Donkey Kong Forum
General Donkey Kong Discussion => General Donkey Kong Discussion => Topic started by: homerwannabee on August 29, 2013, 04:37:40 pm
-
I think I am there now. Recently I have been taking my games fairly deep. If I just keep at it playing 4 hours a day I believe by the end of September I will get a killscreen. For instance today I got a 629,100 game, and a 485,600 game. I screwed up the 485k game as I choked, and killed off two men on the level 13 elevator stage. Despite the last stupid deaths I really think I am getting the hang of the elevators again.
So I was wondering for all of you. When did you realize that you were on the verge of a killscreen? For me I think it is mastering the barrel stage with top hammer that was the clue for me.
P.S. Hopefully I don't jinx myself with this prognostication! LOL
-
After what I perceived were the elimination of the vast bulk of primarily user error deaths. Your chances are greatly increased after that happens.
-
I managed to retrieve a tasty old TG thread from the Wayback Machine the other day which might help answer this question...
http://web.archive.org/web/20110724064547/http://forums.twingalaxies.com/viewtopic.php?f=30&t=14198 (http://web.archive.org/web/20110724064547/http://forums.twingalaxies.com/viewtopic.php?f=30&t=14198)
Fresh off my first Donkey Kong kill screen (MAME), I decided to do a mathematical analysis on exactly how much luck you need to reach Donkey Kong's kill screen. Any veteran player will tell you no matter how good you are, luck plays a role. Sometimes the randomness is just too overwhelming to overcome. My analysis confirmed my gut feelings and also confirmed what was being said by Brian Kuh and Steve Wiebe in King of Kong and interviews.
I'm sure many of you will find this rather discouraging, but I hope an equal number of you will find this motivational. But regardless, these numbers are pretty interesting.
First my mathematical simplifications (to make calculations easier):
1- you score an average of 7500 points per board cleared. i know that barrel boards tend to score higher and elevators lower, but bear with me. that average would give you a kill screen score of 870K plus a few extra points for any deaths and the kill screen itself. That's a pretty respectable non-point pressing score and was roughly Billy Mitchell's high score that stood for 20 years.
2- you have the same probability of passing each screen. i know conveyors and rivets tend to give people the most trouble, but i don't think this assumption affects the odds TOO much and it sure makes calculations a lot easier.
3- the game has no kill screen (ie I'm really calculating the probability of passing 116 screens on a never ending game)
4- your learning curve has plateaued (ie you are not improving your game as you try to reach the kill screen)
Finally the results-- the columns of the chart in order are:
1- your AVERAGE score (not your high score). you have to be honest with yourself on this one for the chart to work. I suggest playing 10 CONSECUTIVE games and averaging them ie no restarts. it's impossible to calculate the odds based on your high score, because it would require knowing how many games you played to achieve that score. even then, that's a very difficult calculation.
2- your AVERAGE level completed
3- probability of passing each screen
4- odds of making it to the kill screen in 1 game
5- how long it would take you to reach the kill screen if you played 3 games EVERYDAY.
6- some real life events of roughly equivalent probabilities.
Simply look up your average score and then look at the last column to see how long it will take you.
120,000 5-2 75% 30,594,967,731 ____ 27,940,610 years (probability of hell freezing over)
125,000 5-2 76% 7,706,778,296 ____ 7,038,154 years
130,435 5-3 77% 1,989,228,620 ____ 1,816,647 years
136,364 5-4 78% 526,070,264 ____ 480,429 years
142,857 5-5 79% 142,541,127 ____ 130,175 years (probability of winning the Powerball with a single ticket)
150,000 5-6 80% 39,573,092 ____ 36,140 years
157,895 6-1 81% 11,258,743 ____ 10,282 years (probability of dying in a plane crash each time you fly)
166,667 6-2 82% 3,283,400 ____ 2,999 years
176,471 6-3 83% 981,908 ____ 897 years
187,500 6-5 84% 301,277 ____ 275 years
200,000 6-6 85% 94,913 ____ 87 years
214,286 7-2 86% 30,730 ____ 28 years
230,769 7-4 87% 10,238 ____ 9 years (probability of dying in a car accident over 1 year)
250,000 8-1 88% 3,515 ____ 3 years
272,727 8-4 89% 1,246 ____ 14 months
300,000 9-2 90% 458 ____ 5 months
333,333 9-6 91% 175 ____ 2 months
375,000 10-6 92% 70 ____ 23 days
428,571 12-1 93% 29 ____ 10 days
500,000 13-4 94% 13 ____ 4 days
600,000 15-5 95% 6.1 ____ 2 days (probability of rolling a 7 in craps)
750,000 19-1 96% 3.2 ____ 1 day
1,000,000 24-5 97% 1.9 ____ 0.6 day (probability of flipping heads)
1,500,000 35-6 98% 1.3 ____ 0.4 day
3,000,000 69-2 99% 1.03 ____ 0.3 day (probability that Windows will crash on your way to the kill screen-- j/k-- that would really suck)
This is consistent with some things that Brian Kuh and Steve Wiebe said.
