ScoreHero Wiki : GuideToReliablyObtainExactScores

ScoreHero :: Forums :: HomePage :: Categories :: UserIndex :: PageIndex :: RecentChanges :: RecentlyCommented :: Login/Register

A Method to Reliably Obtain Exact Scores (v1.0)



Wow this is finally done. Any aspiring Proof Apprentices should feel free to fire away with any questions or clarifications they might want, and any accomplished Proof Journeymen (you know who you are) should feel free to chime in with any suggestions or tips and tricks that they do that i missed in this guide or just to give an alternate slant on any topics contained herein.

Table of Contents:
I. Opening remarks
   a. Overall Strategy
   b. modulo 50
II. Overview of song sections (How to Start)
III. 'Holds Section'
   a. Considerations when choosing a holds section
   b. How to get the right mod 50
   c. Misc tips
IV. 'SP Section'
   a. Basic SP section Strategy
   b. Misc tips
V. 'Miss Section'
VI. 'Combo Section' (Bulk of the Song)
VII. Notes on Practical Execution of the Method (Putting it All Together)
VIII. Notes Specific to Obtaining Cutoff Proofs
IX. Trivia


I. Opening remarks


a. Overall Strategy
What is it that makes exact scores difficult? To the layman, saying 'I am going to score 133,337 on this next run' seems like a silly prediction, or something on the par of a performance mode run where it would require precise memorization of exactly what notes to hit and miss, but in fact it is not the case. The aim of this article is to reveal the Barfo Method for getting exact scores (there may be other methods). I was having a very difficult time figuring out a way to cogently explain this method until i stumbled upon the question that is the first sentence of this paragraph ('What makes exact scores difficult?'). As an interesting aside, in typical Hitchhikers Guide to the Galaxy fashion i had already had the answer to this question in my method, but it took realizing what the question was before i could embark on this project of how to explain that answer.

So then, What makes exact scores difficult? The answer is the interplay between three discrete but related factors. The first is getting the score into the general ballpark (all but the last two digits) of the target score, the second is getting the last two digits to be exactly right, and the third is to do both of those things without failing the song (or minimizing the chance of failing the song) at any point. The crucial insight of my method is to detangle these three challenges so that they can be tackled separately and in a logical order that maximizes the chances of success. In practice the method i present first attempts to get the last two digits correct, and then ensures that the song will not be failed, and only then ensures that the song is finished in such a way that we get the proper leftmost digits (which in essence means getting the exact score, since the rightmost two digits were already solved).

b. modulo 50

In order to systematically approach how to get exact scores, and specifically the issue of 'last two digits', the term 'modulo 50' (hereafter referred to as mod50) is crucial. 'Modulo' is a mathematical operator which essentially means "the remainder after long division". For example, the 'leet score' of 133,337 has a mod50 of 37, while the 'fake score' 123,456 has a mod50 of 6, while the 'OMFG ES FC'd Institutionalized score' of 431,374 has a mod50 of 24.

Because the basic note score in GH is 50, this means that for any scores that have the same modulo 50 are interchangeable just by hitting normal, non-hold notes. This is crucial observation because the method for getting the proper 'last two digits' will be hitting hold notes at varying lengths to the correct score and then not ever hitting another hold again (so as to avoid messing this up). By thinking in terms of "mod50" rather than 'last two digits' we can make our job half as difficult form the get-go as there are only 50 possible states rather than 100. To reiterate, if im trying to get a score that ends in 84, then i should be aiming to get a score that ends in either 84 or 34 during the section where i will be hitting holds to try and change the mod50 of my score.

II. Overview of song sections (How to Start)


The first step of any exact score is always going to be to pull up the notechart in order to analyze the structure of the song and figure out a plan of how to approach it. For songs with no notechart (360 songs?) it is still possible to figure it out on the fly, and sometimes what you decide on paper has to be tweaked when you start working on the song in practice, but just like making an optimum SP path its always going to save a lot of time and headache to work it out at least roughly in advance. In terms of 'analyze the structure of the song', the basic method is going to be to identify two crucial sections of the song, and thusly divide it into four parts. Unlike finding an optimum SP path (for which the 'result' is going to be an exact number and thus it is possible to formulate a concrete optimum strategy completely on paper), an exact scores 'path' results in a higher chance to get that score on a single run. For that reason, there is no one 'best' path to get an exact score, it is much more open to interpretation about the best places to draw the path to maximize your chances of getting a certain score.

Remember that the three principal challenges to exact scores are getting the correct mod50, getting the leftmost digits exactly correct, and doing those two things without failing the song. Leaving aside the third challenge for now, the other two are central to the two key song sections that we need to identify how to solve.

In order to get a correct mod50 what we need to identify is a section of the song (preferably towards the end) with a decent concentration of holds. I call this section the 'Holds Section'. What i will do is go into this section with some arbitrary mod50 (and a rough range of my score), and then i will attempt to hit holds in this section to get the correct mod50, and then I will not hit another hold for the rest of the song. Yep, what you do is let all of the holds pass you by (dont even strum the note). By this mechanism i can solve the issue of getting the proper mod50 without having to worry about going over the target score (you start and finish the holds section with enough breathing room and the plan to solve the left digits problem later), and while also not failing the song (you will be hitting most of the notes in the hold section and thus you should get mod50 with Rock Meter (RM) in the deep green (or yellow), which is usually plenty enough to ensure that while you miss the rest of the holds in the song you wont fail due to those misses. In a later chapter (III) i will go into MUCH more detail on how to pick out this section in advance, how to 'target' your dropped holds to maximize the chance of getting the mod50 you want on a minimum of holds, and the sort of problems you might encounter in this section.

Now that the problem of getting the proper mod50 is solved, we need to examine the song to determine how to 'close out' the song with the proper exact score. The best method that i have found for this is to use a section i call the 'SP Section'. It is called this because what i do is always save a full bar of SP and for the last 8 measures of the song in engage that SP and try and hit only enough notes to get the exact score i want. Thus the SP section is actually fairly easy to identify on the notechart, you just want to pick the last 8 measures of the song that have a high density of non-sustain notes (opportunity to gain RM). Remember we cant hit any held notes so we want to ignore sustains at the end of the song, after you get the exact score in the SP section you will let those just pass by. Based on my observations, you can plan to miss as few as 25% of the notes in this section (because of SP being up) and still not lose appreciable rock meter, while you can easily plan to hit up to about 75% or so and be able to pull the plug and stop hitting notes as soon as you get the right score. This leaves you pretty good flexibility in terms of what score you can go into it with, while still having no worries about losing all your effort by failing the song. I will go into exhaustive detail as to how to execute this maneuver in a later chapter (IV), including commenting on how to decide what the proper range of target incoming score should be (hit 50% of non-sustain notes under SP), as well as a lot of other miscellaneous tips and tricks.

Having determined two special sections of interest, the hold section towards the end of the song, and the SP section right at the end of the song, we have effectively divided out song into four sections, comprising the two sections i just mentioned as well as the bulk of the song from the start to the holds section, and the portion of the song between the holds section and the SP section. (Technically there is a possible fifth section which is any holds at the very end of the song after the SP section, but since the plan will always be to miss all of these holds this section is uninteresting and can be ignored).

