1. Welcome to Smogon Forums! Please take a minute to read the rules.
  2. New to the forums? Check out our Mentorship Program!
    Our mentors will answer your questions and help you become a part of the community!

Typing Rating

Discussion in 'Create-A-Pokémon Project' started by X-Act, Dec 1, 2009.

  1. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    I decided to try my hand into a typing rating. I think I've already nailed the required formulae, but I won't be as bold as saying that these are the final formulae until I have time to research this thing further.

    First of all, we need to consider two facets to typing: its offensive nature and its defensive nature. Its offensive nature is due to the STAB bonus it offers, while its defensive nature is basically due to the type chart.

    It should be noted that the offensive nature of the typing of a Pokemon shouldn't note movepool at all. For example, just because we have a Fighting-type Pokemon does not mean that it will have Close Combat. Conversely, just because we have a Normal/Flying type Pokemon does not mean that it won't have Close Combat (see: Staraptor). Hence, as I like to repeat over and over again, the offensive nature of a Pokemon is affected the most by its movepool, and not by its typing or stats.

    Without further ado, here's the Offensive Typing Rating (OTR) formula:

    Code:
    OTR = (72672 + Type1 + Type2) / 68680
    where Type1 and Type2 are the two types of the Pokemon whose number corresponds to the table below:

    Code:
    Type    Number
    ---------------
    None         0
    Bug       2080
    Dark      2171
    Dragon    2046
    Electric  2200
    Fighting  2213
    Fire      2253
    Flying    2351
    Ghost     1902
    Grass     2193
    Ground    2257
    Ice       2442
    Normal    1804
    Poison    1820
    Psychic   2058
    Rock      2522
    Steel     1857
    Water     2167
    As you can see, the OTR's range is very small, from a minimum of around 1.0844 to a maximum of around 1.1304, which indicates that typing does not contribute much to the offensive nature of a Pokemon.

    Now Defensive Typing Rating (DTR), which provides a number corresponding to how defensive the typing is:

    Code:
    DTR = 68 / Type
    where Type corresponds to the typing of the Pokemon whose number corresponds to the table below:

