Formation Rating System (WIP, input would be appreciated!)

Here is something I salvaged from my old computer's hard drive before it crashed, haven't looked at it for a while but here you go; the formations are somewhat dated for that reason, but I think the general concept is something which could be handy in the long run as a formation rating system, particularly as an objective clan recruitment requirement (as opposed to "elite forms" which will be judged for their elite-ness on a case by case basis)

I haven't really updated the whole thing; the only change I've made before making a screenshot is updating my own formation.


formulas

//


sample formations

//

sample formation ratings

//

just so you can see how dated this is, there was actually an A form rating system


as you can see; there are some problems with this. It is somewhat heavily based on a formation's skill set; giving basic sin wyrms, and covert wyrms incredibly low scores compared to "generic entry level" etc...

does anybody have any advice on refining these equations?

or has anybody developed a similar system of their own?

My 2 cents.

1) Bp multiplier.  Your formula's big problem is that it rates a 13 bp AA with a good skill better than a 45 bp AA with a crap skill.  So some sort of multiplier should be done given an AAs Bp, as sheer overwhelming stats can win.  This will optimally be done my somebody with more time than me to figure out the proper ratios.

2) I'd say your skill ratings are slightly off, yes the holy quintuplets in the Double AA game are health, IPD ,DED, DEA, IPA, but your formula underates guild downs.  And I'd probably argue that DED as a Skill seems to be more important than IPA.  Your value system could use an overhaul over its simplistic formula.

3) Relica and new form combos made certain junk skills like Increase dodge rate way more infuriating and makes winged forms stronger.  It could be some recency effect giving me a bias, but man it seems like you can make teams miss way more often due to Dodge rate plus/accuracy down relica stacking.  And well everybody needed more reasons to run just 5 winged monsters and now they don't even lose a combo.

Interesting concept. But I agree that it needs refinement. I'll see if I can make a different proposition for it later. Off the top of my head, I'd redefine primary as health, ipd, and dea, secondary as ded, ipa, guild down, and species down, and tertiary as anything else(apart from Crit and heal). Also, I would focus on the total bp instead of the number of elite members, though I'm not sure how exactly to weigh that. Perhaps 45%/35%/20% weight given to skills/bp/combos or some such.

Edit: here's a proposition by me: 
If bp < 86, 0% else (ceiling(bp/86)/6)*30 
((#prim*4+#seco*2+#tert*1)/16)*40 
((5-#sc)/5)*10 
(#comb/5)*20 

The sum of those makes the score. #sc is the number of skill changes needed to optimize skills.

MR.C wrote:My 2 cents.

1) Bp multiplier.  Your formula's big problem is that it rates a 13 bp AA with a good skill better than a 45 bp AA with a crap skill.  So some sort of multiplier should be done given an AAs Bp, as sheer overwhelming stats can win.  This will optimally be done my somebody with more time than me to figure out the proper ratios.

2) I'd say your skill ratings are slightly off, yes the holy quintuplets in the Double AA game are health, IPD ,DED, DEA, IPA, but your formula underates guild downs.  And I'd probably argue that DED as a Skill seems to be more important than IPA.  Your value system could use an overhaul over its simplistic formula.

3) Relica and new form combos made certain junk skills like Increase dodge rate way more infuriating and makes winged forms stronger.  It could be some recency effect giving me a bias, but man it seems like you can make teams miss way more often due to Dodge rate plus/accuracy down relica stacking.  And well everybody needed more reasons to run just 5 winged monsters and now they don't even lose a combo.

1) I agree, there is a problem with this being far too skill oriented, however; I'm not so sure about the "overwhelming stats" argument. 5AAAs are equivalent to a 300bp AA team (being that 100bp AAA has similar stats to a theoretical 60bp AA), and 5AAAs with no skills are an absolute joke. (the old mcfly)

2) Yeah, there should be some sort of more detailed stat hierarchy; my guess being HpUp>IPD>DEA>DED>IPA, I chunked everything together in the way I did to get nice round results (for max score /100), do you have a suggested equation for weighting skills against each other? I'm going to stick with my position on considering guild downs tertiary skills; they are devastating when they take effect, but as an "overall" formation rating, the frequency of their taking effect must be taken into account (so; Impulse down may be worthy of a relatively high weight, Pyscho down of a medium weight; but covert down is hardly ever useful)

3) relica is far too complex a system to integrate into formation rating. the whole idea is to get a quick score based on members and skills; you could get far better indicators of effectiveness by getting full formation stats, skills & levels, relica, including player guild in the equation etc... but what I'm going for here is simplicity. However, there does need to be some modification to combo weight; as winged != bp combo in terms of functionality, etc...