The section between the holds and the SP section I dub the 'Miss section', which for lack of a better term reflects the fact that you will be missing a lot of hold notes in this portion (intentionally). In this section the goal is basically to bridge the distance between the holds section (where we got our mod50) and the SP section (where we will finish the song) without failing the song or going over the points target that we want going into the SP section (which is of course the only real reason we would fail the song). The goal here is simply to find a simple and robust 'note hitting rule' to hit enough non sustain notes so that we can reliably get through this section without having any chance of failing the song and while getting the same rough number of points in this section each time. The chief property of the Miss section is how many points it is worth, because once we know that (usually from practical experience but you can estimate it), we can make sure to go into that section within a rough bounds of score so that we can be confident of getting to the score range we want by the time the SP section starts (which will allow us to easily finish out the song under SP to turn that score range into an exact score). If i can reliably play the Miss section under the constraint of not hitting any sustains, knowing in advance within 1000 points or so how much that section will be worth, and not failing the song (or having any chance of failing), then i will have explicitly solved the third main challenge of exact scores which is simply not failing the song.

The manner of determining what score we go into the miss section with is basically just slight adjustments to the initial 'bulk of the song' section which i call the combo section since it is often the only section of the song where you will want to build your combo up to over 1x. Note that the score upon leaving the combo section predicts the score going into the miss section since more or less the holds section is worth roughly a similar number of points every time (which you also need to calculate, usually by practical experience). Discussion of the miss and combo sections will also take place in later chapters (V and VI).

III. 'Holds Section'


As covered above the holds section is the section of the song where you are going to try and get the correct mod50 for the score you are aiming at. If you fail to get the correct mod50 by the end of it then you are going to usually be restarting the song, therefore it is important to choose this section to maximize your chances of being able to succeed. Skill also plays a factor into the determination, as you get better at controlling hold points you can get more aggressive in size of your hold section, etc. You are going to go into the holds section with a score within a certain score range, and for the most part you should find that you leave the holds section without narrowing that range, but with the mod50 exactly where you want to be. Note that the technical end of the hold section (and start of the miss section) may vary each run, because in practice the hold section ends as soon as you get the right mod50, which might be on the first note of the hold or the last depending upon your luck and skill. In a way the hold section is not so much a hard and fast "section" as it is a window of opportunity, during which you had better get the mod50 otherwise it might be very difficult to get the exact score.

a. Considerations when choosing a holds section

Hold section should be towards the end of the song, but not at the end of it. This is obvious because you need enough space for at a minimum a SP section in order to even out any variations in your score. Its also a good idea to leave some space for a miss section where you can narrow down any score variations before engaging the SP section. However for some songs you want to put the hold section as close to the end of the song as you can because there are a glut of holds in the end of the song and otherwise the miss section is very hard to hold RM up in. On other songs the miss section is very large as in cases where there really arent many holds towards the end of the song (Mr. Fixit, LtR).

If your hold section is too small then you wont have much chance to get the exact right mod50 before leaving it and probably screwing up the song, while if the hold section you plan on is unnecessarily large, you might have a bigger variation in the score obtained through this section as sometimes you will get it right away and earn less points while other times you will be going for it for a long time and earning more points. In general it is hard to manage your overall score (thousands digit) during the holds section because you are too busy making estimates and such on the mod50. For that reason as you get better and better at getting the scores you want out of holds section in fewer number of holds, you will want to shrink this section down (always move the start of it to later to get more of the song into the bulk section) so that you can have a more accurate estimate of how much points you are going to score in this section.

If there are big BPM changes in your song, then its not a bad idea to try and pick a holds section with a lower BPM as that makes actually getting the holds a lot easier, but usually the changes arent enough to matter (inside a song), or they are in the wrong places (like you cant use the slow BPM section of Free Bird cause its only the first third of the song)

b. How to get the right mod 50

During the holds section you always want to hit holds at 1x multiplier, because that will allow the score to go up the slowest. Also aside from the hold notes you want to hit as many interspersed non hold notes as you can in order to keep your multiplier up since if you are hitting a lot of hold notes and dropping them right away your meter will not be going up and may even be losing value. For the typical song you want to hit a hold and then hit every note after that up to the last regular note before the next hold, which you miss (to ensure the combo resets to 1x), and then hit that hold, repeating the pattern. If the holds are one after another hit about 5-8 in a row, let one go past and then start again, or if the pattern is like one hold then one regular, do that about four times and then let the next regular note go by (kill at note streak 8 to reset combo to 1x), etc. In general its a bad idea to be having to look at the combo counter as you will be busy processing your current mod50, you just want to be able to instinctively always be letting a note go by if you are nearing a note streak of 10.

Also chords make your points go up a lot faster (but still in single point increments unlike having a 2x multi) so it is easier by far to get the right score using single notes, but generally its a good idea to use every possible sustain you can.

I used to do this totally at random (hitting hold notes and releasing) but with experience i got to where i could predict roughly the score coming off each hold to "aim" for the right score. The methodology here starts by realizing that a whole beat of a single sustain is 25 points (12.5 for a half beat which is marked by a thin line on fret bar). So the method is two-fold. Visually you can see how much of the hold will get you ~12 points so you want to do is get a few points below the target mod 50, and then try and keep adding as few points as possible with each hold until you get it exactly. BPM is a heavy influence here as on a slow BPM section you can easily go up by either 0 or 1 point every time (and be able to do it without overshooting even once almost every time) while on a fast song, the minimum you get might be 2-5 with each try. If you overshoot, then you will be slightly over your mod 50, which is the same as being really far under the mod50, and you can hit a bit less then 2 beats of the next hold (~45 points) or multiple holds to quickly get back to where you are almost at the right mod50 and can start trying to add in minimum increments again. Just repeat the process until you get the right mod50. For chords, its the same thing but keep in mind that a half beat is 25 points, and even at slow BPM its pretty tough to reproducibly add only 1 point, so you will be likely to overshoot a lot more.

c. Misc tips

Also understanding how hold points are rewarded relative to being early/late on a note is important. What happens is that when you hit a note early you will only collect hold points once the hold notes start passing in the fret bar. So there will be a slight delay before you get any points. Whereas when you hit a note late, you instantly collect any hold points you have already passed and the hold points on top of that also start accruing right away. Thats why the points for a hold note are always the same independent of whether you hit it early or late, by the way. In terms of our job what this means is that even if you are dropping the hold note as fast as possible, if you are hitting the note late you wont be able to score 0 points from it. So the best thing to do if you want to add a minimum (ideally a fully reproducible 1 every time) number of points is hit early and time your release to be right when the actual note goes by. Things to think about.

IV. 'SP Section'


The SP section is really the "make or break" section of the song from an exact scores perspective. Sure it can be tricky to get the right mod50 on some songs and you might end up restarting the song several times during the holds due to not getting it right (particularly when you are inexperienced), but there is nothing more frustrating than finally getting the correct mod50 after several tries and then screwing up by going one note over or under (or hitting too many notes too fast too soon and having to watch as your RM starts to fall as you try to run off the clock, only to fail on the last note). In Exact Proofs as in baseball having a masterful, reliable closer is crucial ingredient to a consistently winning record.

a. Basic SP section Strategy

Recalling what was said before, the basic input of the SP section is going to be a score that has the correct mod50, and is within a few thousand of the final score you want, but does not need to be exactly anything. Also you Rock Meter might be solidly in the green if everything has gone according to plan so far and the miss section of the song is easy, whereas it could be deep in the red if you went a bit higher than you wanted in the previous sections score-wise and have been struggling to keep that down (or if the miss section of the song is particularly brutal). Either way, the SP section is designed to take these variable inputs (score short of the desired by a certain amount but not precisely so, Rock Meter could be anywhere on the map) and be flexible enough to easily and comfortably produce a precise result (exact score, rock meter not failed) from these inputs.

