|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 17:18:59 GMT -5
Another thread recently started with the idea that some bots should have their cost fixed to respect their actual combat value. The thread quickly devolved into a "buff this, nerf that, don't you dare to nerf that!" discussion, but I think that it would be interesting to find a way to actual evaluate bots in a semi-objective way.
i proposed, as a starting position:
Let's say a point system. I propose a simple speed*maximum health*cycle damage (as per wiki) of standardized weapons . (Happy to hear other suggestions). For maximum health I'll go for maximum frontal health, so health+shields. Counting energy shield at 50% for simplicity and adding a 10% to dashing, jumping and cloacking bots.
I'll take Destrier, Gepard, Stalker, Gareth, Boa, Golem, Carnage, Galahad, Griffin Fury Lancelot. As weapons. level 12 pins for light, orkan for medium and Thunder for heavy (to keep everything in short range). It's crude, but just to give an idea. Again, happy to hear suggestions for other metrics.
Results are
Destrier: 87(k)*55*(2*22832)
Gepard: 87*58*(3*22832)
Stalker: 90*66*(2*22832)*1,1
Gareth: (80+160)*64*(22832+45664)
Boa: 185*42*(45664+81520)
Golem: 132*44*(22832+45664+81520)
Carnage: (114+34)*46(avg)*(2*81520)
Galahad: (120+199)*50*((2*22832)+45664)
Griffin: 158*35*((2*22832)+(2*45664))*1.1
Fury: 158*30*(3*81520)
Lancelot: (170+360)*36(avg)*((2*45664)+81520)
Ancilot: (170+360+47,5)*36(avg)*(2*45664)
Which gives (/1000.000)
Destrier 218,5
Gepard 345,6
Stalker 298,3
Gareth 1.052
Boa 988,2
Golem 871,2
Carnage 1.109,9
Galahad 1.456
Griffin 833,3
Fury 1.159
Lancelot 2.489
Ancilot 1.471,2
It was pointed out that shields and energy shields should be graded differently, that rather than a standard loadout, a specific/most common loadout should be considered (but then, which would that be), that beacon capping capability should be cosnidered (I don't think so, as it's essentially un-rateable and speed is a decent proxy for it, but that's me) but nothing definitive as the thread devolved again in a "this is OP and should be nerfed" "No way".
So, I'm proposing this again.
PLEASE, without calling for OP, nerfs and buffs, but just looking at the bot as they are now, what would you think would be an objective, or as objective as possible, way to numerically grade them?
|
|
|
|
Post by KaneoheGrown on Mar 15, 2017 17:43:04 GMT -5
Simple, use basic supply and demand methodology.
Everyone wants a Dash? Start it at 5k gold and allow it to increase till only 10% of the player population is buying it (10k+ Au).
No one wants a Vitz? Make it 1 Au till the demand rises.
Realistically, bot prices should reflect the utility within the meta. IMO, Galahads and Griffins are way too cheap when compared to how important they are in a players hangar. Supply and demand pricing would allow them to correctly reflect how much players are able and willing to pay for them.
|
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 17:45:56 GMT -5
That's not an objective way to rate robots on their capabilities, it's based entirely on the subjective tastes and evaluations of the players. Also, it would be based on data we don't have (numbers of bots in play and stuff) so, we couldn't make an actual ranking.
Please stay on topic.
How to rate bots, among us with data that we have, in a seemingly objective, formula based, numerical way.
|
|
|
|
Post by hurtbot on Mar 15, 2017 17:56:32 GMT -5
Well, since a robot's combat value changes depending on map, enemy bot types deployed, the pilot, and the strategy they're employing, I don't think there really is a way to do this completely objectively. IE: The guy who can pilot the bejeezus out of a stalker but can't pilot a Lancelot to save his life is obviously going to find far more value in the Stalker. So to do this in a formulaic fashion you'd have to include variables to represent those factors... which would be slap-dash at best.
|
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 18:04:15 GMT -5
Pilot is irrelevant. When you evaluate cars one assume the same pilot piloting them all and being able to pilot them all. It0s an "on paper" discussion.
