Head Custodian and Caretaker
My personal opinion is that there should be a set of standard rules, that EVERYONE abides by. No i**s or but's.
Those of us setting arcade records are (I would hope) all going to film inside the machine at the end and show DIP switches, what is so hard about putting the machine on the specified existing settings?
Jace is exactly right about what he said from a programmers standpoint, and unless someone is going to analyze every single arcade games code, there is always going to be doubt on certain titles if free play has an impact.
And to add to the freeplay argument, what is the story with inserting multiple coins before starting? It does have an impact in certain games, it has been shown that it does. Certain games also play differently if not hard reset between attempts (Wonderboy is one I know off the top of my head). Why should some games be allowed to not be reset between record attempts and others must be turned off - is it that hard to turn the power off and back on?
I do agree that some of the old TG requirements were probably over the top, but certain things regarding setting world records are just no brainers, they take seconds to comply with - why complicate things by allowing players to do whatever they feel like?
Do you happen to have access to the "old" guidelines? I tried accessing the FAQ on WayBackMachine but I can't. I would be interested in reading them again, and I think it would be a good starting point for debating issues.
I hope I'm wrong, but I can see arguments starting in the future when WR's are community accepted, and then someone from 2010 says "But he did this and that, and I wasn't allowed to"
We've all been waiting for score submissions for so long, and I'm as happy as the next person that they are being taken, it just seems like a mistake to all of a sudden ignore so many of the old rules.
Here is the response by Matt Osborne who develops free play kits today across many different golden age arcade hardware platforms:
The validity of that argument hinges on the coin drop being some kind of magical global reset of the machine state. And that is simply not the case. I've studied the code of many games, and never have I seen one that reset on coin drop. There's no logical reason to do so.
You can never ensure that players are starting from exactly the same machine state. Even if you took two identical machines, turned them on at the same time and tried to drop a coin in each at the same time, the probability of any pseudo-random number generator getting seeded the same is very slim. That's because the seeds are typically generated from some processor register or counter that's changing value thousands of times per second.
As far as freeplay goes, that should not provide any advantage over coin drop. How a game implements freeplay can vary, but one thing you can count on is that when Start is pressed the code goes into game mode just the same as if you had dropped a coin first. The whole coin drop ritual clearly came about based on the assumption that that somehow gave players a known and level starting point. There is no technical argument behind that assumption that holds up.
We have been provided a vehicle to implement our own vision of what we feel is reasonable evidence that a score was accomplished via elite skill on a title. It is up to you, myself and the rest of the community to accept or reject submissions. If you feel a player did not accomplish their score truthfully, or in what I expect is a bulk of unaccepted cases, there is not enough evidence provided to make a clear distinction, then I would recommend downvoting it or not voting as it were, as you see fit. That's the role the community has been tasked with. I'm under the assumption that folks voting on a performance have some degree of familiarity with the title/platform and if the evidence provided supports their score claim, at least that is the way I am approaching it from my perspective. All the votes I've given are on games which I have spent significant time on and feel comfortable with evaluating a performance, though everyone's mileage may vary in that regard, that's the path I'm taking so far.
Looks like the topic is well handled above.
So I'll now throw in a side angle. These programmers at the turn of the 80s were focused on coin profits, they didn't have the time nor the monetary motivation to assess how a game played to a master who had spent excess time learning the nuances of a game. It has been stated that they tried to make interesting enemy dynamics, mixed with a certain amount of unintended bugs that hopefully had the charm to make them "features" and the rest was left to the fate of 1's and 0's and human psychology.
Here is one you guys probably don't realize in that vein of thinking. On blue romset for robo. it was coming back from the field that robo on the original yellow/orange/red romset was too difficult (which affects coin op profits). so they chose to drop the default difficulty from 5 to 3 and they added a "bozo" mode. on wave 1 if the player is doing poorly the difficulty of the game was dropped to the lowest possible for that wave. It made sense and quite ingenious...but probably blows the mind of an intensely pedantic classic gamer.
thankfully the classic game programmers went well beyond the mindset of being a pedantic player (lots of them enjoyed chess and math).
Hey guys! This is my first post on the new TG forums.
I'm in agreement with Robert and the others. Unless there is a proven exploit to a particular game, I see no reason to disqualify a submission that was done on freeplay. The code logic behind freeplay is usually quite simple. It typically works like this: the game will read the coinage setting DIPs on boot, and if the freeplay combo is found it will set a flag somewhere in memory. I search the code disassembly for any read instructions on that memory location, and also set a read watchpoint in the MAME debugger to catch any tricky (i.e. indirect) reads. I then analyze the code around these reads to see how it's using the flag. I have never found a case where is was used during gameplay. It is only used during attract mode (e.g. when a Start button is pressed to determine whether to go into game mode).
That's my experience anyway. And I've probably studied 30+ classic games up to this point. I've found that games from the same manufacturer (at least on the same or similar hardware) use the same basic logic for this kind of housekeeping stuff. Which makes sense, since they are going to recycle as much common code as they can.
Sort of unrelated to the topic at hand, but I thought I'd mention anyway... When I add freeplay to a game that never had it, one simple way to implement it is to simulate a coin drop. I essentially trick the game into thinking a coin has been dropped when a Start button is pressed. I always take the simplest and least intrusive approach first. I know games with freeplay or high score save mods added are not allowed for submissions, but I want people to know that I take great care in making sure any changes I make do not affect gameplay in any way. I sometimes find bugs in the original code too... but I leave them in! They're usually just stupid inconsequential things. Very tempting to fix sometimes... but you have to leave all the quirks in!
Actually, I knew that about PacPlus. The difference here is that one can do it EVERY TIME as opposed to 1 out of every 64. This amounts to a HUGE difference in my mind.
As for MsPac, I wasn't speaking of the fruits (which I already knew were unaffected), I was speaking exclusively about the ghost behavior. In my mind, being able to CFP MsPac goes contrary to the intentions of the developers, who very definitely intended to eliminate this possibility.
Naturally, controlling the fruit in this way would be far more useful than controlling the ghosts, and I don't think having a game set to free play would allow one to beat the WR. As you point out, the fruit behavior is unaffected by this. Nonetheless, the possibility of creating CFP patterns by taking advantage of free play bothers me enough to believe that the technique should be banned, whether or not any records are at stake. Ms. Pac-Man was designed to be unpatternable (or at least, not patternable for the first 30 seconds of every maze).
[EDIT: I meant to say "the first 5 seconds of every maze." Naturally, after the first reversal, it is possible to group and fall into a pattern of sorts.]
Last edited by anningmay; 01-06-2015 at 10:15 AM.