The operating word of the SP section is flexibility. That means that I am going to give the anatomy of the SP section and I am going to talk about the "Ideal" run through that section, but keep in mind as i talk about that that the whole point of the SP section methodology is that it is robust enough to handle if you are far off this ideal, so i will then talk about what the limits in each direction are and way to help you seal the deal even when youve screwed up some and its a tricky challenge (in the misc tips section).

There are two considerations when looking at what the ideal play through a SP section is. Obviously the actual play on any given instance will be to hit exactly those notes you need, but keep in mind that we can control to an extent what score we go into the SP section with so we want to figure out the ideal play, so that we can 'aim' at that score through the other sections of the song. The first consideration is that i want to have a lot of flexibility up or down in terms of how many notes i can hit versus skip. The second consideration is that i want to, where possible be hitting 40-60% or so of the total notes of the song (counting holds because those hurt your RM) so as to have a definite net positive flow of RM so that no matter what RM state i go into it at i should easily avoid any chance of failing. in fact somewhere around 20-25% of the notes is the break even point so even if you are hitting only 25% of the notes you are probably going to stay fairly still (no net gain or loss) in RM which , tho not ideal is okay.

The first calculation to do when deciding what score to aim at for going into the SP section is to count the total number of non-hold points that are going to comfortably fall under your 8 bars of SP (you dont want to be like executing any squeezes or anything for obvious reasons). Say for example i have a SP section with 40 single notes (no chords and no holds). The base score for those 40 notes is going to be 2k (40X50). obviously in order to give myself the maximum flexibility id want to aim to hit 20 of those notes in the SP section, which would also be worth 2k (20x100). If i do that i will be hitting 50% of the notes under SP and even if i start the SP section in the flashing red i will quickly and easily be into solid green by the end of it (and the end of the song). Also note that i could probably easily get through this section only hitting 10 notes or even as many as 30 or 35 if i was very careful about my combo not getting too high and sending me past. So even though i am aiming to be at 2k under the score i want i really can do it extremely comfortably even if i come in anywhere from 1k to 3k short of the score. So basically, the mod 50 notwithstanding, my problem of getting an exact score by the end of the song has been simplified to getting within a 2k wide window by the start of the SP section of the song (Rock Meter state does not matter), which is hugely easier. If i have more or less base points in the SP section, then my window will be smaller or wider .

If you have hold notes in your SP section then you dont always want to pick half of the non-hold notes. That is because hold notes you always have to miss and so your actual portion of total notes hit might drop down below 20-25% which is the danger area where you may not have a net increase of RM. So for example say in my example above my SP section is still 40 non hold single notes, but i also have 40 hold notes (actually a very high proportion compared to most songs). So my base score of 2k points from all the non-hold notes is still the case, but i may not want to aim to hit exactly half of these, because if i do that then i will only be hitting 25% of the total notes in this section. And while that is certainly doable, if i aim for 2k under but i come in at 1k under, i wont be in the situation like before where i can easily hit only 10 notes to correct for that, because if i do that i will be only hitting 12% of the notes and i will SURELY fail. i could still go all the way up till 40 probably, especially if the hold notes are putting in automatic combo breaks that will enforce my combo wont go over 1x (which is not the case for a stream of 40 consecutive notes). So really the range of viable notes hit are 20 to 40 in this case rather than 10-30 like in the previous (no holds) case. So i should aim for 30 notes hit (3k under). Note that even though the lower end of my spectrum moved up, the upper one also moved up (due to the extra ease of controlled hitting lots of notes if there are holds in the way) and my actual window of flexibility is still sitting at 2k.

So in summary what you want to do for the SP section when you are preparing in advance is count the base score of all the non-hold notes in that section. Then you want to plan to hit half those notes (which coincidentally end up being the same base score since you are dividing and then multiplying by 2) which means going into it with a score that is less than the desired score by this base score that you just calculated. Nudge that number up a bit if you have a lot of held notes (or chords).

b. Misc tips

I always pause the game when i get to the start of the SP section. Then i look at my score, compare how far off the ideal i am and then calculate how many notes (+100 pts each) i will want to hit to close the gap (count chords as hitting 2). Then unpause, engage SP and count off those notes as you go (chords worth two). Thats handy because then you can in theory avoid having to look at the score which is distracting (but i always end up looking anyways). If its a very large number of notes, then just hit roughly half of the notes until you get to within 1k on the score, THEN pause and count how many more notes you need and then count it off as above.

You always have to factor chords into the calculation (even though ive largely ignored them). Remember that chords are going to give you 200 points per hit but the same rock meter boost, so the way to think about it is that chords give you more return in terms of points than RM. So in the ideal case youd probably plan to hit half the chords and half the non chords notes, but if you are high or low going into the SP section you'd want to change that. For example if your score is too high and you need to skip extra notes you'd better make sure you are hitting more singles if possible since you will get more RM bang for your buck and less likely to fail. If i am within 1000 points and im staring at 20 chords and 10 singles coming my way and a rock meter in the red, then i can probably pull it off if i hit the 10 singles, whereas if i only hit 5 chords thats a near-certain fail. On other hand if your score is too low then you want to power up on chords to help close that gap and you will be hitting so many notes anyways that it the less RM boost you get is a non-issue your worry is failing due to not getting up to the score you want (not from RM) in that case.

Also keep in mind that (in nearly all cases) its far better to hit the exact score you want in the SP section a bit early than a bit late. Because typically your rock meter will be solidly in the green (or at least yellow hopefully) at this point (end of SP section) so if you judge wrong and get the score a bit early its no big deal to just skip a few extra notes at the end once you get the final score, whereas if you judge poorly in the opposite direction then you are boned because there are no more notes left. You should always try and spread your notes out along the length of the SP section so that you dont waste a lot of green meter boosts (if you need to hit 20 of 60 notes and you hit them all in a row then you will fail the song because you are letting 40 consecutive notes go by whereas if you hit 1 of every three you will easily stay in full green the whole way). But you should always 'front-load' your SP section hits just a slight bit. So for example if i am aiming to ht 20 of 40 notes then i might want to hit 5 in a row at the start then hit the other 15 every other note, and then let the last 5 or whatever just pass by. If my ideal was 20 but i end up going into it short on score such that i need to hit 30, i might front load it a lot more and hit 9 of the first ten and nine of the second ten, and then 12 of the last 20. Whereas if i go into it too high and i actually only need 12 notes of the 40, then i might let the first 5-10 notes go by, and then hit every other note past that point until i hit 12.

Also I always say the SP section is a full bar of SP, but in some rare cases you cannot get that. This usually happens because you need to use some SP in the bulk of the song in order to even approach the score you want going into the holds section, which can happen on a 5* proof for a tough song (or 4* on a REALLY tough song) or if you are going for a 133,337 score and that happens to be 6*+. Keep in mind that in the holds and miss section you will be skipping any hold notes so you might miss some of the SP sections in the song due to that. Basically if you have less than a full bar then its not ideal but the method is the same except you have probably a SP section thats only 6 (or 4) measures long, which means a lot less base points, which means less of a flexibility window for the score you have to go into it with. If you cant get any SP for the SP section than thats REALLY tough because you have like no margin for error, because you really wont be able to vary your play through the end section of the song without taking huge RM hits. You pretty much have to enter that section in a near full green and then space out your notes you hit (take advantage of single versus chord notes) to try and keep your RM up. Where possible you always want a full bar or at worst a half bar.