    Code:
    Pokemon Typing     Type
    ------------------------
    Bug / None           74
    Bug / Dark           72
    Bug / Dragon         73
    Bug / Electric       68
    Bug / Fighting       78
    Bug / Fire           77
    Bug / Flying         84
    Bug / Ghost          72
    Bug / Grass          96
    Bug / Ground         74
    Bug / Ice            94
    Bug / Normal         72
    Bug / Poison         74
    Bug / Psychic        83
    Bug / Rock           76
    Bug / Steel          57
    Bug / Water          70
    Dark / None          68
    Dark / Dragon        68
    Dark / Electric      66
    Dark / Fighting      65
    Dark / Fire          68
    Dark / Flying        66
    Dark / Ghost         54
    Dark / Grass         84
    Dark / Ground        72
    Dark / Ice           86
    Dark / Normal        74
    Dark / Poison        60
    Dark / Psychic       76
    Dark / Rock          84
    Dark / Steel         60
    Dark / Water         68
    Dragon / None        68
    Dragon / Electric    67
    Dragon / Fighting    70
    Dragon / Fire        68
    Dragon / Flying      73
    Dragon / Ghost       64
    Dragon / Grass       85
    Dragon / Ground      74
    Dragon / Ice         78
    Dragon / Normal      68
    Dragon / Poison      69
    Dragon / Psychic     76
    Dragon / Rock        77
    Dragon / Steel       49
    Dragon / Water       64
    Electric / None      66
    Electric / Fighting  66
    Electric / Fire      73
    Electric / Flying    62
    Electric / Ghost     62
    Electric / Grass     75
    Electric / Ground    72
    Electric / Ice       78
    Electric / Normal    66
    Electric / Poison    70
    Electric / Psychic   74
    Electric / Rock      81
    Electric / Steel     58
    Electric / Water     65
    Fighting / None      70
    Fighting / Fire      71
    Fighting / Flying    71
    Fighting / Ghost     65
    Fighting / Grass     84
    Fighting / Ground    75
    Fighting / Ice       82
    Fighting / Normal    70
    Fighting / Poison    75
    Fighting / Psychic   72
    Fighting / Rock      80
    Fighting / Steel     55
    Fighting / Water     70
    Fire / None          70
    Fire / Flying        72
    Fire / Ghost         67
    Fire / Grass         73
    Fire / Ground        72
    Fire / Ice           85
    Fire / Normal        70
    Fire / Poison        76
    Fire / Psychic       76
    Fire / Rock          87
    Fire / Steel         60
    Fire / Water         69
    Flying / None        70
    Flying / Ghost       69
    Flying / Grass       85
    Flying / Ground      70
    Flying / Ice         84
    Flying / Normal      68
    Flying / Poison      69
    Flying / Psychic     77
    Flying / Rock        74
    Flying / Steel       48
    Flying / Water       70
    Ghost / None         64
    Ghost / Grass        72
    Ghost / Ground       69
    Ghost / Ice          74
    Ghost / Normal       56
    Ghost / Poison       68
    Ghost / Psychic      80
    Ghost / Rock         75
    Ghost / Steel        47
    Ghost / Water        64
    Grass / None         80
    Grass / Ground       84
    Grass / Ice          96
    Grass / Normal       80
    Grass / Poison       75
    Grass / Psychic      92
    Grass / Rock         80
    Grass / Steel        59
    Grass / Water        73
    Ground / None        72
    Ground / Ice         82
    Ground / Normal      72
    Ground / Poison      71
    Ground / Psychic     80
    Ground / Rock        93
    Ground / Steel       57
    Ground / Water       68
    Ice / None           82
    Ice / Normal         86
    Ice / Poison         80
    Ice / Psychic        88
    Ice / Rock          100
    Ice / Steel          73
    Ice / Water          79
    Normal / None        68
    Normal / Poison      66
    Normal / Psychic     70
    Normal / Rock        84
    Normal / Steel       60
    Normal / Water       68
    Poison / None        68
    Poison / Psychic     73
    Poison / Rock        81
    Poison / Steel       58
    Poison / Water       66
    Psychic / None       76
    Psychic / Rock       86
    Psychic / Steel      55
    Psychic / Water      76
    Rock / None          80
    Rock / Steel         72
    Rock / Water         81
    Steel / None         54
    Steel / Water        52
    Water / None         68
    Interestingly, the most defensive typing is Ghost/Steel (hello Kitsunoh), while the least defensive typing is Ice/Rock, which is so far non-existent. However, STAB-wise, Ice/Rock provides the best OTR, while Ghost/Steel ranks very low...

    If we incorporate also the fact that types like Ice and Rock tend to deal more damage than types like Poison and Normal, the above chart changes to:

