My concern is that, if there are multiple valid points at which input can be ended, and if it is not clarified in advance which method of ending input must be used for that particular game, then arguments over which movie is faster may occur in the future. The "improvement" criterion ("there are improvements other than end of input") is not always easy to use since it is possible in some cases for both sides to claim improvements over the other! In other words, the so-called improvements are linked to the choice of end of input. As an example, let's say there is an RPG. Person A makes TAS A for which the final boss is beaten and the input is ended at 3:00:00 (3 hours). Person A chooses not to obtain an item at around the 1:30:00 mark which takes too long to get even though it saves a little time in a couple areas around the 2:00:00 mark. Person B makes TAS B for the same game, which ends input at 2:59:00, but this is right at the beginning of the final boss battle for which the boss is defeated at 3:01:00. How is this possible? Person B obtains the item above at the 1:30:00 mark, saves a little time in the couple areas around the 2:00:00 mark, and uses this item at the beginning of the boss battle (the usage is the last input). The item enables the player to defeat the final boss without any input, but it takes a lot longer than just beating it conventionally. Person A claims that TAS A is faster by "final boss beaten". He also claims improvements over TAS B by not taking the item, since it does not pay off. Person B claims that TAS B is faster by "end of input". He also claims improvements over TAS A by using the item to save time in a couple places. If both TASes are submitted at the same time, which one wins? If TAS A is already published and TAS B is submitted, does it obsolete TAS A? If TAS B is already published and TAS A is submitted, does it obsolete TAS B? Fortunately I don't think such a situation has appeared yet (not one which is extremely controversial anyway), and I'm not saying that this submission definitely falls into that category. But in general it would be nice to know more about which parts have been improved independent of the choice of ending, and which are dependent on the choice of ending, especially since luck-manipulation is involved. If that is undesirable, then at least we can claim that the run is faster by virtue of showing a smaller time on this site than the previous run, which is of course because of the end of input. By the way, Nahoc is judging the run, not me. I'll voice more details on my opinion of the choice of ending ("end of input" vs. "true end") later.

edit: Visual help for this post: https://docs.google.com/spreadsheet/pub?key=0AqMjARkWFsmedFJKNVhOV3piS05pRUZKbGZucVF5akE&single=true&gid=0&output=html Please don't forget that that's why I wrote down the possible list of "ending points" because all of them is "crucial" in some way. Let me list it again and detail them each to understand what I'm talking about with examples from my and twistedeye's run (or technically: from the game): 0. Input length: very obvious, this is the point when you don't need to do anything to make the game end. this is the point you stand up and make a coffee or roll a cigarette. This can't be seen by the casual viewer (without noticing them). 1. Killing the last enemy that required to obtain the core : In my submitted run, I obtain the last core faster. This means, this is the point when anybody who watched the movie AND KNOWS THE GAME, will understand that this final hit was the last step I needed to end the game. That's why I've targeted this type of ending in my submitted run. The reason it's slower is because you have to wait for the fade out of the fight. 2. Obtaining the last core (either by standing on it or when the enemy dies on that square) : Anyone who KNOWS the game will know, that when it kills the last enemy, the game will be victory. TwistedEye's (and my WIP4) didn't obtained the last core by killing the enemy (they fight first and than make a "I will win if I obtain the last core" move later). This means, that my WIP3 (not uploaded) actually shows that last step you need to win the game, while in TwistedEye's and WIP4 you have to wait to make the lazy move, which with the game knowledge, you already know that this situation will be victory when he killed the last enemy earlier. 3. When you see the cores are obtained by dark side on the table : This is the point, that you can see, the table's cores are acquired by the dark side. It's ended, it won't be changed. Still needed "game knowledge" to understand it's over. 4. Whe you see "The game is ended..." : The "official" clarification about the winner. EVERYBODY understand that the game is ended. 5. Pressing a button to start the credits at next nolag frame : This is only added to the comparison movies I've linked of my WIP4 and TwistedEye's. 6. When the credits automatically starts without input : The reason I'm totally against this, is because it's timed with something like "number of seconds elapsed since the table is shown". It means, you need an improvement of 100~300 frames to make the ending credits appear faster. And of course be slower to make the game wait more... And really. Who doesn't understand the game is over when the game says: The game is ended? I think it's pretty obvious. Sorry for too much text, I tried to make it as obvious as it can, for 2 reasons. First is to help Nahoc to judge and others to aid some feedback, secondly because I'm curious what will you say, because my Battletoads TAS can be improved. Because... If I understanded what people wrote is that the "game is ended when ending credits is showed". And what happens when the ending credits is a set of different cutscenes and you can trigger the "1st scene" later but making a longer input to activate only the last scene? I've already made a version of that for Dylax's and DjtZ run, and I see big similarites in this case between the two games/runs

