"The best tests around"
Jan 10, 2013 19:20:16 GMT -5
Post by OrochiGeese on Jan 10, 2013 19:20:16 GMT -5
"The best tests around: Nothing will be able to keep your edit down."
-A Fire Pro Returns (FPR) guide by OrochiGeese, the living ghost of Jason Blackhart, and others.
TabLAWL of Contents:
I. Greetings and salutations!
II. Engine Fundamentals.
III. The apple of knowledge in the Garden of Edet.
IV. “I'm thinking of a COM level between 1 and 10.”
V. How to randomly generate utility out of the RNG.
VI. “Fly the friendly sims”
VII. My strategies for testing an edit.
VIII. The League of Extraordinary COMskips.
IX. Input from other community members.
X. Credits.
I. Greetings and salutations!
You are reading this guide for one of several reasons:
1) You are having trouble improving an edit you created.
2) You are comfortable with the testing process but want to consider other ideas.
3) You click on any link you see and ended up here. Sit back and have some chips.
4) You blacked out and don't remember. Seek medical help then prepare to fight zombies.
Chances are that if you are here for reason 1, you may be somewhat new to edit making.
Welcome! This is an exciting process that takes some time to master but is extremely gratifying when you finally do. Notice I say "master" rather then "perfect". That is because edit making is an art rather then a science. “Perfect edits” are not possible to make in Fire Pro R for a number of reasons. And just to clarify, we are talking about "CPU simming" here rather then playing (1P) as your edits. It is entirely possible that you make an edit that you consider perfect for your personal usage against CPU opponents. But this guide is only directed at those who use do "CPU vs. CPU" matches where the game uses the CPU to simulate your edits against others.
This guide is also being used to compile a lot of Fire Pro “engine” info from around the community into one place. I do this because I think it is useful to have that “one stop shopping” (especially as many of the Fire Pro community forums with information have disappeared over the years) but also because it is necessary to understand how the engine works to learn how you can maximize your time testing and simming.
I'm not going to be giving you overall logic or strategic advice here as I've been there and wrote that. And so have many others. However, you won't be able to fully grasp my testing suggestions in the later sections of this guide without understanding how the fundamental nuts and bolts of the simming engine actually works.
If you already know about the "Random Number Generator" feel free to skip to Part III or V of this guide. But if you have no clue what the RNG means in the context of FPR, put on your learning hat and turn it sideways to activate.
II. Engine Fundamentals
Some advance warning: I am setting up a point that pays off with number 3) below. I actually structured this section similarly to how the community grew to understood these concepts over time. It's not until 3) where the actual engine mechanics are revealed after some of the misconceptions in 1) and 2) are corrected. That is not to say that every thing I say in 1) and 2) is wrong though but 3) will clarify both sections. I believe it's necessary to start off with what seems obvious in Fire Pro in order to go further into the less obvious territory.
1) Fire Pro R's “CPU logic screen” is based on setting a system of percentages to produce what you think will be a good progression of moves throughout a match. You want a chop to come out more then any other move in front grapple, then set it much higher then every other move in front grapple. It looks simple but it isn't always easy to master. More importantly, it's not entirely representative of what actually happens when simming matches.
2) Fire Pro R's CPU simming engine appears to be based entirely on the above system of CPU logic percentages rather then any fixed number determination. After all, what are percentages in reality? They allow you to guess in advance that there's a 50% chance that a “heads” will come up any given time you flip a coin. But percentages do not give you the ability to know the actual sequence the coin toss results will be in (say “Heads, Tails, Tails, Heads”) if you flip it four times in a row. Percentages don't allow people to see what will actually happens in the future (only Deloreans do), they just predict the results. So if FPR simming is entirely based on percentages, that means you can't predict for sure what actual moves will be chosen from your edit's moveset, right? You are forced to just use probability which is not an exact science, or math.
Here's an example of a purely probability based simming engine: You look back at the CPU logic screen and you see that each category (for example "Front grapple") is broken up into damage settings and then move percentages. Percentages are not guarantees. So even if your edit has his opponent in large damage and your edit has a powerbomb at 50% in that large damage category, that move may not necessarily happen. It probably will at SOME point in the match but there's no guarantee. And it's entirely possible that it doesn't at all. (What if your edit doesn't win a grapple for the rest of the match?) There's only a 50% chance: A prediction of one coin flip. Same thing as Fire Pro, right? Well, not entirely. Here's where it gets kinda weird.
3) FPR's CPU simming system is actually a system of percentages that interacts with a fixed “Random Number Generator.” (I'll be calling it the "RNG" for the rest of this guide). That whole thing I said above about the engine appearing to be based on %'s and not fixed numbers? Incorrect. The engine uses both %'s and fixed numbers to put matches on which take into account your CPU logic settings. Let's explore what this means.
Important note: I really can't emphasize this enough. This entire “CPU Simming Fundamental” section in this guide could not have been provided without the research, help, and consent of Jason Blackhart. The entire reason we know about the RNG and it's effects is because of JB. He has done extensive work on the "innards" of the game. He has stuffed his mighty forearm up to the "gizzard" of Fire Pro. And he has pulled out the "stuffings" of knowledge. I am just providing it lukewarm to you now after he previously served it up hot to everyone. But since those prior servings occur on multiple posts on multiple boards (some of them now defunct), I just wanted all of this information on one plate...uh, place. The information and core analogies for this section came entirely from him. Anything I say in this section (or elsewhere in the guide about FPR engine fundamentals) is entirely based off of the knowledge he passed onto us. I wrote it out in my customary painstakingly longwinded way. But without his research, postings, and follow-up explanations, this second section of Simming Fundamentals would not be known or understood at all by me or anyone else. Without the “RNG” information, there wouldn't be enough core info in this guide to properly lead you to my own resulting strategies for testing.