    Code:
    Pokemon Typing         Type
    ------------------------------
    Bug / None            86.1886
    Bug / Dark            84.5949
    Bug / Dragon          85.3095
    Bug / Electric        79.2046
    Bug / Fighting        90.5112
    Bug / Fire            92.8707
    Bug / Flying         102.2531
    Bug / Ghost           85.0204
    Bug / Grass          113.2103
    Bug / Ground          86.3917
    Bug / Ice            111.4689
    Bug / Normal          85.0427
    Bug / Poison          86.7143
    Bug / Psychic         95.3202
    Bug / Rock            87.8140
    Bug / Steel           68.4186
    Bug / Water           81.3996
    Dark / None           77.6204
    Dark / Dragon         78.0507
    Dark / Electric       75.8424
    Dark / Fighting       74.5249
    Dark / Fire           79.7530
    Dark / Flying         78.0531
    Dark / Ghost          63.0470
    Dark / Grass          96.0619
    Dark / Ground         83.2983
    Dark / Ice            98.8909
    Dark / Normal         85.4490
    Dark / Poison         69.5913
    Dark / Psychic        85.3852
    Dark / Rock           96.4983
    Dark / Steel          71.2178
    Dark / Water          77.6767
    Dragon / None         77.2439
    Dragon / Electric     76.6520
    Dragon / Fighting     79.2658
    Dragon / Fire         78.6951
    Dragon / Flying       86.8498
    Dragon / Ghost        73.6753
    Dragon / Grass        98.3177
    Dragon / Ground       86.4811
    Dragon / Ice          88.3026
    Dragon / Normal       78.4985
    Dragon / Poison       79.0663
    Dragon / Psychic      85.1754
    Dragon / Rock         87.9060
    Dragon / Steel        57.7592
    Dragon / Water        72.0425
    Electric / None       75.0356
    Electric / Fighting   74.3483
    Electric / Fire       85.1276
    Electric / Flying     72.6233
    Electric / Ghost      71.4670
    Electric / Grass      85.2161
    Electric / Ground     83.0859
    Electric / Ice        89.8150
    Electric / Normal     76.2902
    Electric / Poison     80.6730
    Electric / Psychic    82.9671
    Electric / Rock       93.7707
    Electric / Steel      69.0284
    Electric / Water      73.5647
    Fighting / None       78.8355
    Fighting / Fire       81.0314
    Fighting / Flying     81.7319
    Fighting / Ghost      74.0170
    Fighting / Grass      96.3336
    Fighting / Ground     86.0722
    Fighting / Ice        92.1875
    Fighting / Normal     80.0901
    Fighting / Poison     84.6915
    Fighting / Psychic    80.2603
    Fighting / Rock       90.2034
    Fighting / Steel      64.6240
    Fighting / Water      78.8918
    Fire / None           81.0668
    Fire / Flying         86.5780
    Fire / Ghost          78.5585
    Fire / Grass          84.1158
    Fire / Ground         82.9504
    Fire / Ice           101.0548
    Fire / Normal         82.3214
    Fire / Poison         88.9432
    Fire / Psychic        86.8777
    Fire / Rock          101.9374
    Fire / Steel          71.7999
    Fire / Water          80.0148
    Flying / None         81.7674
    Flying / Ghost        81.6595
    Flying / Grass       100.9962
    Flying / Ground       83.0717
    Flying / Ice          99.1488
    Flying / Normal       80.6215
    Flying / Poison       80.8566
    Flying / Psychic      88.7785
    Flying / Rock         86.0931
    Flying / Steel        57.3711
    Flying / Water        81.2881
    Ghost / None          73.2450
    Ghost / Grass         83.7012
    Ghost / Ground        79.7348
    Ghost / Ice           84.9138
    Ghost / Normal        66.1524
    Ghost / Poison        78.1613
    Ghost / Psychic       89.6230
    Ghost / Rock          84.9281
    Ghost / Steel         55.4392
    Ghost / Water         73.3013
    Grass / None          91.0140
    Grass / Ground        98.6276
    Grass / Ice          109.5367
    Grass / Normal        92.2686
    Grass / Poison        86.5954
    Grass / Psychic      103.1866
    Grass / Rock          91.1146
    Grass / Steel         70.4714
    Grass / Water         82.0806
    Ground / None         82.4916
    Ground / Ice          92.9207
    Ground / Normal       83.7461
    Ground / Poison       81.5899
    Ground / Psychic      90.4230
    Ground / Rock        106.6983
    Ground / Steel        67.1932
    Ground / Water        77.2931
    Ice / None            93.2832
    Ice / Normal          99.3387
    Ice / Poison          91.5266
    Ice / Psychic         98.8143
    Ice / Rock           113.0537
    Ice / Steel           86.5132
    Ice / Water           90.6227
    Normal / None         78.0682
    Normal / Poison       76.3115
    Normal / Psychic      80.0529
    Normal / Rock         96.9461
    Normal / Steel        71.4416
    Normal / Water        78.1245
    Poison / None         77.4574
    Poison / Psychic      82.3926
    Poison / Rock         92.5822
    Poison / Steel        68.6366
    Poison / Water        75.1564
    Psychic / None        84.7451
    Psychic / Rock        96.4217
    Psychic / Steel       63.9781
    Psychic / Water       84.8014
    Rock / None           90.8907
    Rock / Steel          84.7249
    Rock / Water          92.9135
    Steel / None          63.6131
    Steel / Water         61.1307
    Water / None          76.8699
    We still see that Ghost/Steel is top-notch, but the worst defensive type now is Bug/Grass. Poor Parasect. Also notice how all the numbers have increased from the previous table.