Armitaage wrote:Interesting concept. But I agree that it needs refinement. I'll see if I can make a different proposition for it later. Off the top of my head, I'd redefine primary as health, ipd, and dea, secondary as ded, ipa, guild down, and species down, and tertiary as anything else(apart from Crit and heal). Also, I would focus on the total bp instead of the number of elite members, though I'm not sure how exactly to weigh that. Perhaps 45%/35%/20% weight given to skills/bp/combos or some such.

Edit: here's a proposition by me:
If bp < 86, 0% else (ceiling(bp/86)/6)*30
((#prim*4+#seco*2+#tert*1)/16)*40
((5-#sc)/5)*10
(#comb/5)*20

The sum of those makes the score. #sc is the number of skill changes needed to optimize skills.

that's a good point about doing away with the critical skill bracket. that was a somewhat dated notion; and a formation with wings and no health now-a-days can be more powerful than a formation with health and comparatively weaker members (or even equivalent strength members). I still don't think guild and species should be considered secondary, as they are too contextual; they should be rated based on average effectiveness.

the bp idea is a good one.

perhaps rather than the VIP score; we could go with something like bp/200 (considering an "elite" formation to be comprised of 5 40bp members)

adding rather than multiplying member, combo, and skill scores seems like a good idea since one major problem with my system is that if they score 0 on anything; it multiplies to 0 at the end.

thanks for the feedback and keep it coming!

As an aside, I was always curious about mcfly.  Are his triple AAAs plussed level 300?  i could never tell.

Overwhelming stats can be a thing, I'm relatively certain Manu trucks Harrier Mushi in a straight up fight despite the crappier skill.  There is a statistical tipping point where crappy AA's stats X just can't overcome superior AA'S stat Y despite the good skill bump.

That being said team synergy is definitely a bigger thing. And if having Mushi means your your Saws of Time and your 20000 DEF gets the benefit of IPD, well sometimes you just bite the bullet.  But IMO, you can only get away with a team having so many good skill Bp lightweights before it tanks your stats.

It's really what makes the Bp stat breakdown difficult, the team synergy in relationship to be values makes such a huge difference. 

Combos and their place, lol, I'd almost feel you need to do this twice once for offense and once for defense.  And I'd be super lazy and just use the offensive bonus count as a the multiplier.  Winged counts as a 1.15 multiplier, impulse as a 1.10 multiplier, etc...  So Winged Psycho Brutes get a 1.55 multiplier to the score.  Defensive values could use defense, Hp up, as a bonus multiplier.  It leads to ugly numbers, but artificially constraining a system to a 100 point scale, leads to slightly more wonkiness IMO.

MR.C wrote:As an aside, I was always curious about mcfly.  Are his triple AAAs plussed level 300?  i could never tell.

Overwhelming stats can be a thing, I'm relatively certain Manu trucks Harrier Mushi in a straight up fight despite the crappier skill.  There is a statistical tipping point where crappy AA's stats X just can't overcome superior AA'S stat Y despite the good skill bump.

That being said team synergy is definitely a bigger thing. And if having Mushi means your your Saws of Time and your 20000 DEF gets the benefit of IPD, well sometimes you just bite the bullet.  But IMO, you can only get away with a team having so many good skill Bp lightweights before it tanks your stats.

It's really what makes the Bp stat breakdown difficult, the team synergy in relationship to be values makes such a huge difference. 