One hidden (in plain sight) thing that i havent mentioned and may not be obvious is the mod50/mod100 difference that ends up cropping up as you start the SP section. Remember that SP section generally will cover all the way through the last (non-hold) notes of the song. This is a good idea for two reasons, because it means that you wont have a lot of extra notes that you want to miss at the end of the song, and you dont want to have SP run out right before you hit any notes so that you get 50 points instead of 100 and miss the exact score (happens more than you might think for badly done SP sections). Avoiding those two problems are huge advantages to doing it that way, but there is one slight drawback - once you engage the SP, you can no longer increment your score by 50, the minimum unit is now 100 (cause you are under SP). So even though i said earlier that for purposes of exact scores ending in 34 and 84 are the same, you have to be aware if you are aiming for 84 but you have 34 not to engage SP until you can get that right. The simplest thing to do is when you pause the game as i noted above ALWAYS check this, and then plan to unpause, hit one single note and THEN engage SP and proceed as above. Its not at all hard to do, but it is easy to silly miss and then you ruin all your effort. If you do make this mistake then stop 50 points short of the score you want and pray that your SP ends before the last notes of the song. You can always try to zero hold points hit the last note of the song because while its usually not worth messing with doing that because its hard to do and get 0, if you already blew the run then you have nothing left to lose.

You may need to worry about mod200 (or 300) if the song ends on a lot of chords, such that the last chance to change your score even by 100 is somewhere other than right towards the end of the SP section. Your last chance to change it by 100 might be hitting a chord before you engage the SP section (if there are literally no single notes in the SP section which is really rare), to it might just be that the last single note is like 4 measures away from the end of the song. Either way identify this 'last chance' note(usually you will find this out while playing cause you wont have noticed it in advance), pause in advance of it, and check and see if you need to hit it and then unpause and do so if needed and proceed normally.

One side note on this technique is that you can expand it to a more advanced form sometimes to expand the number of scores you can aim at for songs where it is really hard to get the mod50 you want. The way this works is you may happen to notice one (or more) easily identifiable holds that you can calculate the exact value of (for ex: 12 or 25 or something). Now you can define a short pseudo-section which is the transition section. The whole reason why i can aim for 34 in addition to the 84 that i want is that i know that before the SP section (when i 'lock' the right most two digits by making all notes worth 100 points minimum) i can always optionally hit 50 points to interchange those scores (and im going to do this while paused). However if i know that say a hold note worth 13 points is also right there anyways, now i could optionally easily (in theory) hit 0, 13, 50, or 63 points which means in theory if i want to end with 84 then i could go into this section with 4 possibilities (21, 34, 71, 84) rather than only 2 (34 and 84). you can get even more complicated than that if you want by making a combinatorial matrix of possible target holds, but keep in mind that in most cases its unneeded. Only in severe cases where the hold section is really screwing you over is it at all a good idea to do, because it is a lot of extra effort and two holds might look similar but be worth slightly diff points, etc etc). 95% of the time i dont go near this technique with a ten-foot pole.

V. 'Miss Section'


The 'Miss Section' is not as hard and fast a section so much as it just is the space between when you get the correct mod 50 (holds section) and when you close out the deal (SP section). It is dubbed the "Miss" section because you need to be sure to "miss" any holds that are in it (and therefore you wont be able to score very highly). The crucial thing that you want to do in this section is be able to play it in a roughly reproducible score (so that if you go into it within a certain score range you will come out of it with a known score range) and also to not fail the song in this section. You are going to use SP right at the end of this (when you start the SP section) so its okay to see your RM drop into the red (in most cases). One other note about the 'miss' section is that it doesnt have a precise starting point, because it 'starts' as soon as you get the right mod50 in the holds section which could be the first note of the hold section or it could be the last (or even after the usual hold section if you are going to try and use hold notes in the usual miss section to get that mod 50).

So the first thing to note about this is that in most cases you want to use a score for this section that naturally evolves from how you end up usually playing it, rather than 'predetermining' a score that you want and trying to figure out how to get it. The reason for that is that it is a lot easier to tweak your score in the opening section of the song, and then just leave the other parts of the songs where the scoring will be roughly fixed (and known in advance). Also, you dont want to waste a whole lot of time calculating on paper in advance what this section's points will be exactly - although it can be done that way more often than not its better to just play through the song once and 'play it by ear' and note down how much points you get. This will be covered more in Chapter VII of this guide.

The exact method for making your score reproducible through this section is very flexible and also very dependent upon the nature of the note chart in this section. Roughly you can divide the approach into two extreme cases: the case where the section is very dense with holds, and the case where it has no (or few) holds. In practice, most sections will be somewhere between these extremes and so you'll mix and match the advice for both cases as needed.

For Miss sections that are very dense with holds then the strategy is very straightforward (but lacking in flexibility - which means you will need to be that much more exact going into it). What you want to do is just hit every note that is not a hold. You will need to do this anyways probably in order to keep your RM up (to cover all of those holds you are missing), and it should be easy to do because the hold notes will act as 'spacers' which will let you hit all of the non hold notes. Also, you should pretty easily avoid any 2x and greater combos, since the missed hold notes will be breaking them, so that will make you score even more reproducible. Note that if your miss section has so many hold notes that you are tending to fail the song due to RM, then you need to push your hold section back and try and get your mod 50 later in the song. Also such a case where you know you will be hitting every note you can and under 1x its very easy to calculate in advance what the score for this section will be, but keep in mind that due to the variable starting of this part it will never be as exact as you calculate. One good thing to consider is getting and empirical value for the total number of points you gain between starting the hold section and finishing the miss section, that will actually tend to be fairly constant, even with varying points at when you get the mod50. Thats because the play strategy in the holds section is going to be fairly similar to the play strategy in the miss section except in the hold section you miss every note right before a hold, while in miss section you miss every hold proper.

In the opposite case, when the miss section has few or no holds then you have the reverse sort of problems. You have great flexibility in terms of the score you could reliably get in this section (FCing it versus hitting the notes at 1x), but execution of that strategy might be more difficult (it will be a lot easier to start at the right point and then finish too high or too low due to variances in play). Obviously, the base reproducible play is going to be to FC that entire section (aside from holds). Although this will give us a very exact score, it is not very robust as if we make any mistake in the section (and for example break a 4x combo) then that will be an instant loss of 3k or more points. Unless you are sure that you can FC it nearly every time you shouldnt plan to do that strategy. Another way to make your play reproducible would be to plan to miss certain notes, however this method is bad because its no fun trying to remember exact places to do things like drop combos, and you have the same problem if you flub a section thats not one that you planned you will change your note streak.