    For the above table, DTR is then found by:

    Code:
    DTR = 36336 / (Type * 505)

    Anyway, the plan is to multiply the Sweepiness by the OTR and the Tankiness by the DTR to obtain a Base Stats + Typing Rating in the future... :)
  2. The_Chaser

    The_Chaser RVW
    is a Researcher Alumnusis a Battle Server Moderator Alumnusis a Past SPL + WCoP Winner

    Joined:
    Jun 9, 2009
    Messages:
    919
    Great Work, X-Act! I will definately use this in the future as I generally like offensive sweepers with good defences and typing.
  3. DougJustDoug

    DougJustDoug Knows the great enthusiasms
    is a member of the Site Staffis an Artistis a Programmeris a CAP Contributoris an Administratoris a Battle Server Admin Alumnusis a Smogon IRC SOp Alumnusis a Tiering Contributor Alumnus
    CAP Leader

    Joined:
    Jun 26, 2007
    Messages:
    3,043
    I'd like to see if there is some way to incorporate this, or some derivative of this, into the CAP process. I really have no clue how to do that. But here's my motivation -- our current type selection process is a total crapshoot. We just randomly throw out potential types, with the hope that it will support our concept, and we take it from there. If we could somehow put a more analytical framework underneath our type discussions and selection process -- perhaps we could add a bit more structure and reason to our project.

    I have a lot of random thoughts about this right now, and I don't want to go in a million directions at once. But, perhaps others can think about this stuff too and we might be able to come up with something together. Or maybe not. At the very least, it could be an interesting discussion!

    Once again -- great work X-Act! I love when you present this kind of thing. It's always fascinating to see numbers that support our common "gut feelings" about this game we play.
  4. cyberzero

    cyberzero
    is a CAP Contributor Alumnus

    Joined:
    May 27, 2008
    Messages:
    1,010
    This is pretty interesting, X-Act, but I was wondering where the numbers for the DTR came from. It seems as though the more resistances one typing has, the lower the number assigned it is, but how did you come up with numerical values for each, especially the non-existent (as of now) types?

    Steel and Steel/Water both have 11 resistances, 1 immunity, 2 neutrals and 3 super effective moves against them, yet the numerical values for them are slightly different. Is this because of move type frequency in the metagame or some other reason?
  5. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    DTR was actually much more easy to calculate than OTR.

    For DTR, I looked at the type chart of that particular typing, and summed up the numbers, literally.

    So, for Steel, we have:

    1/2 + 1/2 + 1/2 + 1 + 2 + 2 + 1/2 + 1/2 + 1/2 + 2 + 1/2 + 1/2 + 0 + 1/2 + 1/2 + 1/2 + 1 = 13.5

    For Water/Steel, we first multiply out the corresponding columns and then add them up (this is easily done by Excel by the SUMPRODUCT function). So we have:

    1/2 + 1/2 + 1/2 + 2 + 2 + 1 + 1/2 + 1/2 + 1 + 2 + 1/4 + 1/2 + 0 + 1/2 + 1/2 + 1/4 + 1/2 = 13

    The numbers listed in the table are the numbers above multiplied by 4, so that they become a whole number. (I prefer using whole numbers rather than decimals). As you can see, Water/Steel is slightly better than Steel because it has two double resistances (to Steel and to Ice) as opposed to mono-Steel, which doesn't.
  6. Purple Wobbuffet

    Purple Wobbuffet

    Joined:
    Sep 17, 2009
    Messages:
    172
    DougJustDoug, I think we could incorporate it by doing this: Suppose our concept is Wallbraker. Everyone knows that it is easier to break walls if you get STAB on the moves you are using to break them. So we say, "You can not suggest a type with an OTR below 2100 (I don't know what this number should be, I just think that it is a good number)." That elimates the typings Bug, Dragon, Normal, Poison, Psychic and Steel. Name some walls that take super effective damage from these types.

