Care Plan Arrow Sizes

[Sabaron]

I have a hypothesis that the arrows on the Chocobo Care Plan Chart are not linearly distinct, but distinct at a higher order of magnitude (e.g. square function--power of 2 or somesuch). However, I cannot at this time collect the necessary data as I do not have an active adult chocobo.

I need someone who (1) Uses Windower (yes, I'm aware it is technically a violation of the ToS, but people post Windower/FRAPS screenies continually without retribution AFAIK) and (2) has an Adult Chocobo. I want to know what the starting Stamina values (percentage) for a chocobo on each of the three tiers of care:

Tier 1: Basic, Walk, Music, Exercise, Playing
Tier 2: Packages, Exhibition
Tier 3: Messages, Digging, Acting

I wish to use these values to determine relative value point sizes for the three arrows and then use these relations to work through mixed attribute care plans. I should be able to get the first two values, but the third tier value I cannot do at present.

It would also be interesting to note if there are any difference between the Stamina values on care plans of the same tier. An indication that some "medium arrows" could be larger/smaller than others would foul up the results a bit, but as long as there is some sort of basis available for the "Threshold" at least of the arrow size, it will be better than current.

[Nakti]

Re: Care Plan Arrow Sizes

At first I was just visually eyeballing the HP stamina bar, but that wasn't exact enough so I've been using Clipper to take screenshots. After I log off FFXI I use Paint to count the pixels.

Basic Care - 1 pixel down

Listen to Music - 2 pixels down
Walk in Town - 2 pixels down

Exercise in Group - 3 pixels down
Exercise Alone - 3 pixels down
Play with Children - 3 pixels down
Play with Chocobos - 3 pixels down

Exhibit to the Public - 10 pixels down

Act in a Play - 17 pixels down

Caveats: Except for Basic Care which I continue to use, all the other plans were pre-update only. And I only did Exhibit and Act in a Play one time each. Since their stamina reduction is so large, there will probably be a variation in the number of pixels based on how much total stamina the chocobo has for that day.

Edit: Not sure if "success" or "failure" affects stamina, but...

CrystalLight made some children cry during Exhibition (failure)

CrystalLight showed up all the actors during Acting (success) - this one was very unexpected since her attributes are all still Poor.

[Sabaron]

Re: Care Plan Arrow Sizes

Hmmm. I suspected that there might be varying "sizes" of "small", "medium" and "large" arrows. That will complicate matters a bit. If there's a variance among the medium/large arrows. Ok:
Value Chart 1:
All medium/large arrows are fixed at 10/17. Small arrows are always the same size when in the same plan:
[CODE]
Plan STR END DIS REC AFF ENG
BASIC 1 1 1 1 -1 -1
REST 0 0 0 0 0 3
WALK 2 2 -2 -2 -2 -2
MUSIC -2 -2 2 2 -2 -2
E-ALO 3 0 -3 -3 -3 -3
E-GRP 0 3 -3 -3 -3 -3
P-CHI -3 -3 3 0 -3 -3
P-CHO -3 -3 0 3 -3 -3
PACK 10 10 -10 -10 -10 -10
EXHIB -10 -10 10 10 -10 -10
MESS 17 0 0 -10 -17 -17
DIG 0 -10 17 0 -17 -17
ACT -10 0 0 17 -17 -17
[/CODE]

#2: Balanced Exercise/Play (using STR-DIS affinity, END-REC affinity)
[CODE]
Plan STR END DIS REC AFF ENG
BASIC 1 1 1 1 -1 -1
REST 0 0 0 0 0 3
WALK 2 2 -2 -2 -2 -2
MUSIC -2 -2 2 2 -2 -2
E-ALO 3 0 -1 -2 -3 -3
E-GRP 0 3 -2 -1 -3 -3
P-CHI -1 -2 3 0 -3 -3
P-CHO -2 -1 0 3 -3 -3
PACK 10 10 -10 -10 -10 -10
EXHIB -10 -10 10 10 -10 -10
MESS 17 0 0 -10 -17 -17
DIG 0 -10 17 0 -17 -17
ACT -10 0 0 17 -17 -17
[/CODE]

[Sabaron]

Re: Care Plan Arrow Sizes

The High Energy care plans seem to reflect a "loss of value" over the medium energy care plans since the arrow is only worth 17 in this scenario. Were it worth 20, it would be equal value. Obviously, the mid-arrow care plans are very effective. Unfortunately, the data I wanted doesn't seem to be evident at this time.

