076,000 (arcade, KS)
Member for 11 Years DK 1M Point Scorer DK Killscreener Twitch Streamer