    Anyway, an Ice/Rock sweeper wouldn't work. In four words I will tell you why:
    Scizor used Bullet Punch.
    Ice/Rock, believe it or not, is the only possible typing that can be 4x weak to Steel.
  7. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    Just wanted to note again that OTR is only a number between 1.0844 and 1.1304, which means that typing does not really affect the offense of a Pokemon much.
  8. DKG

    DKG

    Joined:
    Oct 9, 2009
    Messages:
    30
    Umm noob here, anyone care to explain to me what this is?
  9. zarator

    zarator Credits to Mos-Quitoxe for the cute sprite^^
    is a Community Contributoris a Contributor Alumnus

    Joined:
    Mar 12, 2008
    Messages:
    4,975
    Nice work X-Act, but could you care to explain more in-depth the OTR formula? Where do "72672" and "68680" come from? And also how did you calculate the OTR of single types? Thanks for your time^^
  10. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    Sure.

    I considered that a hypothetical Pokemon can have any type of move in its movepool. Be it a Dark/Flying Pokemon or a Water/Dragon Pokemon, they both can learn Dragon Pulse, say. Hence I needed to consider the offense of every Pokemon from each of the 17 types. This would be equal for all Pokemon. To do this, I first took the typing of all the existing Pokemon into account, and calculated the type chart for each Pokemon. Then I added up all the Bug, Fire, Flying, etc. types separately. When I found the average of these type offenses, I got the offense for each Pokemon without taking STAB into account, and I got the number 72672/68680 (I actually got a more simple fraction but I prefer working with whole numbers, as I said before... in fact, 68680 = 17 * 505 * 8 [number of types * number of Pokemon * 8]).

    Then, since STAB multiplies the offense of that particular type by 1.5, all I needed to do is to add half of the value for that particular type, because I had already considered that particular type to be multiplied by 1. So I divided each of the numbers for the types I found before by 2, and I got the numbers in the table.

    Let's give an example. If I have a Water/Ice Pokemon, I considered that if it attacked by a Water move, its power would be 1.5 times the usual power, and the same thing would happen if it were an Ice move. For all the other moves it uses, the power would be the usual amount of power. Hence its ODB is:

    Bug x 1 + Dark x 1 + Dragon x 1 + Electric x 1 + Fighting x 1 + Fire x 1 + Flying x 1 + Ghost x 1 + Grass x 1 + Ground x 1 + Ice x 1.5 + Normal x 1 + Poison x 1 + Psychic x 1 + Rock x 1 + Steel x 1 + Water x 1.5

    And this can be written as

    (Bug x 1 + Dark x 1 + Dragon x 1 + Electric x 1 + Fighting x 1 + Fire x 1 + Flying x 1 + Ghost x 1 + Grass x 1 + Ground x 1 + Ice x 1 + Normal x 1 + Poison x 1 + Psychic x 1 + Rock x 1 + Steel x 1 + Water x 1) + Ice x 0.5 + Water x 0.5

    The thing in brackets will be the same for every Pokemon, and I found that to be equal to 72672 / 68680. Ice x 0.5 and Water x 0.5 can then be found from the table (and then divided by 68680 as well). This way, I simplified the calculation considerably.
  11. Cartoons!

    Cartoons! !!!
    is an Artist Alumnusis a Forum Moderator Alumnus

    Joined:
    Nov 2, 2007
    Messages:
    1,055
    Skarmory has the best DTR of all existing Pokemon, but Special Fire and Electric moves are not all that uncommon. Is there any way of combining your DTR results with the actual % of use of specific types of moves in the OU metagame?


    Results like these are a great read, and worth keeping in mind when creating our CAPs, but I'd hesitate to consider applying a formulaic approach when it comes to deciding Type. With our concept tending to be more complex than just 'sweeper' or 'wall', how do you justify calculating an optimal Typing one way or the other without some serious polljumping?