Combos and their place, lol, I'd almost feel you need to do this twice once for offense and once for defense.  And I'd be super lazy and just use the offensive bonus count as a the multiplier.  Winged counts as a 1.15 multiplier, impulse as a 1.10 multiplier, etc...  So Winged Psycho Brutes get a 1.55 multiplier to the score.  Defensive values could use defense, Hp up, as a bonus multiplier.  It leads to ugly numbers, but artificially constraining a system to a 100 point scale, leads to slightly more wonkiness IMO.

oh the system can always be translated to a 100 point scale; e.g. if the "new" system has an optimal formation score of some "value" = 538, we can just divide everything by 5.38 for final score

in regards to mcfly? I'm not entirely sure; I suppose so? I've never fought him with an unskilled team; I suppose that would probably lose. However; I think IPD isn't the best example; if you think about something like DEA or DED, that does allow a much weaker monster to gain a substantial advantage against something highbp and unskilled; on account of proportionally decreasing skill of enemy; but then stat distribution comes into play too (a:d:h ratio, with attack heavy monsters, def heavy, etc etc...)

Ok, after some consideration, I've revised my proposition for a rating system:
If bp < 86 then 0 else ((ceiling((bp-(AAA*40)-(AAA+*25))/86))/5)*30 
((#prim*4+#seco*2+#tert*1)/16)*40 
((5-#sc)/5)*10 
(#comb/5)*20 

A bit more complex, but it attempts to deflate AAA/+'s bp. I chose 86 as the base bp as that is the bp of an entry level AA/+ form.

Armitaage wrote:If bp < 86 then 0 else ((ceiling((bp-(AAA*40)-(AAA+*25))/86))/5)*30 

((5-#sc)/5)*10 

looks good! I'm somewhat flummoxed by the first equation though, what does AAA stand for?
I'm assuming ceiling is 200

(if AAA is some special feature to use this to rate AAA forms, we can worry about that later (or perhaps never since... i mean it's AAAs, who cares, that's about as important as an A form rating system now ) and try get the AA form equation down first)


and in equation 2, what is #sc?

Sc is still skill changes. And AAA/AAA+ is the number of AAA/+ monsters in the form(since a AAA is as strong as a 60 bp AA, for each we subtract 40, and a AAA+ is as strong as a 75 bp AA(iirc), so we subtract 25 for each)

Thank you for the feedback! I've made some changes. It's still WIP





relative formation scores of sample formations seem to make a bit more sense now, although I'm sure this requires a fair amount of additional fine (and rough) tuning


now although the scoring system in itself may still be under construction; could anybody advise on some good names for tiers or brackets?
(names which could applied to the final refined system)

going purely off of a normal distribution

"Elite" [score 97-100]
"Superior" [score 84-97]
"Standard" [score 50-84]
"Deficient" [score <50]

or maybe

Elite
Advanced
Intermediate
Beginner

idk, I'm not very creative but the idea is that instead of subjective assesment of every single recruit;

I could say "clan requirements: standard grade form"

or if I was in PU I could say "clan requirements: elite grade form"

i don't quite understand what's the pc/sc in the combo rating, and how that works. also, i don't really agree with the 200 bp being considered the baseline here, as this is like trying to measure a human in relationship to a skyscraper. i keep pushing the 86 bp as a base in determining the bp score as 86 is the bp of the minimal AA+ formation that is barely stronger than most A+ forms and which has all required skills. trying to measure other, elite, AA+ forms to the 86 bp AA+ form is like trying to measure a skyscraper in relationship to a human. where the first analogy would give a value more like 1/284th(the average skyscraper being around 1700 feet, a human of 6 feet would be about 1/284th the size), the second analogy would give a more natural value of 284.

other than that, i like the shape that it is taking ^-^

Pc Sc are primary and secondary combos; I got rid of rarity/level altogether since we're talking about AA forms here so those are assumed given.

Pc are primary combos; or the largest stat affecting combos namely guild, species, and wings; having all three can net 3 Pc and gain a combo score of 6; which would mean that the form in question is a 5x rarity, level, winged, species, guild formation (e.g. sin or covert winged wyrms, psycho winged brutes)

Sc are secondary combos; or miscellanea like bp, skill, or the X-Y-Z species combos which do provide tangible but less substantial benefits than their primary counterparts.