A better solution to getting a reproducible play-through is to use the combo meter to your advantage to modulate how you play. In other words, i will hit all of the notes except i will break my note streak at a reproducible point. The simplest play rule for this would be to break combo every time i see a 2x. The advantage here is that even if i make a few "extra" mistakes im not losing more than the note value of the single note i miss since the fact that i break combo is irrelevant. Also, you can plan to break combo every 3x or every 4x, which will give you higher scores for that section, but may be more sensitive to play variations. I find that 75% of the time an enforced 1x multiplier is the best way to go. If you are aiming for a high exact score, than it may be the case that you simply cannot easily get your score high enough in the main part of the song where the amount of points that a 1x multiplier through the miss section will just not be enough to close the gap and you need to plan for a higher score. Alternately, sometimes i find for very difficult sections, that i am more reliable if i try and hit all the notes than if i intentionally miss some (because that will throw me off my concentration and ill miss a lot, potentially endangering my RM). Also, if you plan for 1x and then on a specific play through find that you are too low you can go to higher multipliers in an effort to close the gap so that you can get to where you want to be by the end of the miss section.

Note that most miss sections will be a mix of sections where hold notes enforce you into a 1x multiplier, and sections with notestreaks that in theory would let you get higher than 1x, and what you want to do is a straight mix/match. IE plan to hit all of the notes in sections where you cannot get a 2x anyways due to holds, and then if there are larger note streaks, break at 2x during those. This mixed case is actually better than either of the extreme cases.

One last note: watch out for HOPO'able hold notes. These can often happen at the end of a run of HOPOs and your hand can hit the note on autopilot even as your strum hand is very clearly receiving the message from the brain to miss the note by not strumming it. Especially nasty is the HOPO'able hold note with the same note (non-hold) right after it - you consciously do not strum the note, and leave your fret finger off the note so you dont accidentally trigger it, but then when you put your finger on the fret after letting the hold note go by to get ready for the next one you are still within the HOPO window for the hold note and so it goes off. The first several times this happened to me i couldnt figure out why when i got to the SP section my mod50 had changed, when i was CERTAIN i hadnt hit any holds. (Yet another way that Red Lottery is evil!)

VI. 'Combo Section' (song bulk)


The Combo Section (aka the bulk of the song) is the last section i will talk about, which is odd because it is of course the first part of the song. The reason for this is that of course what you want to do in the combo section depends completely on how many points you are going to tend to get out of the other two. The ONLY goal of the combo section is going to be to get to the hold section fairly close to a target score, that you probably have worked out in advance based on empirical runs through the song. So say for example i want an exact score of 150,000. If i have worked out that my hold+miss section (i report it this way due to the variation in dividing line between these) is going to be worth ~10k, and my SP section should be able to easily get me 1k-4k (decent back end flexibility), then my target for the combo section is going to be 136k-139k. If i am sufficiently skilled then i will on nearly every run be able to extract the mod 50 that i want out of the song in the holds section and play it and the miss section to 146k-149k and then as long as i keep my cool i should 100% of the time be able to finish it out in the SP section at the score i want, and basically my problem of how do i score 150,000 exactly on the song has been reduced to merely scoring 136-139k on the song up to a certain point, and then engaging in a rote method to convert such a score to the exact score i want. Obviously, this is a MUCH easier task.

While im sure it would be possible to go over the notecharts in advance and pick out exactly how many times to drop a combo, etc, etc in order to get my score to the level i want by the start of the holds section without ever picking up a controller, this method is frankly to me overly time consuming. What i do is usually work out this part almost entirely 'on the fly'. I will in many cases need at least one or two runs to get an idea of how much the hold/miss sections are worth and how easy the SP section is to execute, and i can also use those runs to fine tune some basic guesses about how to play the combo section. Ill go over things like practice runs and how to extract the most info out of them in 'Chapter VII' (Putting it Together).

Roughly, the combo section can be divided into two cases: where the song is easy enough relative to your skill level that you can FC or near FC it anyways, and where its one that you generally have trouble with (maybe your best score is <5*). One way to think of a combo section of a song is like a very very large 'miss' section, except where you dont have to worry about dropping hold notes. IE the most points i can get out of that song (without using any SP cause i want to save that for later - plus i dont want the extra score variation of activation SP in varying places) is going to be if i FC it, while the minimum points i can get out of it is if i break combo at 1x by missing every 10th note (actually you can go lower by missing even more notes but you rarely ever would want to go that low, and your RM would get to be a problem). If you can near-FC a song then there are few exact scores that are outside of your range, while if you are worse on the song then you will notice a lower ceiling as you will not be able to take the combo section up as high as is the theoretical maximum.

The key to being pretty good at getting the score you want (+/- a few k) out of the combo section is properly using "landmarks" to figure out how to keep your score reproducibly below the upper limit. One useful thing to keep in mind is that if you are always getting back to a 4x multiplier after you drop one, then the amount of points you lose (relative to a no SP FC) is actually going to be fairly constant (assuming the song is fairly even in distribution of reg single notes and chords). The exact manner of how many drops you need to make is going to be highly dependent on the ultimate score you are aiming at, as well as the nature of the song. Ill just list a bunch of possible 'landmarks' ive used in the past (as used by me on various songs).

If you need to bleed off a bunch of points dropping as soon as you get a 4x can be very helpful. his will play the combo section at an average multiplier of 2x and is probably the minimum youd ever need. Another good thing to do if you need some drops (but not as many as after every 29 notestreak) and you want a good landmark to make them really reproducible is to drop after SP phrases. Collect the SP phrase and then drop the next note, then FC up to and including the next phrase and drop again. Depending upon how much you want to keep your score down you can do that for all phrases or just some. Or you can use song lyrics, or big changes in the note pattern (like entering the chorus, or starting a verse) and always plan to drop at some of those.

If you have a song that for the most part you can FC, but there are a few parts that you often miss, then use those. Otherwise your score might vary a lot based upon how many of those 'tough spots' you hit (and as you play more and get better you will have your score go up if you arent controlling for that). So just play like regular and if the first time you play you drop in three tough spots and you get a pretty good score, then make sure you 'plan' to miss those spots on subsequent plays (my fav way to do this is always try and hit them, and then if i do, drop the note right after to simulate the 'miss' - that way you get the feeling of accomplishment of improving on a tough section). For a sufficiently hard song you might be dropping so much that the random variations will even themselves out and youll get a reproducible play through in spite of yourself in which case you are golden - unless you need to improve your score by a big amount to even go into the hold section with enough points to possibly get the score you want, in which case you are in for a tough road.

Remember that if you need a near FC but not quite that you should always back load your drops. Say i know that if i FC the combo section except for 2 drops which can be roughly anywhere in the song. What i should not do is plan to drop twice the first two times i get 4x in the song, as then i have no margin for error - if i screw up the FC then i am going to be playing catch-up. Rather i should plan to play a pretty long distance with an FC and then put my drops not too far in front of the hold section. That way if i accidentally screw up during the part im planning to FC, then i just count one of the drops as already done and only drop one of the 'planned' drops. Also, be aware of when the hold section is coming up and what you can always do is get a near FC up to a certain score a bit below what you want and fairly close to the holds section and then immediately start dropping at every 2x multiplier (or even after every 4 notes hit if you really need to keep score down) and your score growth will slow way down (and if you do it right you should stay solidly in green) and hopefully come into the hold section about where you want to be.