1- Brian Kuh's said he scored about 220K at a DK game at Funspot. Assuming that's his average score (obviously a sample size of 1 is not great), it would take him 9-28 years to reach the kill screen-- which explains why he has never reached the kill screen.
2- Steve Wiebe in an interview said his average game was 600K. Someone also said it took him 7 games to reach the kill screen at a live performance (my calculations say 6).
These numbers are pretty consistent with my personal experience. I was averaging 200-300K for a long time and felt like I needed a MIRACLE to reach the kill screen. When I was finally able to average about 400K I felt like I needed just a little luck and sure enough it only took a couple of weeks.
This chart should also give some people insight as to how to maximize your points ie risk vs. reward. I never understood why people try to point press when they haven't even reached the kill screen. You can see from the chart that each 1% increase in chance of death on a screen reduces your chance of making it to the kill screen 3 fold and I'm pretty sure a lot of point pressing techniques increase your chance of death by several percentage points. It's a lot sexier to talk about point pressing and reaching 1M, but my advice is that unless you've reached the kill screen, only point press conservatively-- this will maximize your score AND your probability of reaching the kill screen.
SUMMARY:
If your average score is:
less than 200K - you're better off buying a lottery ticket, don't hold your breath and don't fly any planes
between 200K-250K - you need a major miracle, and drive less often than you play donkey kong (otherwise you will die in real life before reaching the kill screen)
between 250K-300K - you need a lot of luck, don't get married, have any kids, or shower because you will need all that gameplay time.
between 300K-400K - you need a little luck and a lot of quarters
greater than 400K - you're reaching the kill screen fairly regularly, consider point pressing
I'm curious if these numbers are consistent with other people's experiences.
Hank
-
There it is! Nice find, Chris!!!!
-
OK, last time I did a 25 game average I was getting to level 7-5. But here is the thing, my elevator effectiveness was way off. I was rusty. Also what Hank's chart does not take into account is that some people like myself can get on a roll per say, and do much better than one would usually do.
Also I wonder. It says if one averages 500,000 their odds are about 13 to 1. OK I am about half that, but I have reached 500k 9 times now. Does that mean if I reach 500k 4 more times I am about due for a killscreen? Also I have reached 600k now 3 times. 2 actually in the past 10 days. So I reach 600k 6 times does that mean I am also due for a killscreen?
Just wondering if the odds can work that way as well?
-
I realized it when I hit 19-2 with 2 men left. Think that is basically my VIP pass to Killscreen.
-
I beat my best of 505,600 (top hammer), which was a score that eluded me for about 11 months, with my two hammer game of 696,800. I watched Ethan and a couple others play the rivets, I played a success rate of 89% about of 100 attempts on the conveyor, and started killscreen attempts with top hammer only. Got the killscreen 10 days later. I have not ran top hammer games since.
505,600 4/28/12 MAME US Set 1
696,800 4/1/13 MAME US Set 1
714,200 4/8/13 MAME US Set 1 Beat Brian Kuh
893,800 4/11/13 MAME US Set 1 Killscreen!
-
OK, last time I did a 25 game average I was getting to level 7-5. But here is the thing, my elevator effectiveness was way off. I was rusty. Also what Hank's chart does not take into account is that some people like myself can get on a roll per say, and do much better than one would usually do.
Also I wonder. It says if one averages 500,000 their odds are about 13 to 1. OK I am about half that, but I have reached 500k 9 times now. Does that mean if I reach 500k 4 more times I am about due for a killscreen? Also I have reached 600k now 3 times. 2 actually in the past 10 days. So I reach 600k 6 times does that mean I am also due for a killscreen?
Just wondering if the odds can work that way as well?
George, I don't think you're reading Hank's old analysis correctly. His analysis and chart is mainly an attempt to approximate the variability of the randomness of a full game and whether those factors generally factor in your favor or against you in any specific kill screen attempt. In other words, the idea is to take a player of a specific skill level at a snapshot in time (assume this player will NEVER get any better regardless of how many months that player plays all of these future games). So, for any given situation, his decision making, precision of execution, percentage chance of a brain fart, etc, etc is always identical. That player, ON AVERAGE, gets through a certain number of screens (or gets a certain score). Based on that, how LUCKY does this player have to get for all of the stars to align and he gets a kill screen. So, the wild barrels are not coming at difficult times, the barrels are steerable at the right times and are not coming down on their own at the wrong times, the star pattern never breaks down, he gets a huge number of free passes, etc, etc. How many things that are controlled by randomness have to go unusually well for this player to reach the kill screen on any given attempt.