Also, I'd like to thank Jason Blackhart very much for lending me the approval to post all of this engine info here. He also checked out this guide and confirmed the validity of this engine section. So rest assured, this guide has the JB stamp of approval...until he finds out something new about the game that corrects this information!)
Back to the nuts and Bolts: The RNG is the mechanism that turns probability into reality. It's what determines which moves are actually chosen based on the probability values and also the move slots. In order to fully explain this to you, we have to switch metaphors. No more "coin tosses". Now we turn to an old favorite pro wrestling problem-solving method: "spin the wheel, make the deal". We're dealing with a roulette wheel now. And the "flipped coin" from before is now the "ball" in a roulette wheel. The coin flipper from before is now the roulette wheel handler that triggers the ball around the wheel. He is the RNG of the simming engine. But before we go further into this, remember we are using an analogy. (FPR stands for Fire ProWrestling Returns, not Fire ProWrestling Roulette). Furthermore, it carries with it some deviations from an actual roulette wheel. But first let us examine the fundamentals of the roulette wheel so we can then explain how FPR compares to it.
Imagine a roulette wheel. You're in Las Vegas looking to show everyone how lucky you are. So you go up to the roulette table and offer 10 strawpennies a turn. You are quickly escorted out by security. Six months later you come back under a pseudonym and a fake mustache and you're all right to play. Let's put aside the colors on the roulette wheel and just focus on the numbers for now. So you look at the wheel and guess a number: let's say 7. The ball moves around the wheel and could settle in any of the slots but finally falls into the slot for the number 7. We won....this hypothetical. Congratulations, I guess. it really would have sucked if we had lost.
Remember that every number on a standard roulette wheel has only 1 slot assigned to it. So every number seemingly has an equal chance of actually turning up. If the roulette wheel has numbers 1-50 that you can guess, you have a 1/50th chance (2%) of guessing the right number cause there are 50 slots: 1 for each number. And the mechanism to see which number actually wins is to spin the wheel. Of course, there are two major variables: Which number the ball starts out on each spin and which order the numbers are in. As a result, you don't get that perfectly clean probability math that you crave. But you understand that the equality in this situation lies in the fact that no number has more then one slot for it to be chosen by the ball.
Let's alter that wheel. Imagine if the 7 you chose had multiple slots for the ball to fall into and still be counted as a 7. Now let's say that no other numher had multiple slots. That would give the 7 an increased chance of winning, right? Because if the ball came near the first slot for the 7, it could land into a few potential slots and still count as the 7. This unequal weighting concept takes us back to CPU logic and simming.
When you are CPU simming a match, each move category has it's own roulette wheel. Fire Pro treats each move as a set number of places on it's own roulette wheel. Imagine if the numbers on the wheel now represent what each of your MOVES are per move category and damage setting. For example, the “weak damage” front grapple roulette wheel would just have these moves as potential slots.
[] - Chop to chest
UP + [] - Bodyslam
LEFT/RIGHT + [] - Snapmare
DOWN + [] - Drop Toe Hold
For simplicity sake, let's just stick to those moves for now rather then filling out the entire front grapple list. In fact, to use a practical example. let's say you're someone who only uses the [] moves at the start of the match (when the opponent has low damage). So you have four moves. Let's say you set them in an unequal way in CPU logic. Of course, they have to add up to 100% or the game won't accept it. Now imagine we have 100 slots on the roulette wheel. And they correspond to each move's % like this:
Slot - Button - Move Name - CPU logic %
Slots 1-35 - [] - Chop to chest - 35%
Slots 36-55 - UP + [] - Bodyslam - 20%
Slots 56-80 - LEFT/RIGHT + [] - Snapmare - 25%
Slots: 81-100 - DOWN + [] - Drop Toe Hold 20%
Fire Pro sees those moves as spaces on the roulette wheel. The unequal percentages mean that each move will get an unequal amount of spaces on that roulette wheel. The [] move (chop to chest) will have 35 spaces out of 100 for the "ball" to land in. That means that there is a higher chance that move is chosen then others, right?
Mathematically that would be true if we were still ONLY dealing with percentages and probability. But Fire Pro isn't based ONLY on math. The RNG is the other variable and is the person unleashing the ball onto the roulette wheel. And, it turns out, that man has an agenda. Yeah, you knew he'd back. You can't fight fate. Uh...
The "Random" Number Generator. Notice how I put quotes around the "random" part. That's because, as it turns out, it's not so random after all. The person who spins the wheel in Fire Pro is quite different then an honest spinner in a casino. The honest spinner doesn't have any specific number he is shooting the ball to go into. The FPR spinner already knows what number he wants and you can't fight it. It's set. So when we say "random number generator", the jokes on us because WE are the only audience that the number is seemingly random towards. FPR already knows. That ball is landing where FPR has determined it will land. Calvin was right about pre-destination.
In fact, FPR already knows every number that it's going to pick in a match. There is an entire string of "random" numbers that FPR has in it's memory and uses in the same order every time you start the game. In fact, every time you turn the game on and start a new match, the RNG uses the same sequence of numbers in a row. Once that sequence has been completed, it repeats. For those of you who SWORE your game was haunted because you've seen the exact same match twice...rejoice! Your game may still be haunted...but not for THIS specific reason!
Now you're probably calling “shenanigans” on me. But try it. Turn the game on and pick any two wrestlers. Use whatever your default ring/rules/referee settings are (so you won't have to remember them when you turn the game on again). Sim the match so it's CPU vs. CPU. Try to remember it (it's easy to remember who wins and how). Turn the game off. Turn it back on and try the same match again with the same wrestlers and rules. Provided you chose everything the same, the match should be EXACTLY the same. If it's not, you've got a haunting on your hands and I'd appreciate you stop reading this guide. "Spooked by association," and all.