    In any case, Typing is one of those attributes we deem least critical to the success or failure of our CAPs. Syclant was at one point monstrous with its Bug/Ice combination, today it is going through a Revision process in order to buff it up. Even there, changing it's Typing to something that might help it deal more easily with the Scizor problem is not on the table.
  12. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    It would be possible to do this, yes, but we need to be as objective as possible here. And that means that the typing rating (and the base stats ratings also) should be independent of the metagame the Pokemon are used in.

    For example, you mention that Fire and Electric moves are common in the metagame. Maybe they're common just because Steel types are such good defensive types? In other words, the fact that Fire moves are common in a metagame does not make Steel types any less defensive - if anything, it underlines the fact that, since they are so defensive, drastic measures are required to stop them, and one of these measures is, apparently, spamming Fire moves left, right and centre.
  13. Purple Wobbuffet

    Purple Wobbuffet

    Joined:
    Sep 17, 2009
    Messages:
    172
    In response to X-Act's post after mine, doesn't STAB affect the damage a move does by 1.5? So shouldn't this affect the OTR? If we made two pokemon, one having Ice/Rock typing and the other having Normal/Poison typing, but everything else the same, it's obvious that the Ice/Rock would do better than the Normal/Poison by a good amount, because of its STABs.
  14. zarator

    zarator Credits to Mos-Quitoxe for the cute sprite^^
    is a Community Contributoris a Contributor Alumnus

    Joined:
    Mar 12, 2008
    Messages:
    4,975
    Wow, thanks for the explaination man^^ Where do you teach? One of these days I should attend a course of yours after I finish with philosophy!:-)
  15. petrie911

    petrie911

    Joined:
    Aug 27, 2005
    Messages:
    859
    X-Act, is it not reasonable to assume that moves that have good offensive type will be used more than moves of poor offensive type? As such, it would seem like having a resistance to a good offensive type would be more useful than a resistance to a poor one. After all, Hippowdon's rock resist is a very important part of its walling ability. Its Poison resist...not so much.

    A possible way to handle this is to divide each weakness/resistance by the offensive type rating of each type in question. In other words,

    DTR ~ 1/sum(type1mod*type2mod/OTR)

    Where the sum is carried out over all types and ~ here means "is proportional to".

    This also reminds me of two other modifications. Abilities that affect type interaction (Flash Fire, Dry Skin, Thick Fat, Levitate, etc), and Stealth Rock weakness should probably factor in somehow as well.
  16. familyguyman

    familyguyman

    Joined:
    Apr 6, 2009
    Messages:
    499
    I am glad you went through with this and I am liking the results! Objectifying things we already know from experience is a pretty great thing, now we have another dimension to add to our CAP process.

    *Is happy*
  17. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    The OTR formula provides a higher number for an Ice/Rock Pokemon than for a Normal/Poison Pokemon already, so what is the problem?

    (Also I'd argue that an Ice/Rock Pokemon would always be a better attacker than a Normal/Poison Pokemon - you're assuming that an Ice/Rock Pokemon will have a good Ice move and a good Rock move, which, while a natural assumption, is not always the case [see: Flareon]. This emphasizes once again that movepool is what makes a Pokemon offensive.)

    I'll think about this.
  18. armogohma

    armogohma

    Joined:
    Sep 9, 2008
    Messages:
    74
    I was just about to bring that up too.

    Imagine two hypothetical types, henceforth known as A and B. A is immune to ice and rock, and weak to poison, while B resists poison, normal, steel, and ghost, and is weak to ground. Both have defensive typing ratings of 64, yet I'm pretty sure A is a better type, especially because of the stealth rock immunity.

    If we were to adjust the numbers used in the ratios to signify just how useful a resist or immunity to a specific type is, or just how bad each specific weakness is, the DTR would probably be a bit more accurate.
  19. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    I agree with you after thinking about it, and I've managed to fix it. Your particular example now yields a DTR of 65.727 for Type A and 69.112 for Type B.