On maps, I would also disagree. Considering that all bots have some maps in which they are better and others they are worse, that'd would even out. But for mental exercise sake, let's assume a totally balanced ma with some terrain fit for all of them.
|
|
|
|
Post by mijapi300 on Mar 15, 2017 18:07:09 GMT -5
Well, since a robot's combat value changes depending on map, enemy bot types deployed, the pilot, and the strategy they're employing, I don't think there really is a way to do this completely objectively. IE: The guy who can pilot the bejeezus out of a stalker but can't pilot a Lancelot to save his life is obviously going to find far more value in the Stalker. So to do this in a formulaic fashion you'd have to include variables to represent those factors... which would be slap-dash at best. Thats kind of what objective means. The intent of the op is to determine objective values of each bot, excluding outside factors such as maps, pilot skills, and the such. Just the intrinsic value of the bot itself with the weapon setups available to that bot. |
|
|
|
Post by hurtbot on Mar 15, 2017 18:17:28 GMT -5
Well, since a robot's combat value changes depending on map, enemy bot types deployed, the pilot, and the strategy they're employing, I don't think there really is a way to do this completely objectively. IE: The guy who can pilot the bejeezus out of a stalker but can't pilot a Lancelot to save his life is obviously going to find far more value in the Stalker. So to do this in a formulaic fashion you'd have to include variables to represent those factors... which would be slap-dash at best. Thats kind of what objective means. The intent of the op is to determine objective values of each bot, excluding outside factors such as maps, pilot skills, and the such. Just the intrinsic value of the bot itself with the weapon setups available to that bot. Ok... I'll re-phrase. There are too many subjective factors involved to quantify this. This is akin to the "Best Robot" discussion. The only way you'll wind up with something resembling an accurate quantification is by doing some translation of subjective values into concrete ones. Incidentally, he's asking for "as objectively as possible," not "strictly objectively" |
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 18:30:21 GMT -5
Yes, but I'm also asking, or rather inviting others to come up with me, for something reasonably quick and straightforward, not a multivariate analysis with focus-group defined subjective variables.
mijapi300 actually got the spirit of it.
For instance, beacon-capping capability is subjective (stalker is better than a gareth?), but speed is, I think, a pretty goof proxy variable: faster usually means getting there first and being able to retreat and come back.
Jumping is harder to find a proxy for, as it generally means you can avoid one attack every 23 seconds and also works as extra speed and, yes, it has tactical use, here really subjective.
What about cloacking?
Also, phisical shields: I've put full value, but someone said, and it's correct, that they have splash damage pass thro. Also, they don't cover all sides. What would be a reasonable multiplier for them? 0.6? 0.5 for Lancelots and 0.3 for britbots, that can only cover 1/4 at any given time?
What about Energy shields? They cover everything, but let energy pass thro. 0.5?
What I'd like is a reasoned discussion (if that is ever possible on an online forum.. it's rare) on good direct proxies or on how to reasonably translate the non proxy-able stuff stuff in a multiplier. But "There's too many subjective variables so it cannot be done unless you make 100 rankings" kinda defy teh whole idea.
|
|
|
|
Post by gr3ygh05t on Mar 15, 2017 18:46:11 GMT -5
So it's like comparing cars via engine displacement, RPMs, 0-60 time etc. I like it.
How about we standardized slots by assigning an arbitrary weapon on the slot that is comparable across different slots. How about the punisher, punisher Mk2 and the thunder?
|
|
|
|
Post by mijapi300 on Mar 15, 2017 18:51:09 GMT -5
I came up with the following so far:
HP (per thousand): straight number
Physical Shield: 33% decrease because it can't stop splash, 25% decrease because you actually have to aim your shield to block.
Ancile Shield: 33% decrease because of plasma, 10% increase because it recharges
Speed: Ratio of bots speed compared to top speed in game (66)
Damage: 1 point per light, 1.5 points per medium, 2 points per heavy hardpoints
Size: Ranged the bots in target size from 1 (most lights) to 4 (Raijin). Bot rating gets a 10% hit per size rating
Ability: Bot gets a 25% boost in rating for having an ability. Did not enter for Raijin or Fujin. I believe including the shield HP in the base offsets the ability of raising the turret.
Does this sound like a fair rating system to you?