I don't quite understand what you're saying about the bp part... your skyscraper analogy appears to have gone over my head

could you give me an example of how you would rate the bp of the following formations? and what you would consider an optimal bp rating? (with a formation example; e.g. I used psy winged brutes with skillchanges as the baseline to determine optimal in my initial draft)

bp rating of an 80bp form
bp rating of a 150bp form
bp rating of a 200bp form

you can still consider say 200 optimal, but by comparing it to a singular unit. in my proposed example, i base bp score based on how many times stronger is the form compared to the unit form, aka the 86 bp.

the whole business of "ceiling(bp/86)" is to eliminate fractional parts. ceiling(2.7) and ceiling(2.4) both equal 3, and ceiling(2) equals 2. max possible value is 6 or 5 times, depending on whether you consider the possibility of a 500 bp form  to be the max bp a form could possibly have(which is 5.5 times stronger than a 86 bp form, approximately) or 375(accounting for the fact that a AAA+ is about as strong as a 75 bp AA+, so 75*5=375, so max is 4.something times a 86 bp form's strength, so ceiling of that is 5)

not too sure that the combo weighing is a good idea, particularly considering that winged is somehow seen as a primary combo when it has a very localized effect on only attack. personally, i consider winged vastly overrated, it should be a secondary(if we're sticking with assigning weights to combos).

additionally, i still hold that skill changed forms should lose some points for the fact that they do in fact require skill changes to reach the same potential as another form with similar stats but which does not require skill changes. yes, that includes my own, which does feature two skill changes.


edit: regarding the bp ratings under my proposal:

80 bp form, since it is less than 86, gets 0
150 bp form gets 2/6(or 2/5 if max bp possible is 375)
200 bp form gets 3/6(or 3/5 if max bp possible is 375)

Sorry this response took so long! had a bunch of irl shenanigans happen which needed to be contended with

Armitaage wrote:you can still consider say 200 optimal, but by comparing it to a singular unit. in my proposed example, i base bp score based on how many times stronger is the form compared to the unit form, aka the 86 bp.

the whole business of "ceiling(bp/86)" is to eliminate fractional parts. ceiling(2.7) and ceiling(2.4) both equal 3, and ceiling(2) equals 2. max possible value is 6 or 5 times, depending on whether you consider the possibility of a 500 bp form  to be the max bp a form could possibly have(which is 5.5 times stronger than a 86 bp form, approximately) or 375(accounting for the fact that a AAA+ is about as strong as a 75 bp AA+, so 75*5=375, so max is 4.something times a 86 bp form's strength, so ceiling of that is 5)

edit: regarding the bp ratings under my proposal:

80 bp form, since it is less than 86, gets 0
150 bp form gets 2/6(or 2/5 if max bp possible is 375)
200 bp form gets 3/6(or 3/5 if max bp possible is 375)

I think I understand what you're saying here, however as I said earlier; this is a rating format exclusively for AA formations, AAAs or AAA/AA hybrid formations would require a different system than what I'm working on.

Armitaage wrote:
not too sure that the combo weighing is a good idea, particularly considering that winged is somehow seen as a primary combo when it has a very localized effect on only attack. personally, i consider winged vastly overrated, it should be a secondary(if we're sticking with assigning weights to combos).

It may be vastly overrated, but it is also vastly overpowered

When dealing with relatively similar teams (a.k.a. top end ~180+bp teams with a reasonable array of skills) Dark Summoner is first move win. Wings grant considerably cheaper and arguably weaker teams the ability to first move win against much more expensive and well balanced formations.

I had my doubts initially, but the tremendous amount of battles outputted in clan wars changed my opinion decidedly in favor of wings being a cheap but devastating alternative to adhering to traditional species/guild/skill combos as opposed to a gimmick. A formation with wings can attack and win well outside of its "weight class"

On top of that; the mechanics of this game reward offensive wins more than defensive wins
e.g.
on the offensive;
you get honor if you win, you get gold, you get CP if you win, you get crystals, points in a battle event, also in mobius/vl type events where skills don't take effect wings are your best friend for efficient progress

on the defensive;
best case scenario: you get some gold (not even that sometimes when they disable offensive defeat gold loss)

therefore I think that wings DO deserve designation as a primary combo due to the list of aforementioned benefits

Armitaage wrote:
additionally, i still hold that skill changed forms should lose some points for the fact that they do in fact require skill changes to reach the same potential as another form with similar stats but which does not require skill changes. yes, that includes my own, which does feature two skill changes.