Note that earlier i said the max points is an FC with no SP, but in practice remember that you can make this even higher by planning on one or more early activations of SP, which you should do if you are going for a very high exact score and you need to do it in order to have the score you want at the hold section you pick out. Its actually fairly common if you are aiming for a 5* or higher score to potentially need to do this, becoming especially more likely the higher proportion of the song falls at or after the hold section. Make sure to leave enough SP to try and get a full bar where possible, keeping in mind that in the hold sections and onwards you may not be able to get all SP phrases because they may contain holds after you already settled on a mod50. Even if you arent going for a terribly high score if you are playing a tough song you may need to do this anyways if you are having problems getting to the score you want at the holds section otherwise. Note that in some cases you still cant get the score you want even with using the best SP, if thats the case then you either need to get better at the song, or you need to go back to the drawing board and back your hold section up so that you can get more notes into the combo section and thus more points (this obviously has the effect of making the back end of your play less of a 'sure thing')

VII. Notes on Practical Execution of the Method (Putting it All Together)


Okay ideally this section is going to have no new data in it, the goal of this section is just to go over all of the above steps in the order you would reach them if you were sitting down to get an exact score. Also i will consider the picture in terms of making multiple runs at a song and how to get an exact score in the least number of tries, whereas up until now ive been usually talking about things only in the context of a particular run at a song. I am going to discuss this in the context of an example (which will be the next proof that i do).

The first step is to know the score you are going to be aiming at. For the purposes of my example here, lets go with the 5* cutoff of Search and Destroy (expert). That score is 197,330. The next step is to note the rough decimal star rating of that score which will help you in terms of deciding how to approach playing the song. In the case of my example this is obviously 5.0 (since its the cutoff), but say i was going for the 133,337 score, thats actually a bit below the 4* score of 140k.

Now i need to bring up the notechart (for X-SnD its here) and enter the planning phase. How in depth you plan is up to you. Ill go over the planning maximum in depth, and you can subtract from it as you see fit. These days im lazy and i just barely glance at the notechart i dont crunch all of the numbers like im going to do here. However these days i almost never get it on the first try - i need one or two calibration runs. And when i change difficulties (like ive been proving X then H then M) i usually dont even get up from my PS2 to check the differences in the notecharts.

So the first step of the planning is to decide on a hold section and a SP section. For SnD the hold section that jumps out at me is the series of holds that start at measure 99 and last until about measure 114 or so). Theres a lot of holds there and at a ~150BPM im fairly sure i could get the right mod50 in that range, and if not theres still a lot of holds later in the song to try and get it if i miss it.

The SP section is extremely obvious - from measure 134 to 137 there are a whole bunch of dense non-hold notes - i had better be hitting a decent number of those or im going to fail the song. However, since i will have 8 measures of SP (that section i just mentioned is only 4 measures), lets expand it a little bit. In this case im going to go for the easy landmark which happens to be the last SP phrase (m130). notice that if i activate right after that then ill have SP all the way up until the end of that section with it running out only before the final held chord of the song. This spot also is good because it has a few single notes right at the start of it, so i can hit one 50 point single note if i need to in case my score happens to be ending in 80 rather than 30 at this point.

Having picked my sections lets crunch the numbers, SP section first. measures 134 to 137 have no hold notes so i can fin the base score of that just from the chart itself (800+1200+1000+800 = 3.8k). Then before that point i count up the number of notes (chords count two) and i find 11, or another 550 points - make sure not to forget those blue notes at the end of the RB chords - those count too. Finally theres an extra 100 points from a single chord in measure 138. Overall base non-hold note SP section value is 4450. Theres not many holds in this section, and so long as i manage to get to measure 134 without failing there will be no issue of not having enough target notes to hit, so i estimate that ill want to hit about 30-50% of the notes overall for a good play. For sake of an estimate, lets plan on 40%, so the amount im going to on average want to go into the SP section on is going to be 4450*2*0.4 (the total value of all the notes at 2x times 40%) = 3560 which we will round off to the even number of 3.5k. I will note to myself that theres enough points that probably i could do it even at less than 3k, and even a lot higher - even as much as 6 or 8k (since i could build up to 4x multi for some of that long stretch of notes in m134-137) although that section is a bit fast so it might be hard to carefully control the number of points i hit if i have to make up too many.

Filing that 3,5k number away, lets now count up the miss and hold sections. Ill just do the same thing, count every single non hold note, and chords as 2 , and multiply by 50. I wont bore you with the details but i counted up 220 such notes, so 220*50 = 11k. Now analyzing the note distribution, there is nowhere that there are enough hold notes to really screw me up in terms of failing, but there are two stretches with long combos that i will have to break combo on (m115-118 and m122-123). The good news is that 11k is thus a low estimate as if i combo those four note scales in 115-118 ill be potentially able to get a LOT more points. Ill lose a few points dropping some of the notes in the streak sections, but ill gain some trying to hit the holds at the start so overall i think 11k is a good estimate for this section.

So taking my two estimates together, we see that i want to go into the holds section roughly 14.5k below my cutoff (11k + 3.5k). Also note that in both cases i could theoretically make more points if i had to, so that is a soft bottom, in theory its still doable even if im further below that - ill just have to remember to combo as much of those two streaks in the miss section as i can. My cutoff is 197,330 so we will subtract 14.5k from that and get 183k (note i round off as it is an estimate).

Looking at the main section of the song we know that we are shooting for a 5* score and in this particular case my hold section is rather early in the song (measure 100 of a total of 140 measures) so im going to need to play the crap out of the opening bit of the song to hope to get up there. Luckily i know that i can pretty much FC most of the repetitive opening of SnD - where i usually screw it up is in the ending anyways. Looking at SP, i now want to check and make sure that ill easily be able to get a full bar of SP. There is not a HUGE amount of SP in the combo section - only 5 phrases - luckily they are all fairly easy and i shouldnt miss any of them. after the hold section starts there are another 3 phrases, however the first two involve holds and come at the end of the holds section and later so i should hopefully have my mod50 by then and have to pass those up. The third one is gettable (no holds and short and easy), but it occurs right before my SP section so that could in theory be a hassle.

Overall i can tell that im going to need a lot of points in the opening section, so im going to plan to use SP after i get the first two phrases (probably in the chord-rich sections between m23 and m37), and then hope i can hit the other 3 phrases and the one at the end to get my full bar back. If i still cant get enough points going into the hold section then ill have to see if i can get more points out of the miss section in those two long streaks, and if even after that i cant get enough points ill need to not start the holds section so early and instead hit some of that stuff at 4x. But all of those issues i can figure out on the first few playthrough (hopefully), i am not going to try and really calculate that sort of thing in advance (cause it would take longer than playing the song once anyways).

Okay so I said i would faithfully report how this went and being honest there were a few snags on the first run through. , but i wont whitewash those as they might be instructive to others. Anyways the first step when you finish the planning and go to sit down on the song is to take note of any crucial landmarks you need to remember. In this case those are the part where i want to activate SP, the start of the holds section, and the start of the SP section. So i want to hold into my head as i sit down that the score im aiming at is 197,730, that i want to use the SP after collecting 2 then collect 3 more and 1 at the end, that the hold notes section starts a the YB holds after the string of RB chords the second time it does that in the song (later i can try and associate lyrics with that maybe), and that the SP section is after the last SP phrase which will be easy to remember because its the only one in the miss section that has no holds in it. Normally its not a big deal to note this stuff, cause you will plan and then play it right away, but in this case i had a two day pause between planning and playing because of writing this FAQ.

