Nintendo Popeye Bottle Glitch

I finally captured video of the Popeye bottle glitch on the Nintendo NES.  Normally jumping down on the see-saw will kill you if you get hit by a bottle on the re-bound.  To avoid it, after you jump, press to the right and at the precise moment when the bottle should hit you press the punch button.  Popeye's animation will not change but if you time it right you'll get the points and not die.

Popeye World Record

I'm still working on the Popeye World Record for Nintendo. My last update was on April 25th, where I really didn't have much progress at all. Setting a world record is very difficult when you're displaced. It's even more difficult without a television!

Although I'm still displaced, I finally got a television, and have been mulling the Popeye World Record again. Like most things I've been studying it programmatically - looking at how the programmers actually designed the game and to look for flaws in it for an advantage. Not necessary glitches as that would distract from earing an actual record, but more along the lines of knowing how the AI was designed.

I haven't been able to figure it out completely but I am certain that Bluto can be controlled. He only has a few "modes" to which he operates:

1. Search: When he's "looking back and forth for Popeye"
2. Chase: When he walks in the direction of Popeye
3. Attack Above: Jumping up to hit Popeye
4. Attack Below: Reaching down to hit Popeye
5. Throw Bottle: Throwing a bottle

To determine what triggers each of these events simply takes some studying. If the modes are triggered by events within the game then it should be pretty easy to figure out. However if there is a random element to it, it will be more difficult.

It seems pretty obvious that the direction of "Chase" is simply a matter of determining if Popeye is to the right or left of Bluto and simply to start moving in that direction. Never once have I observed Bluto to walk in the opposite direction of Popeye when they are on the same level platform after completing a "Search Sequence" (unless of course you've eaten the spinach.) This is the easiest of the AI to figure out (If Popeye is LEFT of Bluto then Walk Left, else Walk Right).

"Search" seems to be the most difficult for me to figure out. What the heck actually triggers Bluto into "Search Mode"? I'm suspecting it has something to do with the animation frame that Popeye is currently standing in but I haven't been able to prove it. This is actually important to figure out because if you can trigger Bluto into "Search Mode" on the first stage while under the barrel you can knock it down for big points. The lower down on the screen Bluto is the higher the point value becomes. Of course the timing becomes more difficult too.

Again however, if these events are triggered randomly, such as a random number generator then it won't do much good to decipher the AI... infact it might be impossible to do so.

There is two ways that this can be figured out. The first is to use a debugger on an emulated version of the game, but thats the cheap and chump way to do it. The honest way to do it is to try to create a pattern not unlike the PacMan pattern. If you can move around the first stage in a predetermined way and have Bluto follow exactly each and every time then you can be certain that its not random.

Having the next two days off will give me ample opportunity to experiment!

Videogame Box Art History

I've been reading "Racing the Beam" which discusses in intimate detail what it took to design games for the Atari 2600. It's moderately technical and probably would lose many readers due to it's jargon.

However there are a couple sections that I found to be particularly interesting that dealt more with marketing than programming. The book repeats the well known story that some of Atari's finest programmers left Atari and created Activision after they felt they were not receiving the recognition publicly and financially that they felt they deserved. The book provides an example of two games: Indy 500created by Atari and Grand Prix created by Activision. A photo shows the box art and screen shot for both games. The caption points out to the reader that Indy 500's box art features an "intricate, realistic painting of the game's subject" where as the game is more abstract than what the art alludes to.





The Activision box art (and not-coincidentally most of Activisions boxart for the Atari 2600) closely resembles what the actual Grand Prix game looks like.





I've always been a fan of box art and I believe that the art on many of the Atari 2600 boxes are some of the finest ever created. Although true, some people may have been convinced in buying a sub-par game because the art alluded to an experience that just wasn't going to happen, it brings back to full circle my belief that less is often more in a video game. And although Indy 500 doesn't feature the in-game thrills that the art implies, to a child of the '80 with an endless imagination the game did feature what the art implied plus more, but not simply represented on the screen.

