Stat Flattening and Relative CR Scaling: What the heck are they?

Discussion in 'Oracle’s Database (Guides)' started by Remander, Apr 16, 2015.

  1. Remander Steadfast Player

    Pretty sure Tunso said that blocking does triple defense. Easy way to test defense would be to use a tank powerset in DPS stance. Get full DPS and tank sets and titrate the tank gear in to up your defense against an NPC, like a CC bounty. CR would remain static. Compare average hits taken. Sort of like what I did with the CR scaling against Captain Cold.
  2. shiny mackerel Committed Player

    The defense conversions definitely changed. It seems to depend on your max CR now, which means it doesn't stop at level 30 or the CR differential is affecting it some way. I have no idea. I tested at 116/124 CR on different mobs and got the same conversions for the same max CR + differing equipped CR + differing CR mobs, but it changes when your max CR changes. So far they've been in the range of 1% mitigation = 450 defense to 480 defense. The conversion on live hasn't changed and should still be around 70/1%

    Also you can check that blocking really triples your defense by quickly opening your inventory/stats as soon as you block.
    • Like x 2
  3. Rokyn Dedicated Player

    How can we can calculate fair base numbers for might and precision powers now? Now that your opponent and you are tied to CR? I tried running base numbers on different sparring targets (level 1 and level 126) (With no defense).....and I'm not getting the normal numbers I usually get. For example an easy test is a burst might power:

    With PI: 30-42 damage and without PI: 20-27 damage

    But I'm getting totally different results. Are the old formulas not viable anymore or is there another way we can test base numbers after update 47?

    Base numbers were useful to determine discrepancies and similarities between powers and it served as a fair medium to use to compare powers. I'm currently trying to clear the smoke between Earth and Sorcery and calculate pet base numbers.
  4. Remander Steadfast Player

    That's actually around what I would have predicted. If you consider that on live, with 86% armor piercing at T6, every 1000 defense will mitigate approximately 2% of incoming damage. Take away the armor piercing, and to get the same feel, you'd be looking at about 1% mitigation per 500 defense. As for the relative CR scaling, I suspect that is a direct scaling of damage, rather than of defense. Could be wrong, though. Not sure how to best test that. Was hoping Tunso would enlighten us. Thanks for the info!
  5. L T Devoted Player

    I haven't worked that out yet. I believe the scaling is supposed to be based on your equipped CR, rather than MAX CR. Based on tests in the Patrol Catastrophe Solo the mobs seem to be built around your CR+5 giving the same base damage as on live.

    Earth and Sorcery Pets always seem to spawn at your MAX CR. Test them against the high level targets.

    For other powers, you could take off all your gear so that you are CR1 and then gear up so that you are 121 (against the 126 level target) and check the numbers. If you switch to Fire, you can compare your numbers to the base damage numbers I have here:

    https://forums.station.sony.com/dcuo/index.php?threads/trailblazing-guide-to-fire.632/

    The Dark also has reliable numbers in the gadgets guide.

    I think (I hope!) things are still in flux on the test server in terms of CR scaling and enemy toughness. I was running Patrol Catastrophe at CR105 with no mods and only 10 skill points and even with those limitations it really wasn't difficult.

    As I type and think, if you just want to compare the two powers, it probably doesn't matter which method you use to test as all powers should scale the exact same way with CR. The numbers won't be useful for comparing to live, but should be accurate enough for comparing one against the other.
    • Like x 1
  6. L T Devoted Player

    I think, based on my numbers here:
    https://forums.station.sony.com/dcu...-the-heck-are-they.249944/page-2#post-3223557

    That I was getting 550 Defense = 1% mitigation. That was with CR105 gear against (IIRC) CR112 mobs.
    • Like x 1
  7. shiny mackerel Committed Player

    The CR differential scaling is just an independent multiplier so you just need to find out how it's affecting your damage and normalize by that. I use weapon attacks as my reference since they're quick to get samples of and I already know the base values. Find out however much your reference move is being reduced/boosted by and adjust your results based on that. Until someone figures out exactly how the CR differential affects damage or a dev tells us, it's gonna be a pain in the *** testing with sparring targets.
    Or probably an easier way - use another character as the sparring target and just duel them. Make them put on full pve gear at 0 toughness and you won't have to worry about defense. You'll have to wait for them to heal for 20 sec when they get low but at least the cr scaling doesn't matter there.
    also btw, the burst might power should be 20-28 damage. there was that might damage roll update a while ago that bumped up the max damages of a few powers.

    hey... yeah that makes sense. and I do agree that it's most likely only damage is being scaled by the CR differential. that's what the patch notes seem to say anyway.

    I can confirm this

    What was your max CR for these tests? Also how are you getting 550 defense?
    For your first test (11627 defense), I'm getting 1% mitigation = 455 defense, which is exactly what I got for my tests at 116 CR. I'm assuming you weren't at the cap when blocking here, which was my experience at around this defense. Results would be strange otherwise.
    For your second test (13441 defense), I'm getting 1% mitigation = 411 defense. Assuming you're at the cap when blocking here, which was what I found at around this defense at 116/124 CR.
    For your third test (18534 defense), I'm getting 1% = 470 defense. You're definitely at the cap when blocking here.
    • Like x 3
  8. L T Devoted Player

    I *think* it was 116 or 118, but I was testing with all CR92 gear (equipped CR105 in all cases) and I *think* that's all that matters.
    Here is my thinking:
    I looked at blocking to find the mitigation cap. At 11627 defense (assuming x3 = 34881 while blocking) I'm definitely not at the mitigation cap when blocking, because damage I take while blocking goes way down when I swap in some tank gear. At 13441 (x3=400323) I think I'm close but not quite at the cap because damage still goes down when I went to all tank gear. At 18354 defense I'm definitely hitting the cap. You can see when I block average damage still goes down from the 13441 samples, though not by much.