This phenomenon is what we in the community call a "Boot Sim". I will use this concept again in this guide and it means you can always watch the same match between the same wrestlers (provided you don't edit them in between boot sims) any time you turn the game on and set the rules the same way.
As you'll see later in this guide, "boot sims" have usage for testing your edit.
But back to the "RNG". So Fire Pro knows your matches in advance. It sees through time. It's the singularity, or something. But for our human purposes, it's further proof that the RNG is not actually "random" and that there is a string of numbers that Fire Pro uses to decide what moves to use. Let's go back to our move list example:
Slot - Button - Move Name - CPU logic %
Slots 1-35 - [] - Chop to chest - 35%
Slots 36-55 - UP + [] - Bodyslam - 20%
Slots 56-80 - LEFT/RIGHT + [] - Snapmare - 25%
Slots: 81-100 - DOWN + [] - Drop Toe Hold 20%
So the move slots are represented by non-traditional roulette wheels. There are 20 slots available for the body slam move. The way the moves are chosen is by the RNG which is represented by a roulette ball that goes exactly where the casino employee wants it to go. It's predetermined.
Let's say you boot the game up and start a match. Both wrestlers stare across the screen at each other, sharing a gentle silence and camaraderie we'll probably never know. The first number that's chosen applies to the wrestler on the left and the second number to the wrestler on the right. (And the "move situation" being used is the first one in the CPU logic screen, the "standing/striking" category) Let's just say that both those numbers lead to a situation where both wrestlers end up in the front grapple position. Then we'll say that your edit (with the above front grapple logic) wins the grapple. Then we'll say that the next “random” number to pop up (after your edit wins the grapple) is a 25. The 25 determines what move your edit will use from the front grapple.
What would the number 25 lead to? Your edit's chop would be chosen. But let's exit the match and go to edit mode. Now you are changing your chop to 20% and are keeping your bodyslam at 20%. (of course, you adjust the other two moves so your total adds up to 100%). Now you decide to turn the game off then on, and do the same match again with the same rules (so that we are using the boot sim principle).
Now we get up to the same scenario in the match as happened the first time leading to you winning the grapple. But this time, the RNG choice for 25 would result in a bodyslam rather then a chop! Furthermore, this deviation from boot sim 1 would lead to an entirely different match. Marty, you changed the past!
Let's look at our other chosen moves. What's unlikely to happen on an RNG choice of 25 is for the drop toe hold to happen. That is because it is in the 4th move position and the RNG counts 1-100 from the first move position ( which is the [] move). So the actual move slots that you put your moves in your moveset actually matters for purposes of simming. (It also makes a huge difference for other reasons related to gameplay.)
The point is that even though the RNG sequence of numbers never change, your matches can change if you adjust your logic or moves. But, as we've also seen, you may have to change a lot to get some moves to "trigger" due to the RNG chosen number being too far away from the move you want. Just as where the roulette ball starts on the wheel may determine where it ends, the same goes for move slots and the RNG.
In summary, a logic setting of 10% doesn't mean you have a mathematical 10% chance of seeing the move. It just means there are 10 slots open in that "location" of the move and the RNG's pre-chosen roulette ball may land in one of those 10 open slots. But it may not. A 99% move will have 99 slots open and so it's very likely that any number the RNG chooses, that number will be flund in those 99 slots. This is the crux of the probability behind the CPU simming engine. Once you absorb all of this information, you will start to think about your editing processes differently then you did before. And so we are done with the CPU Simming Fundamentals section.
III. The apple of knowledge in the Garden of Edet.
Let's get this concept out of the way before we move on. Even if we do figure out the exact order of the numbers that are used in the RNG, it's only really useful for boot sims (and the “follow-up” sims that use the predictable set of numbers that would follow after the ones used in the boot sims). And even then, putting on the “perfect” match (to your tastes or anyone else's) is very difficult and ridiculously time consuming. Not only do you have to take every roulette wheel spin into account, you have to make sure those values match up perfectly to give you the right moves for BOTH edits you have in a match. You'd have to micro-manage and maximize the utility out of ever spin of that wheel, every % point of logic and move choice. And you'd have to take into account modifiers like breathing and fighting style and a ton of other variables. You wouldn't actually have control here, you'd just be responding to a very complicated and hard to detect set of numbers. Can it be done? I'd imagine it's possible to construct a match that is incredibly well paced, if not perfect. And of course, “perfection” is largely subjective but if you can make your perfect match with boot sims, all power to you. Once you get beyond that boot sim though, you'd have to do it all over again with a different set of numbers. And after a certain amount of time, you're gaming the engine to produce something that it's not really natural to provide. Some people less enlightened and open-minded then myself could call it an abomination. But hey, those alleged crimes of nature are up to you. It's not what I'm looking to do or advise people about here. For the purposes of this guide, I will be assuming we do not know or use the actual numbers or order of the RNG. We're flying blind here. Buckle up.
However, I am not taking anything away from the usefulness of boot sims for creating "one-time e-fed video matches" or testing edits to some degree. As I will mention later, using boot sims to create changes in one match really isn't the type of thing that will help an edit in the long run. But it's possible you can use boot sims to modify your edit in ways that do affect trends. But you won't know that until you sim your edit enough times. I am not taking boot sims off the table, merely explaining why it isn't the centerpiece. A great chef does not merely create one excellent dish, he achieves mastery over the course of his entire career. A well managed and manipulated boot sim can get you a near perfect match, but it won't guarantee you a great edit over time.
Let's get to testing under normal conditions where we aren't focused on boot sims. That doesn't mean that you reject boot sims for being boot sims. They are normal sims, they just happen to be the one where we can replicate easiest. There's nothing about a sim being a boot sim that makes it worse. Remember how to get a boot sim, incidentally? You make sure you have the same wrestlers, the same ref, and the same settings. Let's turn to what settings that we want to use in order to sim (boot sim or not). And we start with COM Level.