|
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 18:59:36 GMT -5
How about we standardized slots by assigning an arbitrary weapon on the slot that is comparable across different slots. How about the punisher, punisher Mk2 and the thunder? that's what I did, I actually used pinatas, orkans and thunders. I don't think changing weapons, as long as they are all of teh same kind (short, mid, long) changes much the end result. Open to suggestions tho. |
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 19:05:44 GMT -5
I came up with the following so far: HP (per thousand): straight number Physical Shield: 33% decrease because it can't stop splash, 25% decrease because you actually have to aim your shield to block. Ancile Shield: 33% decrease because of plasma, 10% increase because it recharges Speed: Ratio of bots speed compared to top speed in game (66) Damage: 1 point per light, 1.5 points per medium, 2 points per heavy hardpoints Size: Ranged the bots in target size from 1 (most lights) to 4 (Raijin). Bot rating gets a 10% hit per size rating Ability: Bot gets a 25% boost in rating for having an ability. Did not enter for Raijin or Fujin. I believe including the shield HP in the base offsets the ability of raising the turret. Does this sound like a fair rating system to you? Phisical shield: that's the idea, I'd argue about the actual %. Lancelot should be terated differently than other britbots, I think, as its shields cover more than 25%. also, 33% of splash reduction seems steep: it assumed 1 out of every 3 weapons in use to be splash producing. Is that the case? Not sure. Speed: why the ratio rather than it's pure form (and wouldn't end up giving the same result anyway)? I don't get the damage thing. You mean as a multiplier, so a destrier would have his result multiplied by 2, while a Griffin by 6? It might help streamline the "depends on which weapons you mount", in that case, yes size, you mean it gets a 10% penalty because the larger they are, the harder to get cover? reasonable on the ability, I think each ability should be treated differently. An 8 second cloack is a whole different matter than a jump- |
|
|
|
Post by mijapi300 on Mar 15, 2017 19:23:19 GMT -5
I came up with the following so far: HP (per thousand): straight number Physical Shield: 33% decrease because it can't stop splash, 25% decrease because you actually have to aim your shield to block. Ancile Shield: 33% decrease because of plasma, 10% increase because it recharges Speed: Ratio of bots speed compared to top speed in game (66) Damage: 1 point per light, 1.5 points per medium, 2 points per heavy hardpoints Size: Ranged the bots in target size from 1 (most lights) to 4 (Raijin). Bot rating gets a 10% hit per size rating Ability: Bot gets a 25% boost in rating for having an ability. Did not enter for Raijin or Fujin. I believe including the shield HP in the base offsets the ability of raising the turret. Does this sound like a fair rating system to you? Phisical shield: that's the idea, I'd argue about the actual %. Lancelot should be terated differently than other britbots, I think, as its shields cover more than 25%. also, 33% of splash reduction seems steep: it assumed 1 out of every 3 weapons in use to be splash producing. Is that the case? Not sure. Speed: why the ratio rather than it's pure form (and wouldn't end up giving the same result anyway)? I don't get the damage thing. You mean as a multiplier, so a destrier would have his result multiplied by 2, while a Griffin by 6? It might help streamline the "depends on which weapons you mount", in that case, yes size, you mean it gets a 10% penalty because the larger they are, the harder to get cover? reasonable on the ability, I think each ability should be treated differently. An 8 second cloack is a whole different matter than a jump- I agree on the ability thing, but just to streamline, I put either has an ability or doesn't, for now. Yes, the speed basically results to the same thing that way, so it was kind of redundant. The main thing the ratio does, is that if a faster bot is introduced, with a 68 max speed. That would affect the rating of other bots because they would now have to deal with facing that new max speed. A Griffin would be 5, but yes that is the point. And I used 1, 1.5, and 2 because that's what I've seen thrown around frequently. (Medium is 1.5x stronger than light, and heavy is 2x stronger than light). And yes, the size was because it's harder to find cover, and you're also just easier to hit in general. Edit: for the shields, the 33% is because there are three weapon types. Each shield can stop 67% of weapons, so under that assumption, the shield is only 67% useful. For physical shields, the 25% isn't an exact discount, it's just an arbitrary discount on the HP of the shield due to the fact that you have to make sure your shield is facing the damage to block it. This case is true for a Lancelot as well, so I didn't do anything different for him. The fact that he has three shields totaling over 300k base HP account for the fact that they cover more surface. |
|
|
|
Post by hurtbot on Mar 15, 2017 19:29:22 GMT -5
Ok... Let me take a stab a this...
