|
|
Post by Sin-N-Terrors on Feb 27, 2014 22:40:33 GMT -5
Contributed by SinA Discussion amongst different managers regarding diminishing returns of speed.---Posted by: Advancement of the Axis Date: Saturday 3:20 PM I read somewhere abt after spd reaches somewhere in the 20's that the returns diminish. 1) is this true?; 2) Are there diminishing returns for str and agil as well? PLEASE EXPLAIN HOW THIS WORKS., THANKS --Posted by: Morton`s Moderator Winner of 19 Regional Titles Winner of 3 World Titles Date: Sunday 10:49 AM This has been discussed many times. So, let me explain it as tersly as possible. Then, if you need those explanations expanded, we can do that.  STR and AGL do not have diminishing returns. In the calculations for damage and defense, they linearly affect the results of the calculations. SPD on the other hand is subject to calculations that use the square root of its value. SQRT(AGG * SPD), so the larger SPD becomes, the less net gain it has post formula. Similarly, in the punches landed formula: PERCENT = SPD*SPD / (SPD*SPD + k * sqrt(k*opp(AGL))) The greater SPD is, the less affect if has on increasing the percentage of punches landed. going from 20 to 25 SPD, you would see a larger percentage increase in punches landed than you would see if going from 25 to 30. That is why they say there is a diminishing returns, at some point, the gains from higher speed do not outweigh the lost damage or defensive potential from APs spent in SPD over AGL or STR
|
|
|
|
Post by Sin-N-Terrors on Feb 27, 2014 22:44:36 GMT -5
written by: cozar
We can illustrate this with some hypothetical numbers. Assuming an efficient 4/8/8, so AGG = 4 we have the following.
SPD = 4; SQRT(AGG*SPD) = SQRT(4*4) = 4;
SPD = 6.25; SQRT(AGG*SPD) = SQRT(4*6.25) = SQRT(25) = 5; (SPD increased 2.25 for a 1 point gain in the SPD damage component)
SPD = 9; SQRT(AGG*SPD) = SQRT(4*9) = SQRT (36) = 6; (SPD increased 2.75 for the next 1 point gain in the SPD damage component)
SPD = 12.25; SQRT(AGG*SPD) = SQRT(4*12.25) = SQRT (49) = 7; (SPD increased 3.25 for the next 1 point gain in the SPD damage component)
SPD = 16; SQRT(AGG*SPD) = SQRT(4*16) = SQRT(64) = 8; (SPD increased 3.75 for the next 1 point gain in the SPD damage component)
SPD = 20.25; SQRT(AGG*SPD) = SQRT(4*20.25) = SQRT(81) = 9; (SPD increased 4.25 for the next 1 point gain in the SPD damage component)
SPD = 25; SQRT(AGG*SPD) = SQRT(4*25) = SQRT(100) = 10; (SPD increased 4.75 for the next 1 point gain in the SPD damage component)
SPD = 30.25; SQRT(AGG*SPD) = SQRT(4*30.25) = SQRT(121) = 11; (SPD increased 5.25 for the next 1 point gain in the SPD damage component)
As you can see, SPD has diminishing returns in damage. However, SPD can still provide higher damage increases when you have low SPD or high STR. Every 1 point gain in STR increases your unmitigated (numerator) damage by POW*SQRT(AGG*SPD), while every 1 point increase in the speed component of damge, SQRT(AGG*SPD), increases your unmitigated damage by POW*SPD. Keeping our 4/8/8 fight plan, and assuming 15 STR and 4 SPD, we can look to see whether a STR or SPD increase is more beneficial.
Increasing SPD by 2.25 = unmitigated damage increase of 8*15 = 120, while increasing STR by 2.25 = unmitigated damage increase of 8*4*2.25 = 72. Thus, we choose to upgrade SPD to 6.25. Next, increasing SPD by 2.75 (to 9) = increase of 8*15 = 120, while increasing STR by 2.75 = increase of 8*5*2.75 = 110, so we still do better off with SPD. Finally, the next SPD increase (by 3.25 to 12.25) will be 120 v. 8*6*3.25 = 156. Thus, this time STR increases are better. Ultimately, we end up with 18 STR (technically we want 18.25) and 9 SPD, which helps explain why people recommend keeping STR = 2*SPD for clinchers.
He, he, let's see if I can turn that math into simpler conclusions:
1) There is no "magic number" where SPD starts having diminishing returns for damage, it ALWAYS has diminishing returns.
2) Keeping STR at twice the value of SPD maximizes raw, unmitigated damage (damage dealt if your opponent had 1 DEF and 1 AGL) when using a 4/8/8 or 4/8/8 (clinch) fight plan. This is merely a mathematical observation that helps explain why clinchers tend to keep STR around 2x SPD. It does not consider the effects of style, the need to win rounds by score, or variations in AGG/POW/DEF other than 4/8/8, so it is not a general rule for all fighters.
|
|
|
|
Post by Sin-N-Terrors on Feb 27, 2014 22:46:27 GMT -5
written by: Sanford
Here's where it get tricky:
"1) There is no "magic number" where SPD starts having diminishing returns for damage, it ALWAYS has diminishing returns."
Yes, but the higher the speed, the great the diminishing return is. I'm no math geek, but fighters with 30 speed have won WT's, so having a lot of speed should be no problem at the beginner level.
"2) To maximize damage, keeping STR at twice the value of SPD is probably a good general guideline."
No. You need to read about fighter types and fighting styles. Counterpunchers can win over most fighter types and they have speed as their greatest attribute. Clinchers and sluggers, on the other hand, do keep str near twice speed.
|
|
|
|
Post by Sin-N-Terrors on Feb 27, 2014 22:47:16 GMT -5
written by: cozar
You are absolutely right, I way overstated my conclusion in #2. If you look at the last line of my earlier post, you'll see my actual conclusion was "Oh, now I see why clinchers tend to keep STR twice as high as SPD." From my calculations, I'm willing to hypothesize that all fighters would benefit from having SPD at least as high as half strength (SPD >= STR/2), but that is an untested hypothesis and it is a lot different from what I actually wrote.
As to 1) I wasn't suggesting that 30 SPD fighters are bad (or that they are good). The point is that all SPD has diminishing returns, so you can't say "no fighter should ever have more than x SPD because after "x" there are diminishing returns." If that statement were true, then x would equal 1. While diminishing returns is a consideration, the maximum SPD that is beneficial for a fighter will depend on style, strategy and your other stats.
|
|
|
|
Post by Sin-N-Terrors on Feb 27, 2014 22:48:43 GMT -5
written by: Sardonic's
My personal belief -- keep speed where it needs to be to stay relevant score-wise, but if you aren't running a CPer or slapper, then those points are better spent in STR or AGL (which aren't rooted).
|
|
|
|
Post by Sin-N-Terrors on Feb 27, 2014 22:49:29 GMT -5
written by: Ghosts Who Walk
There is nothing at all wrong with having speed above 25 on KP Dancers at WT level.. In lower levels (Webl-revival region) and regionals speed rules the roost for these style of fighters so having speed +2,3,4 is fine. As they start to get higher in contenders then you should be bringing the agility closer and probably drawing level or speed -1 or 2 on agility.. In my opinion 15/5 is an absolute must from at least Welterweight and down, 12 chin (This is at WT level obviously at regional level then 12/4 is more than enough).
|
|