IV. “I'm thinking of a COM level between 1 and 10.”
Here is a popular question to ask before simming. “What COM Level should I set in the 'match options' menu?” First, what does COM Level mean? When you are using a 1P wrestler against the CPU, the COM level is basically the “difficulty” of the CPU opponent of yours. But when you have two CPU wrestlers facing each other, the COM level requires a slightly deeper interpretation. Jason Blackhart once said in the 16th century that the “COM level is the response time logic for CPU-controlled wrestlers”. Here are a few traits that higher COM levels lead to.
Above Level 4:
- Many “Tests of Strength” (ToS) early in the match and throughout.
- Submissions held for less time. The higher the COM level, the less time subs are held before broken out of.
- If you do not see a ToS, one of the wrestlers will win the grapple almost instantly after you see the tie-up. The higher the COM level, the less time there will be in between the tie-up animation and a grapple being won.
Below Level 4:
- Rarely more then 3 or 4 ToS to start the match and almost none later on.
- Submissions held for more time. The lower the COM level, the more time subs are held before broken out of.
- The wrestlers will stay in the “Collar and elbow tie-up” position for a few seconds. The lower the COM level, the longer time there will be between the tie-up animation and a grapple being won. This could sometimes take as long as 3-5 seconds until one of the wrestlers finally win a grapple.
I didn't mention Level 4 just because Level 4 is about the perfect balance between the above two categories. There is no objective right answer as to what the best simming level is but Level 4 does represent the clearest line or balance between the factors mentioned above. It is for that reason that many people, including me, use level 4 to sim. I dislike frequent ToS's and I don't like submission moves being broken up right away. However, I can't stand watching wrestlers take 3-5 seconds to win a grapple. So for the purposes of this guide, assume it will be using COM level 4. Although there's really nothing here that doesn't work under the other levels.
So let's get to the next sect-
HERE COMES A NEW CHALLENGER!
Uh...okay I don't know where that came from or why I'm suddenly on a plane to Thailand. But I'm reminded of the fact that there's still one more crucial thing to mention about COM level before we move to the next section.
Grapples: How do they work? A lot like magnets, really. There's kind of a “positive” and “negative” to each side. And what determines that? No, not magic. (Stop reading this immediately if you answered magic.)
If you've been paying attention, you should know that two things control grapple winning:
1) COM level: Obviously...since we are STILL in this section of the guide!
2) Random Number Generator: determines every major “choice” in the game and your life.
The “choice” here is for each wrestler to determine their “response time” in the “collar and elbow” tie-up. The absolute best way to understand this entire dynamic is in the context of a specific conversation a few of us had with Jason Blackhart. I will attempt a summary of the discussion, specifically what JB said, for this guide. JB's quotes are in purple, as is the style of the time.
“While the COM levels you see on the match setup screen range from 1-10, internally, they are also on a scale of 100. - The (internal) COM Level is used as the CPU Response Time logic. The random numbers are checked against this logic to see how quickly they should "input" key presses. This is used for the speed of the various button mashing events (tests of strength, breaking out of submissions) and attempting a grapple attack after locking up.“
The only thing that I know of that isn't completely accurate anymore is COM level having an effect on the pinfall kickout "virtual button mashing". I saw that it used the same RNG function for its "response time logic" and just assumed it was based on COM level like the others, but later found that it apparently always uses a constant 50% fast/50% slow logic (equivalent to level 6, I think) no matter what level you're playing at. I had to check more thoroughly after finding that edits had gained more spirit during a pinfall at level 1 then they did at level 10 in some test matches.
And then he discusses what actually happens in a “collar and elbow tie up”
"The COM Level is still reliant on the random numbers, though. It works just like the other logic values. For an example, it would be like:
CPU Response Time
Fast = 70%
Slow = 30%
(That is actually what COM Level 7 is, btw) Then at every frame (or maybe every other frame) during the relevant situations, a random number is compared against the COM level logic to see if they should press a button. If it was 1-70, they'd respond, while if it was 71-100, they wouldn't do anything for that frame.”
The reason why lower CPU levels lead to longer periods of no one winning a grapple is because of what JB said. He showed that Level 7 has 70/30 response time. Courtesy of JB, here's the response time for all CPU Difficulty levels:
L1 = 5/95
L2 = 10/90
L3 = 15/85
L4 = 25/75
L5 = 35/65
L 6 = 50/50
L 7 = 70/30
L8 = 90/10
L9 = 95/5
L10 = 100/0
So, using that list, Level 1 would have 5/95. That means that the “ball” in the roulette wheel has a 95% chance of landing on “slow” and then the wrestler won't do anything as a result. But why do BOTH wrestlers then wait so long to win a grapple? If one gets “slow”, shouldn't the other get “fast”? No, it doesn't work that way. Both wrestlers get what they get regardless of their opponent. So both wrestlers can get “slow” and have a 95% chance of getting that! It's because they are each operating on the same COM level 1.
The COM level is the great equalize. JB gives us this example:
“You could have a 380-point beast with a loaded move set and MURDERKILL logic going against a 0-point jobber with the weakest moves imaginable and crappy logic, but when it comes to grapple timing (or a test of strength or whatever) they both would have an equal chance, since you can only set the response speed (COM Level) for the match, rather than each individual participant in the match. “
Read that again: YOU CAN NOT CHANGE THE GRAPPLE TIMING FOR THE SPECIFIC WRESTLERS IN THE MATCH. It's always the same for both wrestlers. You can only set the response speed for the overall match.
Now, I see your mouth forming a “U” shape. That looks painful. But more importantly, it won't lead where you think it will. You were about to say that "UKEMI" can give you an advantage in the engine by making one of the edits an expert in countering. But it's way more limiting then you think. JB? If you don't mind...