First, in your initial equation you are multiplying speed, health, and cycle damage. This is going to give results that favor the variable that happens to be measured in the greatest quantity. So, a fast bot is automatically way less valuable than a bot with a lot of health. Whether or not this is desirable is... well... subjective. Second, cycle damage is variable according to weapon and doesn't really fit with your objectiveness criteria, so that should be measured in type and quantity of hardpoint instead. To tie all of this together I would assign a value of 1 per x number of units of the variables. So, let's say that speed gets a +1 for every 10kph and health gets a +1 for every 25k. Light hardpoints are +2, Medium +4, and heavy +6. Add +5 for special abilities. The idea here is to try to equate the value of 10kph of speed with 25k of health, etc.. Add all that up and take a look at what you wind up with. By adding instead of multiplying you wind up with smaller, more comparable numbers. If you find that fast bots are overvalued, adjust the modifier... rinse, repeat. No matter what you do, there is some subjectivity involved with this because there is simply no quantifiable way to compare the value of things like speed, health, and damage potential, so you wind up having to assign semi-arbitrary numbers until you find a solution that "looks" right and is consistent. Hope this is helpful.
|
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 15, 2017 19:56:56 GMT -5
Ok... Let me take a stab a this... First, in your initial equation you are multiplying speed, health, and cycle damage. This is going to give results that favor the variable that happens to be measured in the greatest quantity. So, a fast bot is automatically way less valuable than a bot with a lot of health. Whether or not this is desirable is... well... subjective. Second, cycle damage is variable according to weapon and doesn't really fit with your objectiveness criteria, so that should be measured in type and quantity of hardpoint instead. To tie all of this together I would assign a value of 1 per x number of units of the variables. So, let's say that speed gets a +1 for every 10kph and health gets a +1 for every 25k. Light hardpoints are +2, Medium +4, and heavy +6. Add +5 for special abilities. The idea here is to try to equate the value of 10kph of speed with 25k of health, etc.. Add all that up and take a look at what you wind up with. By adding instead of multiplying you wind up with smaller, more comparable numbers. If you find that fast bots are overvalued, adjust the modifier... rinse, repeat. No matter what you do, there is some subjectivity involved with this because there is simply no quantifiable way to compare the value of things like speed, health, and damage potential, so you wind up having to assign semi-arbitrary numbers until you find a solution that "looks" right and is consistent. Hope this is helpful. I agree with some, but all, of what you say. I agree on your observation that the quantity that is larger will have larger influence by the original formula. That was indeed a basic mistake, which I somehow "felt", even if not rationalize, right away and tried to correct by dividing the end result, but that just changes the scale, not the mistake, my bed. On the other hand, to use standard points for everything and thus aiming for equivalency between three different variables doesn't make much sense to me... it's the epitome of subjective to say "5 kph are equivalent to 10.000 damage", even worse assessing teh result and then rebalancing until one gets the result that "feels" fine. So I was thinking about normalizing the three variables somehow. Probably the best way is by using a ratio to max in game as already suggested for speed. Which is relatively easy to do with HP, not so much with weapons. Maybe points might be used for weapon mounts on the assumption that one light weapon does, on average, 1/x damage compared to mediums and 1/2x to heavies, and then normalizing by maximum mount points in game (I suppose, the Fury, so that a Fury would be multiplied by 1, a Destrier by,say, 0.33. Once all variables are normalized, multiplication is better than addition. Again, and here is probably where subjectivity cannot be walked around so much, I don't think one can assign the same multiplier/variable to all the special abilities. |
|
|
|
Post by hurtbot on Mar 15, 2017 20:21:20 GMT -5
Ok... Let me take a stab a this... First, in your initial equation you are multiplying speed, health, and cycle damage. This is going to give results that favor the variable that happens to be measured in the greatest quantity. So, a fast bot is automatically way less valuable than a bot with a lot of health. Whether or not this is desirable is... well... subjective. Second, cycle damage is variable according to weapon and doesn't really fit with your objectiveness criteria, so that should be measured in type and quantity of hardpoint instead. To tie all of this together I would assign a value of 1 per x number of units of the variables. So, let's say that speed gets a +1 for every 10kph and health gets a +1 for every 25k. Light hardpoints are +2, Medium +4, and heavy +6. Add +5 for special abilities. The idea here is to try to equate the value of 10kph of speed with 25k of health, etc.. Add all that up and take a look at what you wind up with. By adding instead of multiplying you wind up with smaller, more comparable numbers. If you find that fast bots are overvalued, adjust the modifier... rinse, repeat. No matter what you do, there is some subjectivity involved with this because there is simply no quantifiable way to compare the value of things