    So the second test is close to but not quite at the cap while blocking, and I believe that blocking triples my defense. So 40323 Defense is almost 75% mitigation. If it's 550 defense per 1% mitigation, the cap would be 41250, and 40323 Defense would be about 73% mitigation.

    This could be all wrong. It's not the most scientific reasoning, but it seems to fit.
  9. shiny mackerel Committed Player

    I'm thinking that at 13441 defense, you were most likely already at the cap.
    Blocking at 18534 defense definitely gets you at the cap, so we can say that you would've taken 538 * 4 = 2152 damage as if you had 0 defense. Then without blocking, your armor mitigated (1 - 1304/2152) = 39.4% damage, so the resulting defense conversion would be 18534/39.4% = 470 defense/1% mitigation.

    It's a bit trickier when you're not at the cap when blocking, but this is how I did it. Using your first test as an example (11627 def):
    Assume the mob deals D damage unmitigated while your armor mitigates A% damage. Then blocking should mitigate 3 * A% damage.
    So not blocking, you'd be taking D * (100% - A%) = 1939 damage.
    Blocking, you'd be taking D * (100% - 3 * A%) = 605 damage.
    Solving this gets you A = 25.6% damage mitigated by your armor. Then the conversion would be 11627/25.6 = 454 defense/1% mitigation. This is consistent with what I was getting at 116 max CR.

    If you assumed you weren't at the cap blocking with 13441 defense, then doing this would get 562 defense = 1% mitigation. But if you assumed you were already at the cap, you'd get 411 defense = 1% mitigation. And closer if you used the blocking damage for your third test. I just think the conversion makes more sense if you were already at the cap. Maybe the difference in blocking damage could be due to sample size? 554 and 538 are quite close. If you kept the same max CR, you should see the same defense conversion in all cases anyway.
  10. L T Devoted Player

    Sample sizes were not very large. Crits were removed from the run.

    3 not blocking / 10 blocking with low defense (then I died)
    18 non-blocking/22 blocking with partial tank gear
    20 non-blocking/27 blocking with full defense

    So there's room for error, but I would guess that the last two sets are robust enough to at least spot a trend.
  11. Remander Steadfast Player

    With the variance in hit magnitude, 20 may actually be a small sample. For most statistics, 20 samples is decent. I've just found that I pick up additional outlying values when I push to higher sample sizes.
  12. Remander Steadfast Player

    Also keep in mind that Fortified Blocking now increases the mitigation cap to 80%. L T, did you have that mod on during your tests?
  13. L T Devoted Player

    That's a fair question. Answer: Absolutely NOT. I always try to test things with as few variables as possible.
    • Like x 1
  14. Remander Steadfast Player

    Good deal. Didn't think you would, but thought I should ask.
  15. Remander Steadfast Player

    Reading through the patch notes again got me thinking. I suspect the defense formula is actually dynamic with CR tier. It should pretty much mirror the effects of armor piercing, as described above, but it allows it to be carried out infinitely. Each new tier sort of resets it lower, and gearing up through the tier brings tanks at least nearer to the cap. Starts over again next tier. The numbers we're looking at are T6 and T7. I suspect for T7 a higher defense is needed for that 1% mitigation. Seems obvious to me now. Maybe you guys already had that figured, lol! I tend to have epiphanies in the morning. What can I say? ;)
  16. Crimson Mayhem Loyal Player

    I didn't know that was even a question, I assumed it to be this way from the start :D Higher CR = more Defense needed for the same mitigation. Effective mitigation will stay roughly the same for each role though I think it is only bound to CR, not tier. From what I understand there won't be hard tiers anymore with the new system.
  17. Remander Steadfast Player

    Lol, yeah. Late to the party.

    I think there will be a soft tier structure, though. That would allow you to progress toward the mitigation cap within a CR band. Otherwise, your mitigation potential would basically remain static throughout the game. I guess they could do that, though. Would require a lot of testing to figure out.
  18. Crimson Mayhem Loyal Player

    Both ways sound plausible to me really. Just for the purposes of introducing new R&D levels I guess there will be some kind of soft tier structure (and also because the On Duty UI probably can't handle anything else :D) so might be that the mitigation scale actually is tiered too.
  19. L T Devoted Player

    I was thinking the same thing. In fact, I was thinking it might even be sensitive one a piece-by-piece basis of your gear based on earlier tests.

    This poses a question for both damage and defense comparisons: since it's dynamic, where should the point of comparison be? Should you test against an enemy with the same CR as you (I would guess that's end of tier) or is it more important to test performance against something CR+9 or CR+10 (early tier)?
    • Like x 1
  20. Remander Steadfast Player

    That depends. Tunso has already said that we're getting spotted some CR levels against certain NPCs, in order to make it work out. That tells me that we can't be sure when we're on equal footing with a given NPC. Will make testing very difficult, indeed.

    As far as the dynamic nature, if it's really adjusted by each CR step, rather than by bands, that will give us a pretty static mitigation feel. Tanks will be able to cap with block at any CR. Would be interesting, but not sure I like it. I'd prefer some room to grow.