”Ukemi is really about NOT countering your opponent's grapple move. When ukemi is at 0%, it makes the character play "normally", countering based on their remaining health, body size, fight style and return skill. Anything higher will mean they have that chance of "allowing" the move to be successful against them (with the possibility of a later counterattack) regardless of whether the numbers say that they "should" counter the move or not.”
Do you feel your power over the engine fading? Yes, the [unnamed cosmic energy that I don't want to get sued for mentioning in this guide] is growing weak with you. Your apathy is rising but your power is not. Should you stop playing the game forever? Maybe there's another way of dealing with this new found futility?
Maybe it's time to learn to compensate. It's finally time to get to the next section where my advice truly begins.
V. How to randomly generate utility out of the RNG.
This section is best understood as how NOT to get angry at yourself when “simming doesn't work out” or when “you realize that everything you thought you had control of in the game laughs at your foolish display of pride."
We're switching analogies yet again. We're going from roulette to the poker tables. But we're not playing against another person. Rather, we're controlling two opposite sides of a poker game. What we're not controlling though is the deck of cards. All we know is that the deck has a pre-determined order of cards but we don't know what the order is. But that lack of knowledge on our part doesn't mean an objective set order doesn't exist. (No “Schrodinger's Hand Cat” here). This isn't quantum physics, just poker with a set deck.
So imagine we wanted to arrange a "fixed" poker game between two sides to entertain a third party audience (rather then playing a game for the motive of having competition between two players). But even though we're controlling both sides, could we put on the most entertaining game (or even one maximally entertaining hand) without knowledge of what the sequence of cards are? Not reliably or consistently. We could have each side play percentages,but we can't guarantee what actually happens. If one side starts a hand with a pair of Aces and we try to get them three Aces, the percentages we use do not adequately represent what is, in reality, the next few cards available to us in the deck. We may have a low % but still end up with the right card. Or we may not. All that really matters for winning is what the next few cards are, not a mental construct of probability. But we play cause we think we can guess right, and sometimes we do. It's exciting when we do. And we win money.
In Fire Pro, we have a little more control then Poker but we're still operating with a fixed RNG (like a deck of cards) that has an actual order which we try to approximate using probability and logic values. We could be way off in how our percentages match up with the RNG numbers. But without knowing the actual numbers (or memorizing boot sims), percentages are our best guide to setting up the match the way we want.
So what is the result? It's not chaos and the lamentation of our eternal souls, but it's close. It sure isn't the control over the system and our edits that many of us want (and thought we had, as little as two years ago). The fact is that even the best edit makers see their best edits take part in squash matches. The best edit makers have routinely witnessed their edits do stuff they didn't plan to have happen. And so the best way to use the RNG is to realize how much is truly out of our hands with Fire Pro and enjoy what we ARE capable of controlling.
What can we control? Our CPU sequences (which the RNG rarely disrupts), the taunts we give our wrestlers, the sounds that go with the moves. You're in charge of that, Captain. But we shouldn't spend an entire painstaking night of simming and testing worried about whether our finisher should be 1% or 2% with the expectation that our choice will always be objectively correct. It won't be. At least not over time. At least not against every expected opponent edit in every sim. Maybe one random sim (or even a slightly controllable boot sim) will turn out better with 1% rather then 2%. But it's a long shot, and it's not that consistent.
So here's the big secret: We can't micromanage every match. Instead, we're looking for TRENDS in logic because THOSE we have some control over. The best edits are the ones that, over time, have the highest likelihood of meeting certain strategies that you set up. And remember, the word "trend" does not exist without that element of "over time". You won't see a trend in just one match. You need multiple matches to even look for a trend. So let's examine what kinds of "trends" are able to be seen and controlled in Fire Pro matches.
An example of a trend you can look for would be my "layering" strategy that I discuss in my guide. I can't guarantee that will work ALL the time on even my best edits. And I won't pull my hair out if it doesn't occur properly in one match. I can't guarantee it will. But if I can set an edit's logic up in a way where the edit is utilizing that strategy in the majority of sims against their intended opponents, then I'm very content. That's a win, in my book.
Another example of a trend could be someone wanting certain holds to come out in the early/mid-portion of the match and stop before the end portion. Maybe you have a camel clutch that is used merely as a "wear-down" hold. The engine is too out of our control for us to reasonably expect to see the move the SAME number of times every match you sim (unless that number is "O" which you can guarantee).
But a "trend" you can reliably shoot for is to see that move happen maybe "1 to 3" times per match. If you try to set your logic up like that, it's possible to generally succeed over time. However, there will be matches you don't see the hold at all (since "1" was your low end expectation) and there may even be a few matches you see it occur 5 or more times. If those outlier situations happen in the rare minority of matches, you have succeeded in setting a trend for your edit. If you can see that same trend against a wide variety of opponents (especially your edit's primary rival type opponents), then you have done a VERY good job setting that trend.
Just about all of my methods below for improving an edit hinge on the idea of "trends". It will show up in all my suggestions. If a “trend” is the outcome I shoot for, then “multiple sims” and “PATIENCE” are the two tools to achieve it. The idea being that you always want to give an edit a few sims to see how they work before you make any major changes. Patience will help you because one single match may not represent how an edit works overall. It's entirely possible that the first match you ever sim with a new edit just happens to be representative of that edit over time. (or of your edit's best possible matches). But you won't know that for sure until you actually sim that edit a number of times. You can immediately strike gold with an edit in Fire Pro. But you need to consider it “Fools Gold” until you can prove your edit can keep wrestling like that over time.
The best edit making and testing isn't just about achieving success, but recognizing it as lasting success.
And this whole guide isn't just about how you manipulate the game, but what your own expectations for it are. You can't micromanage and expect perfection out of an engine that acts on percentage with so many strange variables and "moving parts" as FPR does. And if you could manipulate everything, you'd lose that unpredictability that makes you want to watch to begin with. The middle ground here is key.