Basically, if the BIG arrow is worth less than 2 times the medium arrow (17) then there is a requisite loss of overall value in taking the high-energy care plans. If the BIG arrow is worth exactly 2 times the medium arrow (20) then there is neither a loss nor a gain in overall statistical value. If the BIG arrow is worth more than the medium arrow (23) then there is a benefit in taking the high-energy care plans over the medium energy ones.

Note that the high-energy plans have a net gain of 7 points, but that this will likely be useless for most builds, since the attribute that the care plan lowers will already be at 0.

[Sabaron]

Re: Care Plan Arrow Sizes

A projection in retrospect of my (failed) END/DIS build @ 50 days
[CODE]
Plan STR END DIS REC Plan Days
DIG 0 -17 17 0 20
PACK 10 10 -10 -10 20
BASIC 1 1 1 1 10
-----------------------------------------------------
DIG* 0 -340 340 0
PACK* 200 200 -200 -200
BASIC* 10 10 10 10
=====================================================
TOTAL 210 -130 150 -190
[/CODE]

Total of positive values: 360

Projection for the STR/END Build at 50 days:
[CODE]
Plan STR END DIS REC Plan Days
WALK 2 2 -2 -2 10
PACK 10 10 -10 -10 30
E-STR 3 0 -1 -2 5
E-END 0 3 -2 -1 5
-----------------------------------------------------
WALK* 20 20 -20 -20
PACK* 300 300 -300 -300
E-STR* 15 0 -5 -10
E-END* 0 15 -10 -5
=====================================================
TOTAL 335 335 -335 -335
[/CODE]

Total of positive values: 670

The STR/END build is approximately 86% more effective at raising overall stats as the END/DIS build.

[Macht]

Re: Care Plan Arrow Sizes

The High Energy care plans seem to reflect a "loss of value" over the medium energy care plans since the arrow is only worth 17 in this scenario. Were it worth 20, it would be equal value. Obviously, the mid-arrow care plans are very effective. Unfortunately, the data I wanted doesn't seem to be evident at this time.
Basically, if the BIG arrow is worth less than 2 times the medium arrow (17) then there is a requisite loss of overall value in taking the high-energy care plans. If the BIG arrow is worth exactly 2 times the medium arrow (20) then there is neither a loss nor a gain in overall statistical value. If the BIG arrow is worth more than the medium arrow (23) then there is a benefit in taking the high-energy care plans over the medium energy ones.
Note that the high-energy plans have a net gain of 7 points, but that this will likely be useless for most builds, since the attribute that the care plan lowers will already be at 0.

I would agree with that if the 3 adult plans actually effected all stats (Considering STR-DUR-DIS-REC only) however it's only effecting 2 stats. If done right you can actually sustain the Strength and Durability you've already gained while gaining 7 points in DIS & REC with 3 Care Plans (Packages, Dig, Act) done equal amount of times. That is if the stat increase is this linear, some other suggestions seen is that some plans were not linear they had percent gains and percent losses.

In it all though it seend Endurance would be the toughest to improve.

[Sabaron]

Re: Care Plan Arrow Sizes

The above is obviously very rudimentary, and one must keep in mind that certain breeds of chocobo may have inherent bonuses to growth in their requisite attributes either from breed (not necessarily color as it states in the June 19th Chocobo post from SE that most chocobos are yellow regardless of breed--I assume in order to add to the confusion) or from inheritance--I'm not sure how either works yet.

If growth is percentage based, then we get (I'm assuming) exponential growth or (not assuming) logarithmic growth. If the former is the correct assumption then a directed careplan (focusing on a single attribute with say high-energy care plans) will yield excellent results whereas if it is logarithmic, the more general careplans (e.g. medium energy or mixing care plans) will be most effective. If the growth is linear, then the determination will be made based on the relative values of the medium/large arrows.
I'm curious to see actual data on exponential growth. I assume that the measurement involves counting the number of days between "ticks" in the chocobo's attributes (e.g. 5 days to Substandard, 4 days to Bit Deficient, etc.). Of course, we are again working under an assumption--that the attribute labels are a linear representation of value, which is a fairly safe assumption.

With respect to increases in attributes marked with "--" in the Care Plan Table, I think that if 0 means "something greater than 0" then the table is deceptive. I doubt SE intends to deceive us using the table. I will, therefore, have to state that is is possible for a chocobo doing nothing grows a little bit in everything all the time and that leaving a "--" on the care plan for a long enough period of time will be transparent to the underlying non-care related growth effect. Feed and Direct Care may also be responsible for these increases; for instance, telling stories may increase mental attributes while racing may increase physical ones. I very much doubt SE purposely lied to us about the effects of the care plans so I'm assuming only the marked attributes are affected by care.