Now that I've been made aware of this interesting concept of Activision representing their box art as a close resemblance to actual game play it got me to thinking about the black label Nintendo games that got me interested in collecting Nintendo games in the first place. There is little dispute that Nintendo was very careful in releasing the NES in the United States alleging to go so far as to sanction games with their Seal of Quality. Almost all of the initially released games for the NES feature a pixelated representation from the game on the label and box art. Typically it was almost an exact representation and in some cases there were a few exaggerations but all in all there was little question as to what the game looked like. This quiet possibly was genius marketing from Nintendo.

Popeye Box Art



Close up of screen shot



Actual NES screen shot

Popeye Progress

Having a little time to practice Popeye I learned something new. The counter that represents the rounds changes to letters on the tenth round. This brings up a curious possibility: what happens after round Z? Does it wrap around to 0? Will it start displaying funky symbols? Will the game crash?

I also learned that the sea hag will only throw one skull when the barrel is still hanging. Drop it and she starts throwing three at a time.

Popeye

It's finally arrived! The Popeye cartridge came in excellent condition. There was no manual but it did come with a sleave. I was also able to negotiate a trade of an old computer for Dragon Warrior, a 2nd copy of Dr. Mario an additional copy of Super Mario/Duck Hunt, two RFU adaptors, 1 power cord, and 1 SNES controller which I'll probably toss.

Now I can officially start on the record.

Setting A World Record: Popeye for Nintendo

I've decided to attempt and achieve a WR score on an NES game. I've narrowed it down to one of two games. The first being "Original" Mario Bros. (not super!) or Popeye. Although I enjoy both games it seems the real challenge has been finding the games. After scouring eBay I finally have a copy of Popeye on its way. Finding Mario Bros. has not been as successful, and therefore Popeye will be WR of choice.

As with most things that I involve myself in (specifically WR scores) I like to pick apart the approach necessary to achieve the record or the desired end result. In this case, I obviously had to do a little research. For starters the current record for Popeye is 1,485,610 points set by Tom Votava on 12/01/2005. I don't know Tom Votava personally but I'm going to assume that this is probably a pretty substantial score because 1.) Tom Votava has plenty of records already, and 2.) the second place score for Popeye is 212,250.

Now that I know what the goal is, the next thing I like to do is determine how long it would be necessary to play the game to achieve the record. In this case I'll either benchmark the game by timing a few levels and recording the score, or in this example of Popeye, I'll refer to http://www.youtube.com/ for a reference. When bench marking any game that has repetitive levels such as Popeye or Mario Bros the longer the bench mark is the more accurate of an estimated time of completion you can come up with in achieving the record.





I decided to use the above video for the benchmark. The player in it takes a casual approach missing opportunities to point press. When benchmarking always take the casual players time and score. You don't want to compare yourself with a relative expert because you'll burn yourself out. After the first level the player scored 10,430 in about 1 minute 25 seconds. He completed the second level in about 1 minute 7 seconds with a score of 5,600 points more then what he had at the end of the first level, and he completed the 3rd level in 2 minutes 38 seconds earning about 16,700 points. All in all, in about 5 minutes 10 seconds you can easily earn 32,730 points.

To complete the bench mark you take the world record and divide it by the total points for all three levels. In this case it would be 1,485,610 divided by 32,730 which equals about 45. This means at an exaggerated minimum it would take completing all three levels 45 times to get the record. Determining how long it would take involves multiplying 45 by however long it took to complete the three levels (which in this case is about 5:10) which adds up to about 3 hours 52 minutes.

Now that I know how long I'll have to play for the next step is determining my availability to practice. Setting a recording isn't as easy as simply sitting down and playing, especially when you have a full time job and a small boy. Those two things obviously take precedence. So a little creativity will be necessary to pull this off. This is where I'll turn to my Nintendo DS. Without going into the exact details I can play the original Nintendo version of Popeye on my DS. We'll just leave it at that. The DS has similiar controls to an NES controller so for the most part it should be seamless in terms of practice and game play. Now, if necessary I can practice anywhere. Well almost anywhere- although my employer probably wouldn't mind me playing games at work (it is part of what we do here) they might not take too kindly if we were playing on a competitors system.