like speed, health, and damage potential, so you wind up having to assign semi-arbitrary numbers until you find a solution that "looks" right and is consistent. Hope this is helpful. I agree with some, but all, of what you say. I agree on your observation that the quantity that is larger will have larger influence by the original formula. That was indeed a basic mistake, which I somehow "felt", even if not rationalize, right away and tried to correct by dividing the end result, but that just changes the scale, not the mistake, my bed. On the other hand, to use standard points for everything and thus aiming for equivalency between three different variables doesn't make much sense to me... it's the epitome of subjective to say "5 kph are equivalent to 10.000 damage", even worse assessing teh result and then rebalancing until one gets the result that "feels" fine. So I was thinking about normalizing the three variables somehow. Probably the best way is by using a ratio to max in game as already suggested for speed. Which is relatively easy to do with HP, not so much with weapons. Maybe points might be used for weapon mounts on the assumption that one light weapon does, on average, 1/x damage compared to mediums and 1/2x to heavies, and then normalizing by maximum mount points in game (I suppose, the Fury, so that a Fury would be multiplied by 1, a Destrier by,say, 0.33. Once all variables are normalized, multiplication is better than addition. Again, and here is probably where subjectivity cannot be walked around so much, I don't think one can assign the same multiplier/variable to all the special abilities. You are correct. Like I said this is all subjective to some extent. Value itself in this context is subjective. Using multiplication and ratios will gain you some granularity, but it doesn't minimize subjectivity and adds complexity in addition to magnifying any intrinsic flaws. As far as special abilities go, yeah I think that some are more valuable than others so this is an instance where you sacrifice some objectivity. The main thing is that the result is consistent and provides a meaningful comparison. Let me know what you settle on... it will be interesting to compare our approaches. Edit: Oh... and the reason I "equate" those values is because that's how Pixonic has put things together... there is a general relationship between speed and health in that the faster robots tend to have lower health, as do the bots with special abilities. |
|
|
|
Post by hurtbot on Mar 15, 2017 21:25:13 GMT -5
I agree with some, but all, of what you say. I agree on your observation that the quantity that is larger will have larger influence by the original formula. That was indeed a basic mistake, which I somehow "felt", even if not rationalize, right away and tried to correct by dividing the end result, but that just changes the scale, not the mistake, my bed. On the other hand, to use standard points for everything and thus aiming for equivalency between three different variables doesn't make much sense to me... it's the epitome of subjective to say "5 kph are equivalent to 10.000 damage", even worse assessing teh result and then rebalancing until one gets the result that "feels" fine. So I was thinking about normalizing the three variables somehow. Probably the best way is by using a ratio to max in game as already suggested for speed. Which is relatively easy to do with HP, not so much with weapons. Maybe points might be used for weapon mounts on the assumption that one light weapon does, on average, 1/x damage compared to mediums and 1/2x to heavies, and then normalizing by maximum mount points in game (I suppose, the Fury, so that a Fury would be multiplied by 1, a Destrier by,say, 0.33. Once all variables are normalized, multiplication is better than addition. Again, and here is probably where subjectivity cannot be walked around so much, I don't think one can assign the same multiplier/variable to all the special abilities. You are correct. Like I said this is all subjective to some extent. Value itself in this context is subjective. Using multiplication and ratios will gain you some granularity, but it doesn't minimize subjectivity and adds complexity in addition to magnifying any intrinsic flaws. As far as special abilities go, yeah I think that some are more valuable than others so this is an instance where you sacrifice some objectivity. The main thing is that the result is consistent and provides a meaningful comparison. Let me know what you settle on... it will be interesting to compare our approaches. Edit: Oh... and the reason I "equate" those values is because that's how Pixonic has put things together... there is a general relationship between speed and health in that the faster robots tend to have lower health, as do the bots with special abilities. Ugh... just thought of this. I think we both have a flaw in the way we are approaching this. We are both assuming that the value of a given variable is linear, when it's quite possible it's logarithmic. In other words, 10k of health is more valuable if you already have 100k than it is if you only have 50k. |
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 16, 2017 4:35:14 GMT -5
um. Possibly, but than maybe not. The marginal utility of the last 10k depends if they allow you to survive an "average salvo" more that you would had without, thus being able to return fire. Essentially, if they give you a chance more to return fire, kill the enemy and and survive the exchange. Considering there's no way to know what an average (or median, if deemed more appropriate) salvo is, I think the most reasonable, and feasible, way to use HP is still a normalized linear scale.