This does NOT literally mean that a player with an average score of 250k will literally need 3 years to reach a kill screen when playing the game every day. When playing that often, this player will obviously improve his skill level naturally, plus, as you said, a player can always get on a roll (in the zone) on any given attempt and PLAY BETTER than they usually do.
It's easy to mix up the subject of this thread with what Hank's post was talking about when reading that chart -- but it can still give people a decent understanding that they might be less than a month away from a killscreen if they are consistantly going deep in games, etc. But, the timeframes may not line up with reality because those variances or improvements in the player's skill is not factored in.
-
I got 685 800 on MAME, and I knew I would get to level 22 very soon, I went to my friends place to play on the arcade version and that day I made it to level 22: 879 500, June 10, 2006. I almost had a heart attack on 21-6, they were making me earn it, and I I was on my last man. That was more than 7 years ago now, granted I do not play as often as I did back then but I do want to get there again. I have attached a picture, unfortunately there is no video recording.
-
It's easy to mix up the subject of this thread with what Hank's post was talking about when reading that chart -- but it can still give people a decent understanding that they might be less than a month away from a killscreen if they are consistantly going deep in games, etc. But, the timeframes may not line up with reality because those variances or improvements in the player's skill is not factored in.
I also forgot to mention that I have had 6 of my 10 best scores all time in the past 10 days. So if I am doing that in the past 10 days, I presume I will probably get there in the next 32 days (end of September)
Here are my top 10 scores
8/20/13 690,500
8/29/13 629,100
6/20/13 603,500
6/25/13 554,200
7/4/13 540,700
8/27/13 535,500
8/28/13 533,200
8/21/13 519,500
5/23/13 504,000
8/29/13 485,600
-
Thanks for not making me update all of these George, lol ;) Great work.
-
Thanks for not making me update all of these George, lol ;) Great work.
No, I made you update every single improvement. Look closely at the dates! ;D 8)
-
I've hit 19-6 with two men and was dead at 20-1. Lol
-
Lol. Rick I was just watching you on Dwaynes interview about a few hours ago.
-
Thanks for digging this up Chris! This was one of my first posts to the old TG forums (I only signed up for an account once I had killscreened, hence lots of people said 'I came out of no where') and one of my favorites. I'm glad it's not lost.
George, I think my simplifications 2 and 4 cover your concerns:
2- you have the same probability of passing each screen
4- your learning curve has plateaued
As with any mathematical calculation of a complex real world problem you often need to simplify things to make a reasonable/manageable model. Obviously there are lots of real world variables that are not taken into account and really can't be since so many issues are individual (how often are you 'hot' and not, what's your threshold for restarting, etc.). Note this is based on your AVERAGE score not your high score, although given your high scores recently I don't doubt that you will eventually reach the killscreen. Your AVERAGE score is just shy of what's needed to have a realistic chance of reaching the killscreen. The chart does NOT tell you how long it will take you to reach the killscreen as you are obviously improving as you play. The chart does tell you that when you are averaging 300-400K you have a realistic shot.
What's interesting is that these calculations (despite the simplifications) I've found to be pretty accurate when talking to people or hearing stories about players' progressions throughout the years. I gave a few examples in the original post. I remember John Marks posted an extensive record of his scores to this thread and his numbers pretty much confirmed the chart (he was averaging maybe 350-400K before he killscreened).
Anyhow to answer the OP, when I had my first breakout game of 640K (my first submission to TG) that's when I knew I just needed to get a little bit better and I'd be able to reach the killscreen (I didn't make this chart at the time but this was just my gut feeling). After that I had maybe 4-5 700K games and then BOOM killscreen.
EDIT: Oops I'm on yoomsss' computer
Signature
World record for most kisses to Hank
-
World record for most kisses to Hank
Geez, you can't leave yourself open like that!
Imagine if Vincent had been lurking...
-
As competitive as Vincent is, I don't think that's a record he would go after He is strongly #nohomo.
-
As competitive as Vincent is, I don't think that's a record he would go after He is strongly #nohomo.
I'd rather lick Billy Mitchell's beard instead of kissing Hank (no homo)
1- Brian Kuh's said he scored about 220K at a DK game at Funspot. Assuming that's his average score (obviously a sample size of 1 is not great), it would take him 9-28 years to reach the kill screen-- which explains why he has never reached the kill screen.
But Brian Kuh reached a kill screen of 849k in 2002 (at home, as he told us at Funspot). That obviously draws controversy, but I do believe him.