Anyways I sit down to play and its my first song of the day so im a bit cold at the beginning, and i make about 4-5 silly drops in the first 15 or so measures of the song, including the first SP. So theres no longer enough SP to have a full bar at the end if i activate after 2, so i quickly decide, well might as well test the hypothesis that i actually even need an activation - maybe ill be in an okay spot playing it with no SP activation. I then proceed to FC the rest of the song up to the holds section, and i have a score of about 178k. So my plan here is just to hit the first few notes at 4k in order to get the score i want - i stop and start hitting things at 1x when my score is 179368. Careful observers will note that this is well short of the 183k that my plan was - i actually screwed up and thought here that i was aiming at 180k for some unknown reason. I paused the game and write down 179368 so i can see how much the hold+miss section is worth. now very quickly i get the proper mod50, when and i stop hitting holds when my score gets to be xxx,x70. Note that this was actually wrong, for a score ending in 30 i want a score ending in 80, not 70, but i didnt realize this right away. The pressure of doing it for an FAQ apparently got to me - like i said its not pretty. Anyways i continue playing the miss section, during the big scales what i try and do is hit two set of four and then the first half of the next (10 notes and 2 missed) and then repeat. I get through that and the other long string fine and get through the last SP of the song and pause. My score right now is 189,620. Oh fuck this is where i realize that i stopped on the wrong mod50 what a rookie mistake. But my estimate of the miss section was pretty good, i got 10,200 on this run which is close to 11k, but then i am like wait why and i not at 194k, and i realize then that i also fucked up by starting at ~180k and not ~183k. GUH. So rather than just quit there i engage SP and play out the song trying roughly to hit about half the notes. I get 4.5k points so if i finish close to 3-5k i should be fine assuming the next run i dont screw up so badly. Note one other thing takeaway from this run is that if i FC the combo section with just a few drops it looks like im in a fine spot on score at the holds section so i can forget my plan of using SP after collecting 2 at the start of the song.

Okay so second run. I FC both of the opening SP phrases but i dont activate cause im pretty sure i dont need the points. In the rest of the combo section i FC except for two drops which were in the middle somewhere, which is a fairly average run for that song, and i finish into the holds section at 184,238. Note that in the measures right before i got to the hold section i was watching my score to see how close it was getting to 183k and i purposefully missed the last 5-6 RB chords to keep my score from overshooting. Overall im 1k over where i wanted to be but thats cool cause remember in the miss section last time i only got 10,200 so im WELL within the window of doability. So in the hold section i overshot 3-4 times or so but i still got the right mod50 before even the first of the three end of the song SP phrases, so plenty of time. My score at this point was 185,780. I proceed to close out the miss section and at the start of the SP section im at 193,830. I pause the game, note that i have the correct mod100 (so i dont need to hit any more single notes, i can activate SP right away). Then i unpause and activate SP and collect the 4 trailing B notes (at the end of the RB holds), and pause the game again. I pause the game because i know that this end section is going to throw a lot of notes at me pretty fast, so i want to quickly plan a strat. My score at this point is 194,230. I remember that theres a lot of chords in a row in the SP section so my quickie plan is to hit several chords then pause again and get the last 1k after that. I quickly calculate in my head that 15 chords should be about right. NOTE that actually this is wrong: 15 chords would be 3k which would put me too close to the exact score to be safe, i should have planned for hitting 10-12 chords then pause. Anyways i unpause with this wrong number in my head, and start hitting chords. I notice that my score is going up fast after about 13 or 14 chords and i go 'WTF' because it hit 197k. I pause right away and realize that 15 was the wrong number, it should have been 10. Luckily my score at this point is 197,130 so actually im still okay. My RM is in high yellow, but i still have a lot of notes to go by. In this case what i want to do to give myself the best chance of not failing is plan to get that last 200 points from 2 single notes rather than 1 chord, because that will give me better RM boost. But i need to watch out cause i remember that the last notes of the song are chords and i dont want to hit one single note and then pass the last single note without realizing it. Right now im basically in the middle of the second set of RY chords right about where the two single HOPO notes are. So i unpause with the plan of hitting two notes in the G-R-Y-R-G scales bit. I unpause and i hold down just the red button and strum two of the 4 red notes that are going by. I do it this way because the single notes here are all HOPOs and i need to be very careful to not accidentally hit an extra one. The best way to do it is to plan to hit two of the same note. Anyways i do this and i let the rest of the song go by and (luckily) i pass it tho i got into the flashing red so it was a bit close.

Now i go to grab my camera and i notice holy shit my camera isnt anywhere. turns out my wife had it in her purse from when she had been out taking pictures yesterday. I had forgotten the crucial unspoken rule which is 'always check that you have your camera, and it is charged up before sitting down to do proofs'. Anyways since i dont have a camera right now what i am going to do is unplug my controller (so i dont accidentally bump it off the newspaper screen, and leave my PS2 on until my camera comes home. Unfortunately that means i dont get to do the other leg of the proof right away.

Thats messy but its basically a good example of how to play it 'by ear'. As a result of those two plays you can see that my basic path is solid, and if i happen to break combo more than a couple times early on i should always just play to about 184k under 4x in the start of the hold section. Unless i screw up the mod50 on the next play or make anymore bonehead calculation mistakes it should be a three play proof (one practice, one for each cutoff).

VIII. Notes Specific to Obtaining Cutoff Proofs


Cutoff Proofs are slightly different than just doing a single exact score, for two main reasons: (i) you always need two pictures as opposed to one, and (ii) sometimes you dont know at the start precisely what score you want. So the first step is to minimize the impact of (ii) by always trying to get good estimates if you can. Riff did some estimates for PS2 and those were invaluable to me in my first run through hard and most of expert. As far as PS2 GH2 is concerned since at least one proof is done on every song the estimates on the song stats pages in red are exact. I dont know if there are any estimates for the 360, although i do think that for most songs the base score is the same (even ones with SP path only changes). Hopefully, when GH2-80s comes out somebody will rip that open and come up with estimates pretty fast on that, as well.

The nest thing thats special about cutoff proofs is that a star rating gives a somewhat normalized snapshot of the song. So if you are aiming at 4* you will always be aiming at double the base score of the song, while 5* will always be 2.8 times that (on GH2 at least). So this can give you a big head start on how to approach getting the right score in the combos section. In general a 4* is a overall 2x multiplier so one good trick for 4* scores is to play most of the song breaking combo every time you get a 4x - this will cause you to on the average have a 2x multiplier through these sections. Then a bit before your hold section FC it in order to make up the difference to where you want to be (because keep in mind that those points in the hold and miss sections you will be getting at roughly 1x multiplier). Or if the middle of the song is too tough, hold an FC in the early part of the song up to certain measure/score - and then play from there to the start of the hold section breaking at every 4x. For 5* scores a good baseline play is to try and FC the whole song with no SP up to the holds section. Usually this would actually end you up a bit higher than you want so either figure that you will break the FC at a few random points due to the song being tough, or be prepared to if you get too close to the target score going into the holds section to start breaking on every 10th note in order to slow your score growth down while hopefully not messing up your RM too much. For 3* scores - ie scores near the base score(which you obviously dont need to do for cutoffs, but it might come up as a 133,337 score), you do it like 4*, except break combo at every 2x multiplier. And then if you are short, FC a section right before the holds section. Thats how i did free bird 133,337 yesterday, which is basically a 3.1 star score.

