View Full Version : What is your mit and avoidance?
Thanous
09-29-2005, 08:45 PM
What is your mit/avoidance in percentage (as well as level)? I'm going to run some stocastic tests and see how things work for different fighter classes. Thanks, <div></div>
Thanous
09-29-2005, 09:13 PM
<div></div>Just as a background of what I'm doing... I'm using a Latin Hypercube Montie Carlo simulation package to run some numbers for how avoidance and mit really do relate to each other. Using this, I can simulate thousands of attacks against your percentages in a few seconds. With this, I hope to be able to come up with a relation between the importance of avoidance and mit that we can all relate to and live with. Granted, this will only be valid for fights that are your level, no arrow, non-heroic. Here's a link to see what I've done so far in case you think this might be a witch hunt. I am just trying to understand the numbers and what they mean for each of us. <a href="http://eqiiforums.station.sony.com/eq2/board/message?board.id=3&message.id=20073" target=_blank> Link B Here</a> <div></div><p>Message Edited by Thanous on <span class=date_text>09-29-2005</span> <span class=time_text>12:01 PM</span>
<P>Any test you do will be fundamentally flawed. Why? Quite simply because the combat result is dependant on many more factors that just the Avoidance % and Mitigation % score on your persona.</P> <P>For instance, you have absolutely no idea how con colour affects the %'s. </P> <P>Nor do you know what combat arts sub-classes use during combat, be it stuns, stifles etc. </P> <P>Short use, long reuse buffs also affect combat since it generally only lasts 1min but would skew the infinate time test.</P> <P>DPS output also affects combat, as well as chance to hit, mob being fought.</P> <P>Quality of combat art effects things.</P> <P>So, comparing 2 numbers whilst informative does not allow you to accurately model any relationship between the subclasses.</P>
Thanous
09-29-2005, 11:10 PM
<div></div>Nemi, All of those are very valid, and I accept those arguements. I state that this is only for melee defense, auto attack NPC damage. No casting, no resists other then defense stats. Now, is that practical? No, it is not. If you choose not to provide information based upon that, it is cool with me. However, the most noise lately has been coming off those two numbers. I'm not trying to make arguements on who is better then who, as we all know it becomes a complicated arguement very quickly. I just want to understand how those two simple numbers do affect each class. Because of that, my test is not flawed. It is flawed if you try to apply it to a scope greater then which it is applicable though. FYI, Pasted your reply into the top of the thread I linked to as a warning/disclaimer, as well as my reply. It is a very important point to make. <div></div><p>Message Edited by Thanous on <span class=date_text>09-29-2005</span> <span class=time_text>12:15 PM</span>
bonesbro
09-30-2005, 03:23 AM
Going from memory, I think I have about 28% mitigation and 60% avoidance. I can temporarily buff myself up a bit higher, but not much.
Tilane
09-30-2005, 12:37 PM
<DIV>37,1% mitigation , 63.3 avoidance at lvl 52 in defensive stance winding dragon </DIV> <DIV>30.3% mitigation , 50.1 avoidance at lvl 52 in offensive stance Halcyon Blessing</DIV>
Thanous
09-30-2005, 08:13 PM