--

Again, I'll explain that by including the two "off-days" on my care plan (STR/END see above) doing Exercise, I'm attempting to achieve Receptive status every 7 days so that I can feed SolarWind a Parasite Worm with greatest efficacy--hoping to boost stat growth overall. I haven't decided how long I'll continue this cycle. At present, I haven't determined whether or not it will be of benefit to interrupt his care plan with Deliver Messages occasionally once he reaches adult stage (e.g. just after his weekend) if it is not at least equally effective wrt stat growth as I'm attempting to get the maximum "bang-for-buck" riding chocobo by maximizing the sum of STR and END. I'll probably throw it in just for experimental curiosity.

[Zeenna]

Re: Care Plan Arrow Sizes

Hi, don't see where the post I replied is from last night, but I did reply that the pixal system seems to only reflect the energy bar and doesn't take into account the affects of the additonal care plans like reading and competeing and what that has on the "point" system.

[Sabaron]

Re: Care Plan Arrow Sizes

This particular analysis is looking at relative magnitudes of the arrows used by FFXI in their Care Plan table. It doesn't take into account any direct care or food. Basically, when one creates a diagram, one uses a similar frame of reference to draw the various sections of the table or the table has absolutely no meaning unless one is attempting to distort the truth. Therefore, I'm assuming the relative effect of the Big Down Arrows on Energy is at least proportional to the Big Up Arrow effects on the attributes and that the magnitudes of the different arrow sizes have a variance which is significantly smaller than the relative magnitudes of the arrows. There isn't really a lot of data here. Therefore, the pixel counting for the energy bar only reflecting change in the energy bar is a moot point since I'm not actually trying to determine exact numbers of "points", but only very rough magnitudal differences between the arrows on Square Enix's Table.

[Macht]

Re: Care Plan Arrow Sizes

Well the only other problem with what you are showing is that a data mining of the chocobo's attributes stated that they can't exceed 255 yet you are saying your chocobo achieved 335. That should be impossible and if the chocobo had achieved 255 then they should be maxed out in that stat.

Also since 255 is the max getting a negative of 335 also seems impossible since Poor is the lowest range and again from the data mining showing that the stats are controlled by 1 byte each would mean the lowest number you can achieve is 0.

Kind of why for me I believe that the arrows has to be indicating something close to this:

Small = 1 point (Take 255 days to max)
Medium = 2 points (Take 128 days to max)
Large = 3 points (Take 85 days to max)

Then the matter of food I would guess them to be something like:

Vomp/Zephram = ~.46 points (189 days to max)

This would explain some statements of the carrots with a small negative effect would take 52 days to raise a stat. If you are feeding it 3 carrots a day you are getting 1.39, however if it has a small negative arrow you get only .39 (Amount low enough that it doesn't ping a byte for it). So that would mean that STR won't gain a point till about 3 days of feeding (9 carrots) which I'm figuring would be about 52 days (Assuming that stat changes take place around every 20-40 points).

If they do work that way then feeding straight out carrots that go with the stat you are focusing on then you would get this result:

Small = 2.35 points (109 Days to max)
Medium = 3.35 points (77 Days to max)
Large = 4.35 (59 Days to max)

More then enough time to max stats especially if after that peak message the chocobo get even more points then if you stuck to your guns on specific stats to raise you should have it at First-class I'm calculating by like day 64 or 65.

Mine I've obviously messed up on, and gotten no were near that with only like 1 at Average so far. Then again it's also a quandry of if the genetics had anything to do with the results as well.

[Sabaron]

Re: Care Plan Arrow Sizes

The negative values are implied 0's. I just didn't change the values in the table. As for the 255 cap thing, that would imply that the magnitudes of the arrows are dissimilar as the growth rate would have to be much smaller to show changes. Just because there is a 255 cap on what you can see in the data mine does not necessarily correlate to the action on the server side. The server can keep a "buffer" over time of "XP" in each of the four attributes and "level up" your chocobo according to server-side rules, but it is also entirely possible that you always get a "tick", so the minimally effective set of boosts would be this (If Walk/Listen/Exercise/Play is only 1 point, then they are useless since you get a much better effect with basic.):

Tier 0: 1 tick (Basic)
Tier 1: 2 ticks (Walk/Listen)
Tier 2: 3 ticks (Exercise/Play)
Tier 3: 4 ticks (Packages/Exhibition)
Tier 4: 5 ticks (Messages/Digging/Acting)

Let us assume a full strength care-plan:
[CODE]Baby: WALK
Young: PACK
Adult: MESS
[/CODE]

Baby (14 days): STR+28, END+28
Young (10 days): STR+40, END+40 => STR=68, END=68
Adult (31 days?): STR+155 => STR=223, END=68