That's an interesting point, however I think that should be considered more in a cost benefit analysis than actual combat utility.

And I also think consideration of skill changes as a major cost is a somewhat antiquated notion. There was a time where it really made sense to look for the prime natural formations when skill changes were astronomically expensive, and most AAs had good skills, but the flood of no-skill freebie event AAs, highbp crap-skill AAs, and precipitous drop of skill change prices due to their nearly constant availability in weekly battle, and crash post-forgotten island 1 makes them essentially a standard operating cost for all but the least affluent and experienced players.

how about if a formation has more than 200bp, you subtract the difference and decrease bp rating accordingly?

bp bloated formations are a problem since they cost so much to use which cuts into how useful they are based on how often you can use them?

like a 250bp formation would be looked at as a 200- (250-200) = 150bp formation?

also for naming the different sections, nobody wants deficient, beginner, intermediate, or standard

so why not.......

97-100: platinum
84-97: gold
50-84: silver

and not assign a name to <50 at all?

also I get the idea about using the normal distribution amounts for the tiers, but there's not enough formations for that to really work well imo

so why not stagger the distribution a bit?

like
84-100: platinum
66-83: gold
50-65: silver

If this equation is solely for AA+ forms, then atk is not really the primary stat to concern over, but rather defense. So, despite the obvious gains for winning offensive battles, combos, such as winged, that only affect atk would be better and more accurately placed as secondary. Winged or not, what really determines if you win offensive battles is the defense of both teams and the skill set. Due to combo distribution, atk is always higher than your own defense, but not always higher than the enemy's. As an example, consider what would happen if your 143 bp Impulse winged wyrms were to face a 220+ bp psycho form with Impulse down(it's quite common, and psycho forms of that bp generally have 110k+ def). Despite your higher attack, without Psycho down, you will lose. A form with winged, under combo weighing, deserves to be higher rated than a form without it, but not by such a huge margin when winged is a) quite rare and b) easy to neutralize(triply so due to relica. My own reduces enemy winged monster's health by 10%, so facing me with winged isn't that great of an idea) 

On the subject of skill changes, cost aside, it still requires an extra step to cover up skill deficiencies using a skill change so I still hold that skill changed forms should take a penalty for that.

Armitaage wrote:If this equation is solely for AA+ forms, then atk is not really the primary stat to concern over, but rather defense. So, despite the obvious gains for winning offensive battles, combos, such as winged, that only affect atk would be better and more accurately placed as secondary. Winged or not, what really determines if you win offensive battles is the defense of both teams and the skill set. Due to combo distribution, atk is always higher than your own defense, but not always higher than the enemy's. As an example, consider what would happen if your 143 bp Impulse winged wyrms were to face a 220+ bp psycho form with Impulse down(it's quite common, and psycho forms of that bp generally have 110k+ def). Despite your higher attack, without Psycho down, you will lose. A form with winged, under combo weighing, deserves to be higher rated than a form without it, but not by such a huge margin when winged is a) quite rare and b) easy to neutralize(triply so due to relica. My own reduces enemy winged monster's health by 10%, so facing me with winged isn't that great of an idea)

On the subject of skill changes, cost aside, it still requires an extra step to cover up skill deficiencies using a skill change so I still hold that skill changed forms should take a penalty for that.

while that may have certainly been true at some point (wings being gimmicky), combined with the proliferation of skill changes; which have made obtaining a balanced skillset nearly trivial (with the singular exception of random skill souls to get health up, unless you're running winged brutes) I would say that wings on their own won't win the entire battle, but a winged platform can go further and for substantially less.

143bp impulse wyrms were rated very poorly in the initial system; 16! (due to skill x member multiplication and a truly pathetic skillset) and no similar rarity/level (unless the 220bp team are all As, or all level 1) 143bp team will beat a 220bp team. However; in clan wars; I ran the following 143bp formation (largely to get more attacks out) consisting of raminas, banished, moonc(DEA), kanna, venus and experienced a surprisingly large win margin; the most surprising part of which was the fact that a lot of these were against ~180bp formations WITH health up. Some were 50/50 wins in the prolonged battles [e.g. I win this time, they win the next time, etc etc... crits and misses] but the fact of the matter was that wings allowed an incredibly cheap and lightweight formation to play quite competitively with heavy expensive formations.

however, I do concede that some sort of interplay should be included for skill/bp/combo in a multiplicative sense... as 61 is a clearly incorrect rating for sin wyrms (a 143bp AA form with only IPA and DED is NOT a good form), and while sin wyrms scored very poorly on skills and bp, it did get full marks on combos...