Ok, while in <b><u>Defensive Stance</u></b> Mean Taken 23.12888505 Percentile Taken 0.05 0 5 0 10 0 15 0 20 0 25 0 30 0 35 0 40 0 45 0 50 0 55 0 60 0 65 46.63220522 70 54.25937634 75 58.38868408 80 62.17776565 85 65.316515 90 69.23257212 95 73.63554119 99.5 83.18967335 Damage that hits... Mean Taken 62.89990818 Percentile Taken 0.05 29.32171215 5 47.37274058 10 50.80441736 15 53.12081843 20 54.95127548 25 56.53286714 30 57.94324425 35 59.26205833 40 60.50571827 45 61.70885624 50 62.89583527 55 64.0816962 60 65.29013687 65 66.52995627 70 67.84623961 75 69.26298641 80 70.83448811 85 72.66334822 90 74.9895428 95 78.41513086 99.5 87.18468944 Little bit of commentary...looks a lot like a standardly buffed Guardian on the mean value. When you get hit, it hurts a lot worse but don't get hit nearly as often (as expected). In <b><u>Offensive Stance</u></b> Mean Taken 34.85223538 Percentile Taken 0.05 0 5 0 10 0 15 0 20 0 25 0 30 0 35 0 40 0 45 0 50 0 55 56.280638 60 60.78668982 65 64.24228737 70 67.05916916 75 69.81637359 80 72.53623009 85 75.20418157 90 78.65435184 95 83.22503083 99.5 94.74686901 Damage actually taken... Mean Taken 69.70445808 Percentile Taken 0.05 33.05405545 5 52.5005645 10 56.30072068 15 58.86169178 20 60.89553659 25 62.64691209 30 64.20823497 35 65.65882442 40 67.04931432 45 68.38537603 50 69.69464853 55 71.01148387 60 72.34412292 65 73.72593974 70 75.17177968 75 76.75018973 80 78.49495801 85 80.5269632 90 83.07673296 95 86.88722891 99.5 96.29501088 Again, I've seen Guardians with simular means. The median behavior however is different for both classes though. It would scare me to try and tank a long fight with your offensive stance, I'll say that much! <div></div>
x0rtrun
09-30-2005, 08:43 PM
Trust me, fights don't last long while in offensive stance. Someone hits the floor pretty quickly. <div></div>
Thanous
09-30-2005, 08:58 PM
See, this is why I'm trying to do this. Honestly, I do play a guardian, but I'm not going to complain about other classes because I don't play those classes. I'm trying to use these numbers to educate myself and others after the revamp. Any more data from you all would be very much appriciated. <div></div>
<DIV>67.1% avoidance, 34.2% mitigation</DIV> <DIV>42 monk</DIV> <DIV> </DIV> <DIV>btw you CAN tell how much con color affects avoidance.</DIV> <DIV>heres how you test it:</DIV> <DIV> </DIV> <DIV>when your about 90%-99% ding'd, write down your avoidance #s (make sure all your skills are maxed)</DIV> <DIV>right after you ding, your avoidance skills will still be that of 1 level behind. so when you check your stats, your avoidance will be lower because it is against a higher level mob. usually its like 2% lower. thats the difference between LVLx.9% and LVLy.0 though. so im pretty sure each mob level difference is about 3-4% avoidance difference.</DIV>
<DIV>double post</DIV><p>Message Edited by woode on <span class=date_text>09-30-2005</span> <span class=time_text>10:06 AM</span>
Thanous
09-30-2005, 09:23 PM
Thanks for the tip. As for your results... I have no idea your level or buffs, but here ya go... Mean Taken 32.91209796 Percentile Taken 0.05 0 5 0 10 0 15 0 20 0 25 0 30 0 35 0 40 0 45 0 50 0 55 53.26068562 60 57.80198458 65 60.84377266 70 63.59370198 75 65.97880268 80 68.45034917 85 71.05904508 90 74.08740942 95 78.27716315 99.5 88.67061701 Damage taken... Mean Taken 65.80274707 Percentile Taken 0.05 33.12390464 5 49.5510664 10 53.14715832 15 55.56991873 20 57.48316611 25 59.13998454 30 60.61857147 35 61.99263905 40 63.29702833 45 64.55487133 50 65.78996899 55 67.03945619 60 68.29809626 65 69.59543148 70 70.97565875 75 72.45286124 80 74.10031684 85 76.02572076 90 78.43947172 95 82.01472777 99.5 91.22370545 <div></div>
bonesbro
10-01-2005, 01:15 AM
<BR> <BLOCKQUOTE> <HR> Thanous wrote:<BR>See, this is why I'm trying to do this. Honestly, I do play a guardian, but I'm not going to complain about other classes because I don't play those classes. I'm trying to use these numbers to educate myself and others after the revamp. Any more data from you all would be very much appriciated.<BR> <BR> <HR> </BLOCKQUOTE><BR>A good goal. I would also consider including the standard deviation - avoidance tanking is a bit more risky than mitigation tanking because of damage spikes. Miss a couple of blocks and you took four hits in a row with 30% mitigation instead of 50%, which gives healers a frowny face and your xp bar a reddish tinge.