    I'll update the table in the original post shortly.
  20. Deucalion2

    Deucalion2

    Joined:
    Feb 23, 2007
    Messages:
    575
    Would it be possible to make a "combined rating", much like you did for the Base Stats Rating? I'd like to see which typing is the overall best.
  21. The Great and Powerful Oz

    The Great and Powerful Oz

    Joined:
    Dec 3, 2009
    Messages:
    8
    Ha! Very interesting concept, nice work. I'll probably end up consulting the chart in the future.

    The obvious point that's been mentioned is the divide between theory and results in the actual environment. Strategies will compensate and increase the priorities of some typing and decrease others. And being a dynamic environment, this constantly shifts. (Not saying anything that isn't obvious)

    I can see this chart as a useful tool looking from a broad perspective.
    Of course you can't use it as the definitive answer of course, because that's where this project's creative and innovative sides come to play.

    Long time lurker, thought I might as well go ahead and say something.
    Thanks for your work.
  22. armogohma

    armogohma

    Joined:
    Sep 9, 2008
    Messages:
    74
    I was thinking, since DTR=68/type, and "type" is so much larger when adjusted (for instance, 68->78 for normal), should the DTR equation be changed for adjusted stats as well? Maybe something like DTR=~78/type? Because there are several pretty defensive types with adjusted DTRs above 1 (even bug/steel), while the only typings that really should be above 1 are the defensively below average ones.
  23. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    68 was chosen for a very specific reason. Originally, it was 17 (the number of different types), but since I had multiplied each number on the table by 4 to make them whole numbers, I had to multiply 17 by 4 as well to compensate. Now that the numbers are all weird decimals, this step is redundant, but we can't just replace the 68 without changing the decimals written on the table as well.
  24. X-Act

    X-Act np: Biffy Clyro - Shock Shock
    is a Site Staff Alumnusis a Programmer Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis an Administrator Alumnus

    Joined:
    Feb 17, 2006
    Messages:
    4,675
    The more I look at the new chart, the more I'm satisfied by it.

    Here are the new numbers for the DTRs that previously all corresponded to '68':

    Code:
    Typing           New DTR  Old DTR
    ----------------------------------
    Water / None     76.8699       68
    Dragon / None    77.2439       68
    Ground / Water   77.2931       68
    Poison / None    77.4574       68
    Dark / None      77.6204       68
    Dark / Water     77.6767       68
    Dark / Dragon    78.0507       68
    Normal / None    78.0682       68
    Normal / Water   78.1245       68
    Ghost / Poison   78.1613       68
    Dragon / Normal  78.4985       68
    Dragon / Fire    78.6951       68
    Bug / Electric   79.2046       68
    Dark / Fire      79.7530       68
    Flying / Normal  80.6215       68
    Mono-Water now becomes the best out of these, since it resists Ice and Fire and is neutral to Rock. Normal/Flying becomes the worst since it is weak to both Ice and Rock and is neutral to Fire. Exactly as we wanted!

    So thank you to all those people who suggested improving the DTR measure.

    By the way, to get these new numbers, I took the type chart again, but, before I summed up the values in the type chart corresponding to the types, I first multiplied each value by the number of the OTR chart of the type that's attacking it divided by 2020 (= 505 x 4). For example, if a type was neutral to Bug, I added the OTR corresponding to Bug in the OTR chart divided by 2020 (= 2080/2020 = 1.03) instead of adding 1.
  25. armogohma

    armogohma

    Joined:
    Sep 9, 2008
    Messages:
    74
    What I meant was, shouldn't we use (2080/2020+2171/2020+2046/2020+2200/2020+...)x4 instead of (1+1+1+1+...)x4? When unadjusted, a typeless (and therefore defensively average) will take 1x damage from all attacks, and thus 68 is an excellent unadjusted base number. But when using adjusted numbers, it's not (4, 2, 1, .5, .25 or 0)x1 plus 16 others, but instead (4, 2, 1, .5, .25, or 0)xOTR plus 16 others.

Users Viewing Thread (Users: 0, Guests: 0)