“Fly the friendly sims” or “Change over time”.
Now we get further into the stated purpose of this guide. How do you take an edit (old or brand new) and get the maximum value out of the simming tests you run on them? Let's look at a few ideas.
1. "First symdrome" - What happens when you don't have the patience I suggested above? The universe will laugh at you as it manipulates your fragile psyche. Let's say you have just put a new edit in his first sim and things are going well up until the final damage category. Your edit then hits his finisher as the first move of that damage category (and also as the first move of his “O” set of moves). In addition to that move occurring before other moves meant to proceed it, the move also occurred much earlier in the match then you wanted, in terms of the clock.
What's the wrong response? Scream in rage, exit the match prematurely, go into CPU logic, put the move at a lower %. What's the outcome of that? You wake the neighbors and they call the cops. Then when you finish explaining to them your dilemma and get a summons, you sim your edit once again against the same opponent. And you somehow end up with a 20+ minute match where the finisher doesn't occur at all. Your battleship, it is sunk.
What's the correct response? Relax. Don't change your move logic. Regardless of whether the move is set right, you over-reacted in this specific situation. It's entirely possible you got one fluke sim or maybe the RNG landed very close to the "line" between your finisher and another move. The point is, it may not be a TREND. What you saw was one move happening once, just a bit earlier then you wanted. That alone such an outlier situation. And it sure isn't necessarily indicative of a trend, especially after one sim. Could it turn into one? Yes. But you didn't give yourself a chance to prove if it was a trend since you didn't let it stay as it was and test it over time.
In this specific example I gave, it probably isn't a trend. Or at least not an overwhelming trend. The biggest problem my Orochi Geese edit had was affectionately referred to as the “7 minute Ganso.” That meant that there were some matches he did his Special “Ganso Bomb” move around the 7 minute mark. The “7 minute Ganso” was a trend, but a trend that only occurred around a rate of once in five matches. And even in the matches he did it in, he was fine otherwise. The other matches he did the move when I wanted him to. Since I loved everything else about his edit, I accepted that small trend because his logic was set up in a way to produce what I wanted to see overall. And changing the logic for that one move would have ripple affects that I did not want to occur.
So why did the “7 minute Ganso” even happen? Chances are that if your finisher is at a number not all that different from some of your other moves in that category (like a bunch of 7% and 8% moves in front grapple), you may see it pop up early. But if you lower it too much (especially in an angry reaction to one sim), you may lose it.
I accepted the situation after weighing the costs and benefits. It also helped that the personality behind Geese's edit had a mean streak so it wasn't out of character for his edit to occasionally spike someone on their head earlier then they may have been “ready” for in a technical style build-up match. Sometimes surprises are okay.
2. "Recognize the real" - This is the exception to the “don't immediately change logics” suggestion offered above. There are a few situations when you absolutely should make changes after only one sim. These are the DRASTIC mistakes (that often occur after making a brand new edit) that can be easily seen and fixed. Don't get annoyed when you see these. Just laugh them off and fix them. We all overlook things when making a new edit (or even fixing an old one). Here are some examples of things you'll want to change as soon as you see them:
- A finisher in the wrong damage category. You may have accidentally changed the "0%" of your finisher to "1%" in weak damage without realizing, Then you see it 30 seconds into the match. Take solace in how easy it is to fix this. Use the occasion to look for other such errors you may have missed. Laugh at how 1 is more then 0. Silly 0.
- "Tightening up the logics" as I say in my above linked strategy guide. If you agree with my position, then you'll be annoyed if your edit does an early match "wear down hold" at the end of the match (win or not).
- A mistaken CPU Sequence. By accident you linked your finisher to your HEAD X "elbow drop" rather then the HEAD O "Pin" as you originally intended. Happens all the time. That's a paddlin', but also an easy fixin'. It's only rage-inducing when it ruins an epic video match and you can't repeat it with the right ending.
The point is that you don't need to wait for trends in this category. Once you see these obvious and glaring mistakes in one match, they will definitely keep popping up and ruining other matches. It's not a question of the RNG or probabilities messing up as much as you allowing them to happen. You can easily fix it to avoid those specific problems occurring in the future. These are the “I forgot to proofread” my edit errors.
3) "Logic Override for great justice" - There some occasions where FPR "overrides" your logic in some way. This isn't something that you necessarily want to adjust due to potentially negative side effects of that change. But it is somewhat within your ability to stop it from happening again. Also, this may show up in trends but you don't always have to wait to see it before you do something. Here are some examples:
- Your edit pauses to breathe while in between two moves of a set CPU sequence. If your serious time (aka "Discretion") is high, this can happen. But think about this: It's not necessarily a bad tendency for a wrestler to want to breathe when they initially feel they need to. In some situations, proper paced breathing will allow your edit to continue wrestling normally and not be standing tired and defenseless at the end of a match. It's just that in this specific case, the breathing disrupted your sequence. It's entirely possible that, without breathing, a sequence would have been interrupted by your edit being out of breath! Live by the sword, die by the sword. Occasionally your wrestler will breathe at a "bad" time. You need to ask yourself if you should reduce Serious Time, maybe do something to help his breathing (like switching fight style or changing some moves), OR leave the situation alone. Sometimes it adds to the drama of a match to see a wrestler too tired to pin after their own finisher. Other times it could take away his one opportunity to win. This is very much up to your...ahem, discretion.