If I get half a hour I'll see to rework the data in the OP following what emerged so far.
|
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 16, 2017 5:22:05 GMT -5
So, here it is with normalized HP and Speed, shields and weapon mounts treated as suggested by mijapi300, special abilities still treated as a generic 10% multiplier for teh lack of a better, for now, variable. The results are somewhat in line with my gut feelings about the bots, I must say, except for the Gareth, but that probably because an analysis like this is essentially done on ideally maximized bots. i suspect if I computed in HP and speed for level 3 to 9 bots, results would change somewhat. The Fury placement might have people going mmm, but for the ones, like me, who think the bot is perfectly fine, but it's the loadout (in particular, the trident splash radius) to have something wrong, I guess it fits just fine.  Carnage and Fury don't have special ability as their rush has been computed in speed, which is calculated as an average of continuous run+rush movement. |
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 16, 2017 6:44:44 GMT -5
Re-thinking of it, maybe the physical shield modifier at 55% ("33% decrease because it can't stop splash, 25% decrease because you actually have to aim your shield to block") is too steep, as shield players will manage most of the time to face the enemy and having some form of cover on one side. Gareth and Galahads (not Lancelots) usually die with their shields shattered, meaning they actually used up all the HP.
Maybe 40% for G&Gs, leaving 55% for Lancelots, whose enemies tend to concentrate fire upon from all directions?
|
|
|
|
Post by Conflict's Student on Mar 16, 2017 7:44:57 GMT -5
A lot of effort has gone into underlying game design to prevent the ability to produce exactly this type of ranking.
Rock-Paper-Scissors is interesting because there is no scalar range on which you could rank order the three different options. Rock-Scissors would make for a boring game because no one would play the weakest option. Therefore higher complexity in multiple dimensions becomes necessary to promote hangar diversity (and thereby motivate purchases).
The closest you could come to objective valuation in a single dimension (e.g. price, hangar score, etc) is classic arbitrage: examine valuations taking place in an open marketplace where two-way exchange is possible and there is a large volume of transactions taking place. Pix can't afford to allow such exchanges because they would crush the freemium business model. Thus, we will always be left with arguments of whether Rock is actually better than Paper because, although Paper beats Rock head-to-head, Rock does better against Scissors. Add in the game developer's need to maintain a low level of churn with constant adjustments to a complex balance with over a dozen options (instead of just the 3 in RPS) and you will find the intentional barrier to objective valuation which is a cornerstone of the game's architecture.
|
|
|
|
Post by hurtbot on Mar 16, 2017 8:15:45 GMT -5
um. Possibly, but than maybe not. The marginal utility of the last 10k depends if they allow you to survive an "average salvo" more that you would had without, thus being able to return fire. Essentially, if they give you a chance more to return fire, kill the enemy and and survive the exchange. Considering there's no way to know what an average (or median, if deemed more appropriate) salvo is, I think the most reasonable, and feasible, way to use HP is still a normalized linear scale. If I get half a hour I'll see to rework the data in the OP following what emerged so far. Well, if you look at the upgrade values for a Taran, the first upgrade will cost you 195Ag per point of damage. Looking at a level 12 upgrade, the value is now 96,535Ag per point of damage which is indeed logarithmic. The only question is how relatable this data is to what you're trying to accomplish. |
|
|
|
Post by AηɗυηєɗнєƖ [ǀƬA] on Mar 16, 2017 9:01:34 GMT -5
I think the upgrade cost increase is there more for commercial reasons (forcing you to convert gold in silver and possibly buying game gold with dollars) than actual utility. There is no practical reason for the 10% final damage increase to cost as much as the first 90% increase, they are just messing with our players' OCD to nudge us towards separating us from our real life cash (and nothing wrong with it, in this form)
|
|
|
|
Post by hurtbot on Mar 16, 2017 12:00:15 GMT -5
I think the upgrade cost increase is there more for commercial reasons (forcing you to convert gold in silver and possibly buying game gold with dollars) than actual utility. There is no practical reason for the 10% final damage increase to cost as much as the first 90% increase, they are just messing with our players' OCD to nudge us towards separating us from our real life cash (and nothing wrong with it, in this form) That was my first thought on this as well. However if you think about it, the upgrade prices really serve two purposes, the first of which is purely financial. The second purpose is to keep players from getting powerful upgrades too early in the game. If driven only by the financial, you would see upgrades for the less popular robots like the Raijin that were much cheaper compared to popular bots like the Gareth. However that's not how it works. Maybe I'm wrong, but when I look at upgrade costs, it appears to be based the power of the bot, or rather(and perhaps more importantly) Pixonic's perception of the power of the bot. Since Pixonic has access to information that we will never see, this may be, albeit indirectly, the most objective data on relative power that we have. |
|
![AηɗυηєɗнєƖ [ǀƬA] Avatar](http://storage.proboards.com/6607038/avatar/qqeAJ4RKEanLTDewVuIX.jpg)