-
214,286 7-2 86% 30,730 ____ 28 years
230,769 7-4 87% 10,238 ____ 9 years (probability of dying in a car accident over 1 year)
How many years did it take him? Only joking... actually I've seen him play. I'm sure his average is higher than 220K.
-
I met him this past year up at Funspot (Brian) and he was a really nice guy and an excellent Super Sprint player as well. He was giving some guy named Elf a run for his money ;D
-
I met him this past year up at Funspot (Brian) and he was a really nice guy and an excellent Super Sprint player as well. He was giving some guy named Elf a run for his money ;D
First time I have ever heard anyone call Brian Kuh a really nice guy. Most of the time people call him standoffish including your friend Richie Knucklez.
-
I met him this past year up at Funspot (Brian) and he was a really nice guy and an excellent Super Sprint player as well. He was giving some guy named Elf a run for his money ;D
First time I have ever heard anyone call Brian Kuh a really nice guy. Most of the time people call him standoffish including your friend Richie Knucklez.
Maybe I just have that effect on people. Love it?
-
I met him this past year up at Funspot (Brian) and he was a really nice guy and an excellent Super Sprint player as well. He was giving some guy named Elf a run for his money ;D
First time I have ever heard anyone call Brian Kuh a really nice guy. Most of the time people call him standoffish including your friend Richie Knucklez.
Maybe I just have that effect on people. Love it?
No, you probably just have that effect on Dbags. ;)
-
I met him this past year up at Funspot (Brian) and he was a really nice guy and an excellent Super Sprint player as well. He was giving some guy named Elf a run for his money ;D
First time I have ever heard anyone call Brian Kuh a really nice guy. Most of the time people call him standoffish including your friend Richie Knucklez.
Maybe I just have that effect on people. Love it?
No, you probably just have that effect on Dbags. ;)
Then what effect do I have on the converse? :o
Have you met Brian Kuh before? I traditionally find it difficult to judge character based on others (likely small sample sized) interactions with them, but then again I play life at a poser pace.
-
I met him this past year up at Funspot (Brian) and he was a really nice guy and an excellent Super Sprint player as well. He was giving some guy named Elf a run for his money ;D
First time I have ever heard anyone call Brian Kuh a really nice guy. Most of the time people call him standoffish including your friend Richie Knucklez.
Maybe I just have that effect on people. Love it?
No, you probably just have that effect on Dbags. ;)
Then what effect do I have on the converse? :o
Have you met Brian Kuh before? I traditionally find it difficult to judge character based on others (likely small sample sized) interactions with them, but then again I play life at a poser pace.
Remember Winky emoticon means: It's a joke
As far as standoffish goes, yeah the small sample size besides you says that. Not saying he is or not, just saying you are the first to really say that Brian Kuh is this really nice guy that I know of. Again calling him a Dbag was just a joke based on the standoffish nature people attribute to him.
-
Actually I've met Brian a couple of times. Every time I met him he was very cordial to me. I think it depends. Of course, partly because he wanted to see me play (to see if I was the real deal) but I didn't mind if he broke into my apartment. My cabinet is full of gummy ... bears.
-
Actually I've met Brian a couple of times. Every time I met him he was very cordial to me. I think it depends. Of course, partly because he wanted to see me play (to see if I was the real deal) but I didn't mind if he broke into my apartment. My cabinet is full of gummy ... bears.
Well you're a different story. You have that effect on EVERYONE! Well that is except Vincent Lemay. ;) ;)
Two winkies just to be extra sure people know it's a joke. ;) ;)
-
We should take a Kuh poll. So far me and Hank have run it up 2-0 in favor of the Kuh. I'm sure Allen would make it 3-0 even though he's never met him before, Kappa.
-
Brian Kuh is a cool guy. He's very passionate about Donkey Kong (me, Robbie and craig spoke to him for at least a hour about a bunch of DK stuff, he enjoyed it).
-
First time I met him was last year through Alpiger and I got a great first impression as far as his character. I didnt talk to him much, but I saw him again with Vincent and Craig. We talked DK for a really long time about DK and he really cares about the game and was interested in the current culture behind the game. He even signed some things for Allen and invited us out to dinner afterwards. When he had told us he kill screened the game in 2002 I actually believed him since he said it with confidence. But at the same time he was afraid to say his highest amount of fireballs jumped in one jump (he thought 3) but said he needed to check his notes before he gave us an answer. I go to Funspot a decent amount but only see him during the tournament, but it would be nice to see him there more often. My buddy for a couple days after the tourney played some games with him and he was willing to give advice. From what I heard he is very knowledgeable about many of those games and is nice enough to help people out.
-
Yer Vincent and Robbie....4-0 for the Kuh!
-
We all know how the kok went down and know to take it with a grain of salt. I hope Brian is genuinely a great guy misunderstood.