- You saw "R" trigger usage (aka "Front facelock + walk") in a match when you have it at 0% for that damage category. Ahh, the phantom R trigger. The bane of many matches. It occurs sometimes when you have one of the "shoot" fighting styles or your opponent has it and your flexibility is high. Even if you have the move set to 0%, you still may see it. Is it worth changing your fight style for? That's your call but I'd advise against it. It may not even happen again. If you want to allow the headlock to keep happening but prevent it from ending the match. you can do that. Just eliminate any "face down" pins or submissions at large damage so that even if the headlock happens, it doesn't lead the worst ending in FPR: "R trigger headlock -> Pin/Sub".
- Your wrestler and his opponent continually "reverse" each others back grapples into their own back grapples. This goes on for like 30 actual seconds. Look, you got caught. You wanted to have “that cool technical edit" who reverses everything into their own back grapple and now have the misfortune of being up against another edit with that same strategy. What's good for the Gander and all. Either let it slide or ask yourself if you REALLY need your "reverse to back grapple" counter at 100%. You probably don't. Spike put in two reversal slots for a reason..
4. "Deja Vu" - Whether you are simming your edit against one opponent or against multiple ones, give your edit a few sims against each. Your first sim of your new edit against Wrestler A may not give a representative sample of what your new edit or the match is capable of. Try it a few times and see the patterns and trends that emerge. Again, patience. Can multiple sims against the same opponent get boring at times? Yes. After the third sim against the same edit, you may want to move on. And if things look okay in those sims, you should move on.
But if you want to improve your edit, sometimes you have to put up with some boredom from seeing repeat matches. In the end, it will be worth it. The newer an edit is, the more repetition of opponents you need to see how your edit acts. I know you want them to run the gauntlet of interesting opponents right away, but you'll be doing yourself a disservice of not fully understanding how they work if there are too many variables. That leads me to:
VII. My strategies for testing an edit.
As do many others (mentioned in Section IX), I take a tiered approach in which edits I choose to face a new edit.
Here is the slow and steady "Gauntlet" of the types of edits that I use when testing a new one.
1) "Control" Edit 1: Even though this isn't a real science due to the RNG and percentages, I do think that attempting a scientific approach is helpful in testing old and especially new edits. An opponent you know is the "control" and the new edit is the "test". It is crucial for early testing of any new edit that you have a "control" edit. The first "control" is an edit you know extraordinarily well because you have watched tons of their matches and even made the edit with it's purposes, strategies, and trends in mind. It is good to have at least two control edits: one that is a "face" and one that is a "heel". Both of them are probably two of your simpler and more predictable edits, likely a rookie. So if something is "off" in a match with an edit you know really well, it's PROBABLY due to the new edit.
Keep in mind that any "control" edit that you simmed tons of times is likely an edit that you made a while ago. It's possible that your edit making has evolved or changed since you made them and any oddities in the match are due to obsolete logic strategies you used for the older edit. Sometimes a "control" test actually makes an old edit better by showing how much better the new edit is in some ways. So it's not true that anything "off" in a match is ALWAYS due to the new edit. Once again, you need to see "trends" before you know for sure what is causing any problems. If everything is going well after a few sims, then I move onto the next edit in this "gauntlet".
2) "Control" Edit 2: These are usually other edits I have made but that I don't know as well as Control Edit 1. However, I still have a lot of familiarity with and confidence in them. I still want some degree of predictability but I also like testing the new edit against someone with more logic variation. See how they do against someone who can use a number of different strategies (or a hybrid) in one match. That way you can test your edit's defenses and ways to cope. You want someone that isn't too much stronger or weaker then your edit. Control Edit 2 isn't that much of a step removed from Control Edit 1 in terms of giving your edit a consistent simming challenge. You still want to make sure that you understand the match more then you are potentially entertained by the styles clash. You're still learning very much about how your edit works and so too much unpredictability or power contrast from an opponent isn't good. This is only a small step beyond Control 1 but it's a further step into the deep end.
3) "Control" Edit 3: This edit is higher up on the card (meaning should be more likely to win in a hierarchy) then Control edit 1 or 2. This is an edit I'm more familiar with then Control Edit 2 but maybe a little less then Control Edit 1. If you have a stronger edit for Control Edit 2 AND one that you aren't as familiar with, any problems in the match can be harder to explain. You are still taking gradual steps in how different the test edits are from the first one and from your test edit. You want someone you know really well here but that is more likely to win. It is okay if your new edit wins a few of these matches but you want to make sure they aren't on a totally even keel. If your new edit keeps winning, make sure it isn't a problem with the control edit. If your new edit and Control edit 1 were meant to be on the same level and were, then Control edit 3 may be the one out of whack in your hierarchy.
4) "Control" Edit 4: Now I start to use edits which I made that are a lot more variable and unpredictable then the earlier ones. Some of these edits may have very experimental styles like "face down" pin strategy or "very heavy on striking or submissions"). They may also be radically different in card placement. This is the first of two phases where you can see how well-rounded your edit is and how they are able to adapt.
5) "Control" Edit 5: Here I use edits from other people whose edit making and specific edits I know really well. It's important to make sure that your style of edit making for this specific new edit meshes well with other people's edits if you are in a multi-person e-fed. This really speaks to the well-roundedness of your edit. Often you'll have to adjust for small equalizers like stats or ukemi. You'll want to test against edits from others that you know well, edits from handlers who have similar edit philosophies to you, and edits you just outright love to watch.
5) Expected short term and long-term opponents: You may already know who your edit is going up against in prospective matches or long-term rivalries. It's good to have your edit ready to face them before you send it off to people. No edit exists in a "vacuum". You're going to have opponents and so you want to make sure your edit puts on the their best matches possible against those expected opponents and rivals in the same division.