Thanous
10-01-2005, 01:17 AM
Thanks for the feedback, I can do that <img src="/smilies/3b63d1616c5dfcf29f8a7a031aaa7cad.gif" border="0" alt="SMILEY" /> <div></div>
zabor
10-01-2005, 01:58 AM
i have no clue what all those numbers mean <img src="/smilies/3b63d1616c5dfcf29f8a7a031aaa7cad.gif" border="0" alt="SMILEY" />
Thanous
10-01-2005, 02:09 AM
This is done using a random number generator. I have it set so that a normal (bell) distribution with a average of 100, a minimum of about 50, and a max of about 150 will give me inputs as damage. I then take that damage, and again using the random number generator, throw out or keep that trial's damage depending on if it is avoided or not (avoided at the %'age rate provided by players). After that, if the damage makes it thru being avoided, I mitigate it by the percentage provided by players (if raw mitigation and level are provided, the percentage= mitigation/(level*100)). Now, I do this at least 5000 times. Basicly, I simulate the same creature, a creature your level, sucessfuly swinging at you 100 times for random amounts of damage, then process how much damage you actually take. The outcome is a curve of how likely a value is, as well as means, standard deviations, and such. What I've been pasting in is the percentiles and the damage at that percentile. What you'll notice is on the first set of percentiles, you see that no damage is done at any percentile that is below the avoidance percentage. After that, it is a matter of mitigation against the damage that was done on any swing. The second chart is just simply if you avoided no swings. If everyone hit, what would the curve look like. It is a raw measure of mitigation. Where the second chart can pertain to avoidance is looking at the differenece between the first mean value and the second mean value. That difference is all provided by avoidance. Does that help at all? <div></div>
Zarvax
10-02-2005, 10:11 PM
<P>63.4% avoidance, 37% mitigation in tank stance at level 54</P> <P>52.2% avoidance, 32.2% mitigation in offensive stance at level 54</P> <p>Message Edited by Zarvax on <span class=date_text>10-02-2005</span> <span class=time_text>11:14 AM</span>
Thanous
10-03-2005, 09:26 PM
<span><blockquote><hr>Zarvax wrote: <p>63.4% avoidance, 37% mitigation in tank stance at level 54</p> <p>52.2% avoidance, 32.2% mitigation in offensive stance at level 54</p> <p>Message Edited by Zarvax on <span class="date_text">10-02-2005</span> <span class="time_text">11:14 AM</span></p><hr></blockquote></span><u><b><span>63.4% avoidance, 37% mitigation in tank stance at level 54 </span></b></u><span>Mean Taken 23.08524424 Percentile Taken 0.05 0 5 0 10 0 15 0 20 0 25 0 30 0 35 0 40 0 45 0 50 0 55 0 60 0 65 47.14497775 70 54.58478029 75 58.6055409 80 61.93948323 85 65.09301278 90 68.39755458 95 73.63936774 99.5 83.90258991</span><u><b><span> </span></b></u><span>Damage Taken Mean Taken 63.00018491 Percentile Taken 0.05 32.86104371 5 47.45295853 10 50.87908991 15 53.20262507 20 55.03103997 25 56.62478243 30 58.04286899 35 59.33918939 40 60.60511662 45 61.8075844 50 62.9981196 55 64.18107959 60 65.38833404 65 66.64103014 70 67.94792448 75 69.37033594 80 70.95116817 85 72.77965909 90 75.0967428 95 78.4757952 99.5 87.19898695 </span><b><u><span>52.2% avoidance, 32.2% mitigation in offensive stance at level 54</span></u></b><span> Mean Taken 32.43794716 Percentile Taken 0.05 0 5 0 10 0 15 0 20 0 25 0 30 0 35 0 40 0 45 0 50 0 55 51.64810612 60 57.52810665 65 61.27569272 70 64.55560054 75 67.56829509 80 70.10370873 85 72.81075436 90 76.11082281 95 80.90550205 99.5 91.63839871 Damage Taken Mean Taken 67.80419345 Percentile Taken 0.05 36.27860788 5 51.04479942 10 54.75394121 15 57.25736157 20 59.23279466 25 60.92695826 30 62.46507521 35 63.87601109 40 65.2137904 45 66.52098695 50 67.78643767 55 69.07669459 60 70.37205991 65 71.71048579 70 73.12730206 75 74.64610906 80 76.35741881 85 78.33673309 90 80.83102842 95 84.51222584 99.5 94.03274965 </span><u><b><span></span></b></u><div></div>
x0rtrun
10-04-2005, 03:44 AM
Lvl 46 offensive: 28% mit and 53.5% avoid balanced: 30.6% mit and 58.4% avoid defensive: 33.8% mit and 64.6% avoid <div></div>
vBulletin® v3.7.5, Copyright ©2000-2025, Jelsoft Enterprises Ltd.