Since the choice of one ending criterion for Archon from many isn't clear (as MESHUGGAH stated), and almost no one had a problem with this choice of end (whether someone noticed or not), I'd say that "end of input" should be used by default in this case, since that is for the time given on this site. Archon isn't a game where "no input on final boss battle" applies anyway. Also, this doesn't apply to greatly glitched runs like Battletoads where the entire endgame is compromised (in other words, "end of input" applies). I think that, if there is any conflict regarding the choice of ending as to which run or runs are faster, then "end of input" has priority and will be applied if the situation cannot be resolved in a satisfactory manner. ---- Anyway, my opinion on "end of input" vs. "true end". Except in a handful of cases like Monopoly, I have generally disliked the idea of ending input early, which delays the defeat of the final boss, and in some cases, for a long time. I feel it is less like "Ah, he did a clever trick and got the boss to die without even touching the controller" and more like "Why doesn't he just finish the game already?" In that way it is like an excuse to stop playing before reaching the very end. There are a few movies like that but the three that stand out are Rosenkreuzstilette "Spiritia", Seiken Densetsu 3, and a River City Ransom submission. I don't have as much a problem with Monopoly since it doesn't fit the mold of the games described above, and relies on luck-manipulation of the whole sequence of CPU rolls after input, which is interesting in some cases. In fact, I'm considering uncancelling one of my old Monopoly submissions. It is, after all, nearly 3 seconds faster than the published run, by "end of input" of course ;) . jlun2 wanted me to uncancel it, so now might be a good time.

Oooh, an improvement over my run! Hats off--I always felt mine was pretty ugly. Love the idea about finishing the battle with the wizard fight, and, personally, I've always loved the idea of early-end-of-input, cheating away frames, hehe. It does definitely make the end of MESH's run seem awkward, though, until end-of-input occurred to me. Then I read the discussion in the thread! So I said to myself, "I'll fix that easy. I'll just re-TAS MESHUGGAH's wizard fight so that the dragon is standing on the correct pixel already when he fires the killing blow! Hold left for justice!" ...Turns out that, despite the Dragon having one of the slowest recharge speeds in the game, he just can't get over there in time. Ending as close as possible still ending input on frame 4207 saves only 162 frames--finishing on the right pixel still requires 34 frames of input. ...And now I'm confused, because, even changing the fight and extending end-of-input of this submission to bring the "THE GAME IS ENDED" message up as quickly as possible, my OLD run is still faster by 32 frames! Unless your chart timed it differently of course should I I ah it's 3am thanksgiving morning i'm going to bed, want to check out MESHUGGAH's v4 but maybe later Great job on the run, Archon is a very fun game, I personally approve of end-of-input myself but support whatever is decided

My WIP4 is uses different RNG to get better spawns and different enemy units moved (to make the "fastest" ones, fewest moving etc) with your strategy of obtaining the cores. Also I removed unrequired lag frames. I can upload WIP3 which is the same as this submited run but goes left and reduces waiting as much as possible within the same input length.

This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [2211] NES Archon: The Light and the Dark by MESHUGGAH in 01:10.00