In fact, sometimes I'll actually skip a few testing steps and start with this one. If I made a new edit with a very specific storyline/feud opponent in mind, then they may be the first opponent I test against. A few years ago I made an edit named Kyle Walker with the intent purpose of having him be a vicious antagonist to Orochi Geese. As a result, I had Kyle immediately simming against Geese cause that was the money match. Unless Kyle put on an excellent match with Geese, Kyle's edit was pretty much worthless to me. Of course, it wasn't enough for him to ONLY put on a good match with Geese. Once I was happy with their matches, I then had him face other prospective opponents. As a result, Kyle's matches with some of my test edits aren't great but his matches against upper card rival story wrestlers have been excellent. That's a balance I'm very happy with as it's the reason for his edit.
Prioritization matters. No edit can have 5 star matches with every edit they encounter. There are way too many different ways of edit making and some style clashes can't be compensated for. But if your edit can put their on their best matches (and have that be a trend) against his expected opponents, then your edit making is optimal.
Part of that prioritization means knowing when not to change your edit even if there are mismatches.
I will never make an edit better against non-rivals if those changes makes the edit worse against his main rivals. My top e-fed edit feud was Orochi Geese vs. Thunder Strike. Any time I would change one of their edits in a way that would make their matches worse, I'd try to compensate for it or outright undo the change. I'm not interested in having an edit work with every edit imaginable. (There are literally thousands of FPR edits). For my e-fed edits, matches with their top rivals out-prioritize matches with random edits they won't face. As a result, I try to strike this overall balance when making and testing my edits. They should be able to put on:
5 star matches vs. approximately 5-10 of their top enemies and rivals.
4 star matches vs. approximately 10-15 of other rivals in the edit's division and low intensity feud opponents.
3 star matches vs. approximately 20+ other wrestlers in your promotion that you may end up facing or have something in common with your edit.
2 star matches vs. wrestlers not in your promotion (ones with a totally different point scale).
1 star match vs. wrestlers you know nothing about and may have been made with radically different logic standards or strategies and point structures that you used. Basically, the 1 star here comes from your edit at least following your own guidelines for him even if the other edit is all over the place. You can't control the opponent.
VIII. The League of Extraordinary COMskips.
Full disclosure here: There was a time early on in my edit testing “career” where I did everything that I am telling you not to do in this section. I learned the error of my ways through experience, reasoning, and being yelled at for being stupid. Never underestimate the positive results of being yelled at for being stupid.
COM Skip is found in the “match options” screen after you select a tournament or league. It allows you to sim matches in tournaments and leagues really quickly by running matches at like 2000% speed. You see the equivalent of a “slideshow” of the match because it's moving so quickly. As a result, you really don't see what is happening at any given time other then a few frames of moves. What are the reasons people do COM Skips?
1) Just for fun: When you're in that Fire Pro “addict” mode, you'll have Fire Pro doing stuff for you even if you're not there to see it. Just knowing you have FP running is enough to give you the warm fuzzies. (Note: If the “warm fuzzies” are in your chest or left arm, seek a doctor. It's not Fire Pro causing it.)
2) Just for curiosity: You don't really care who wins but are just curious what will happen if a set amount of edits face each other. You don't plan on making any changes based on the results or even placing that much emphasis on the results. You just want to see who can win a match at any given time but not actually watch the matches. Maybe you have a few edits who are feuding and you put them together in matches and see who comes out ahead. There's no issue with this reason or reason 1. People “COM skip” all the time and don't place any emphasis in it. It may not be your best use of time, but then again, there's an argument that reading this guide isn't either.
3) Instant rankings: Uh oh. Slow down, Professor. We got problems. You're looking for a quick-fix here: A quick and dirty way to bypass watching actual matches while hoping to get the same information you would get by watching them. Or maybe you realize you won't get the same information but at least you will get it quick. Just like most “get rich quick” schemes, this will only result in pain and disappointment. Why? Simple.
How are you going to identify trends in matches when you can't even see what's happening!?
How are you going to base a “rankings” or “Tier” list on results from one sim that you never actually fully saw? Or even TEN sims that you never saw? Because the fact of the matter is that even if Wrestler A beats Wrestler B ten times in a row in COM Skip sims, YOU DON'T KNOW HOW THEY DID IT. You know the finishing move and you know the time, but you have no idea WHY the match went down that way. Which edit is at fault for a squash?
You often don't know if Wrestler A won by a small margin or a huge margin. If you do know, you don't know HOW Because even if Wrestler A was seemingly crushing Wrestler B by beating them in under 5 minutes, you can't see what moves or tendencies they were using since you didn't actually see the match. What if Wrestler B was losing due to missed opportunities? Sometimes that can't be helped, like in situations where Wrestler B irish whips Wrestler A into the turnbuckle and isn't able to follow up each time. There are many reasons why a match can go one way or another. If you aren't watching the match, you know next to nothing. Your tier list is based only on wins without knowing why it happened. It's not a tier list, it's a coincidence list. I spit on that list.
And the only thing worse then striving for a COM-skip based tier list is modifying an edit based on that tier list. I said earlier in this guide that you shouldn't modify an edit after only one match unless you are sure the problem fell into one of the obvious categories (like a very early finisher or accidental late match stretch). Would you modify an edit after not even seeing the match? You can't see any trends at all that way. Even if you get 10 matches where one wrestler squashes in under 5 minutes, it would be extremely foolish to change that edit without thoroughly seeing HOW they squashed. How do you know whether the reason they squashed is due to points or logic? Or a combination of those? Or skills? Or maybe it was just strange Fire Pro mechanics? Or maybe a wizard did it?
You don't know.
As someone who once put a lot of stock in COM skill tier lists and even edited wrestlers based on it, I can tell you it's fool's gold. If you want to test your edits against each other to find “relative match strength”, that's fine. It's very understandable if you have a pack or fed and you want to see (and control) who's more likely to win. But you won't be able to truly meet that legitimate goal if you aren't watching the matches in their entirety and accounting for every variable. Don't just leave this to the wind. Take control of the situation as much as you can. Look for trends, not tiers from COM skipping. A wise man once maybe said: “COM skipping is tripping.”