In regards to skill changes? also feeding off of what I just said about wings becoming surprisingly relevant due to the proliferation of skill changes; that sounds like the job of a separate rating system altogether; a cost-benefit one of some form. There really are monsters which aren't particularly useful in their natural state which are platforms with tremendous potential once skillchanged and perhaps used in cross guild combos

Moving it to secondary combo would still net wing wyrms full marks on combos, but it would also slightly raise the score of other, perfectly fine forms in the combo department, which I think is fairer. Particularly since you admitted to low bp winged forms depending on chance against higher bp forms. Also, the form you listed lacks an answer to Impulse down(which is quite a common skill in Psycho and Covert forms) which greatly diminishes it's strength. 

The skill change penalty I keep pushing for is intended to reflect the inherent weakness of a form that requires it as opposed to one that doesn't, cost-benefit notwithstanding. The fact that without the skill change(s) the form would be rated much lower deserves being taken into account, imo.

Armitaage wrote:Moving it to secondary combo would still net wing wyrms full marks on combos, but it would also slightly raise the score of other, perfectly fine forms in the combo department, which I think is fairer. Particularly since you admitted to low bp winged forms depending on chance against higher bp forms. Also, the form you listed lacks an answer to Impulse down(which is quite a common skill in Psycho and Covert forms) which greatly diminishes it's strength. 

The skill change penalty I keep pushing for is intended to reflect the inherent weakness of a form that requires it as opposed to one that doesn't, cost-benefit notwithstanding. The fact that without the skill change(s) the form would be rated much lower deserves being taken into account, imo.

I'm not placing wings in primary combo to stroke my own ego by creating a rating system solely for the purpose of giving my formation a high rating (In either iteration of the rating system, my own formation was decidedly low ranked anyways; 45 in the first, well within the "standard" or intermediate in the second)

low bp winged forms do depend on chance against insurmountable opponents which un-winged forms would lose 100% of the time against. Additionally; they do not depend on chance against ALL higher bp forms; only the ones with good skill distributions. Against poor skill distribution; winged lowbp will beat slightly higher bp consistently. I can drop wings to a combo category of its own weighted as 1.5 (as opposed to 2 for guild and species, or 1 for things like bp combo, the X-Y-Z species combos, skill combo, etc); which brings the combo maximum to 5.5 as opposed to 6 [this entire system was built on the basis that psycho winged brutes are the optimal formation; formerly on account of the only species/guild having a winged health up;  now somewhat moot due to random skill changes]

anyways, with wings devalued...



as opposed to the former



the changes made being;




anyways, this doesn't address the larger issue at hand of non-indicative scores which allow an unreasonable bonus based on stats (e.g. sin wyrms basic package are still highly rated, and it shouldn't be) so if we're done with the wings, I think we need to start looking for a fair multiplication between stats (as a combination of combo and bp) and skill rating

Ok, this change does make things more fair. As for rebalancing, how about doing the 4/3/3 as bp/skills/combos instead of skills/bp/combos? As you point out, skills can only go so far.


Edit: any way I look at it, so long as the bp score formula is bp/200, wing wyrms of 143 bp score about 60-67. Multiplication based formulae would suffer the same non-representative issue as your older formula, with the possibility of a 0 score if any of the scores being multiplied is 0. So let's shift focus to refining the bp formula. 

Edit 2: however, the bigger problem with the current formula isn't the rating of wing wyrms, but of the 86 bp form. Under 4/3/3 skills/bp/combos it rates 69.2, higher than wing wyrms. However, under 4/3/3 bp/combos/skills it rates 63.56 to wing wyrms' 66.10. I say that this is a better weight assignment, 40% bp, 30% skills, and 30% combos. Now we just need to fine tune the bp formula and we're set, I think.