So assuming that you know the cutoff you want exactly, then the only real difference from aiming at a single exact score is that you are aiming at two. So the first time through you have two mod50s that are right nest to each other and you can stop hold section if you get either score. Then once you get that one then you have reduced it to an exact score problem again. Make sure to remember which one youve gotten. A couple of times ive gotten one half of the cutoff and then gotten interrupted or something and then i go back to it and i get that half of the proof again and move on to the next song and i dont notice it until i go to enter the scores into SH proof system. Crazy.

Okay so when you dont know the cutoff, its going to necessarily take a lot more plays to get the proof. Luckily, the hard part in many cases is getting a reliable playthrough and the last score, so the scores in the middle you can hopefully rattle off one after another. So if you only know a range for the cutoff then you want to write that down, the high and the low. Note that a lot of times if you open up the SH database on 'all scores' for that song you can skim down it and you will see a much smaller range between the highest score below the cutoff and the lowest score above the cutoff than on the song stats page. Thats because these score probably do not have pics so they can't be used for the 'official' cutoff range, but theres really no problem with using them as initial estimates of range. Anyways what you want to do is take your range and aim to get into the middle of it on each run (so that you get the maximum possible narrowing).

One thing to keep in mind in this case is that when your range is really huge, then you have many many mod50 cores you can aim at, and you should use this to your advantage so that you never leave the hold section with an unusable mod50. If your range is greater than about 100 or 200 then you hardly need to worry at all about mod50 scores. in my experience its still a good idea to play it just like if you need a mod50 score, just dont be too picky, if say you are shooting between 100,400 and 100,700 then you want to aim roughly at 100,550 but really any score ending in 40-60 (or 90-10) is okay to stop on in the holds section. You should still play with the miss section and the SP section, IMO, because it will be good practice to get you a reliable play through which you will definitely need for the final few pics. Once you get your range down to 100 or 50 or so then you still have a lot of leeway for mod 50, but you still want to aim for the middle and the few scores around it. One good rule of thumb is that you should aim within about 40-60% of the middle of the range. So for example if ive gotten my range down to somewhere between 40 and 80, then i want to aim for roughly 60, and id probably stop as soon as i say anything in the 56-64 range (or 6-14). Once you get down to lower then you want to be more precise, so if i have it at 45-54, then ill aim at 49 say but stop if i saw 48 or 50 too. If you are close to being at the end of the holds section than take anything you can get thats in the range, the key is that you want every play to give you a score that narrows your range. Once you get down to where there are only 4 possible scores or less (ex if my range is 45-50) that would narrow your range, then they are actually all equally likely to be a cutoff proof (i think, calculation not shown), and you should take any four of those that you see.

This process is a lot easier once you have a reliable playthrough as you might need to get 10 scores or more to narrow the cutoff to where its proven, and if you are getting a useable score on every play versus on half the time that extra time REALLY adds up.

One other thing to note is that 4* cutoffs are always EVEN. Thats because 4* is always double the base score, which is itself always an integer. And mathematically speaking 2*n is always a multiple of 2, in other words even. So what this means is that the 3* leg of your 4* proof will always be an odd number and the 4* leg of your proof will always be even. So if you are narrowing, and you get a 3* score thats even or a 4* score thats odd, you don't need to snap a picture of it, because it cant possibly be a leg of the eventual proof. You should, of course, still note it in terms of narrowing your ranges. 5* cutoffs can be either odd or even (since they are the rounded up result of 2.8*n) so this does not apply. Generally its a very good idea to snap a pic of every score you get that could possibly be the proof, since it sucks to have proven the proof to yourself but you didnt snap one of the pics because at the time the range was still 200 points or something (it happens).

As far as proof submissions for contribution points, i used a sort of self-imposed honour system when i turned mine in while doing my initial runs through hard and expert, and I think JC can probably have the final word on this but heres what i did and what i propose be the universal honour code (to prevent rampant pointswhoring). (Actually this has less relevance now that the main page shows cutoffs proven rather than merely contribution points). It didnt seem totally right at the time that if i got one proof in two pics and three plays (because i had a perfect estimate going into it) that i would get the same number of contribution points as if i spent like 20 plays narrowing a very wide cutoff but i only submit the final two pics because it felt that the time investment was so much more. One the other hand it felt like points farming to turn in a whole bunch of scores that i knew werent the final proof when i was sitting on 15 pics that were all lower than the range in the system but also not the final proof which i had gotten the two pics of also. The compromise that i did, was that id always look for the estimate where i could and try and get it in two pics only, but in cases where the estimate was wrong then at the end once i had the proof and all pics that i took on the way to the proof id look at the path i took for it and if i had about 4-5 submittable pics that were actual steps in me narrowing down the ranges (incl the proof pics) id turn all of those in, and if i had more i would pare it down by taking about half or so of the submittable pics and turning those in up to a total of about 7. I never really turned in more than 7 cause it felt too much like points farming at that point. I never turned in scores that i would have known at the time i got them could not have been the proof - ie it was an odd 4* or even 3* or it was a screw-up and outside of my current range (in fact i generally didnt even snap pics of those). for the most part i had good estimates about 5-75% of the time so i was turning in only two pics, but on those ones where there were bad estimates i had to put in significant extra time proving them and so i turned in on average 4 or 5 for those. Note that the way the system is set up now in terms of points, the best way to pointswhore is just to do a bunch of narrowing of lots of cutoffs in the DB without ever proving any, but the thing about that is that it doesnt help the community all that much because it doesnt unlock any of the extended scoring for the community, and also because it doesnt save very much time for the eventual person who has to do the proof, since if they are following my method and have a reliable outside estimate then the cutoffs that are there are irrelevant, they will do it in the first two plays after they get a reliable path through the song, and even if they are going from the ranges because there is no estimate, then the time it takes increases only logarithmically with respect to size of cutoff range (ie since you can cut the range in half on every play through, most of the time is spent getting the last several scores anyways which occur when the range is small).

IX. Trivia

Songs with the 'toughest'* individual sections, only including songs that i have had a change of doing (ie not Jordan)

Combo Section: Free Bird (by a mile)
Hold Section: Hangar 18 (Hon. Mention: Trogdor, Yes We Can)
Miss Section: Jessica (Hon. Mention: FTK)
SP Section: Laughtrack (by a mile)
Barfo's Most Impressive Accomplishment: 133,337 / FC score on KitN on Easy.
Strangest SP Section: X-stream

*'toughest'=most likely to cause song failure (or overrun) even by the optimum path, or takes the longest number of runs to work out (combo section category).

Hangar 18 on X is the only song that i gave up purely because i could not hit the hold section right (i could get the combo section that i wanted and i could close out the SP section - had i ever had the right mod 50). All the rest of the songs that i gave up on as too hard it was purely because i could not get my score reliably enough to the point i needed going into the hold section.

Laughtrack ends on about 50 or so hold notes (or it seems like it). Its the only song where my path in the SP section necessarily involves hitting hold notes. And laughtrack path is incredibly difficult to execute (i pause the game and break out a pencil and paper to plan exactly which holds to hit based on my score going into this section, each time).

X-stream, i engage the SP section about 20 measures before the end of the song, and then hit a predetermined pattern of the holds at the end (whose value i have calculated in advance) to avoid (barely) failing the song.

Source


Information originally compiled and published by:
User_Barfo Barfo for ScoreHero on 26 Apr 2007
Source - www.scorehero.com


Category: CategoryGuides
 Comments [Hide comments/form]
Valid XHTML 1.0 Transitional :: Valid CSS :: Powered by WikkaWiki
Page was generated in 0.0205 seconds