Which fits within the parameters and how one would successfully max out a chocobo. However, this plan discounts the medium sized double arrows on the (much easier to maintain) Tier 3 Careplans:

Let us assume an even physical care-plan:
[CODE]Baby: WALK
Young: PACK
Adult: PACK
[/CODE]

Baby (14 days): STR+28, END+28
Young (10 days): STR+40, END+40 => STR=68, END=68
Adult (31 days?): STR+124, END+124 => STR=192, END=192

This chocobo is obviously better, so in order for the Big arrows to be big enough to be (at least) on par with the efficacy of the medium double arrow we need to do this:

Basic: 1 tick on everything
Tier 1: 2 ticks (Walk/Listen)
Tier 2: 3 ticks (Exercise/Play)
Tier 3: 4 ticks (Packages/Exhibition)
Tier 4: 8 ticks (Messages/Digging/Acting)

The Energy bar data shows that the order of magnitude for the energy loss is 1:2:3:10:17 instead of 1:2:3:10:20 implying either an increase in efficiency (Energy:Effect) with the top-tier plans or a decrease in benefit. SolarWind's Care plan then looks like this:

[CODE]Baby: WALK x 14
Young: PACK*8, E-ALO, E-GRP
Adult: PACK * 4, MESS, WALK x 2 (over 31? days) [P19, M4, W8]
[/CODE]

Baby (14 days): STR+28, END+28
Young (10 days): STR+35, END+35
Adult: STR+92+ΔMess, END+92
Yield: STR=155+ΔMess, END=155
ΔMess (@5) = 20 => STR=175, END=155
ΔMess (@7) = 28 => STR=183, END=155
ΔMess (@8) = 32 => STR=187, END=155
ΔMess (@9) = 36 => STR=191, END=155

Now, let me get into some notes on the careplan:

The only reason he takes reduced care plans is in an attempt to illicit Receptive status in the hope that it is beneficial to feed a parasite worm at this time resulting in an overall increase in statistics over that of the straight-through care plan. If I am unable to illicit Receptive state on the first week of his adult care plan (the Young Care Plan Weekend of E-ALO, E-GRP failed to produce Receptive state, btw), SolarWind will very likely only get one day off just to recover from the MESS care plan and it will be a WALK day or an E-GRP day depending on whether or not I can get Parent back with E-GRP in one day.

Additional Stuff:

Observation of Percentage (Exponential) Growth in statistics: Having Exponential growth of attributes mixed with a byte-width ceiling (255), is a bit hairy. Small exponential growth would get truncated (since we can't store a decimal) and large exponential growth would allow you to cap the stats rather quickly since, as you can see, the Even-Physical plan above hits 192 with linear growth.

Square-Enix stated in the last patch release notes that they made Chocobo attributes increase more quickly. Looking at the 256-ceiling, how much, I wonder, could they have done?

Also, if we allow for different breeds to have a ΔStat bonus, we can't give them "small" bonuses because we're restricted to integers, and giving them larger bonuses will allow them to cap faster.

If there is no server-side non-256-based implementation, I would expect to see early capping of stats on chocobos raised post-patch especially in non-yellow ones (definitive breed). It is, however, entirely possible that you only get to see a single byte-width client side because it allows SE to generate smaller data packets on an already laggy server. This is what I would do had I made the system--since the client needn't be concerned with fractional growth, we needn't send them that part of the mantissa, saving us several bytes per packet.

Carrots: This also implies that the minimum effectiveness of a carrot is 1... assuming 2 carrots per day for 63 days, that's +126.



------------------

The "points" I'm referencing above are indicative of a server-side thing, not what you actually get and they do not have to correspond to the 256-based values you see client-side--they can be modified by an unknown fractional coefficient. If Walk/Listen/Exercise/Play is only 1 point, then they are useless since you get a much better effect with basic.

[Sabaron]

Re: Care Plan Arrow Sizes

Even more stuff: Minimal Efficacy of Care plans
Tier 0: 1 tick (Basic)
Tier 1: 2 ticks (Walk/Listen)
Tier 2: 3 ticks (Exercise/Play)
Tier 3: 4 ticks (Packages/Exhibition)
Tier 4: 5 ticks (Messages/Digging/Acting)

At this level, note that Exercise/Play is inferior in raising overall stats to Walk/Listen. If we make the assumption that a care plan is at least as effective as the one below it then we get:

Tier 0: 1 tick (Basic) [Total Effect: +4]
Tier 1: 2 ticks (Walk/Listen) [Total Effect: +4]
Tier 2: 4 ticks (Exercise/Play) [Total Effect +4]
Tier 3: 5 ticks (Packages/Exhibition) [Total Effect +10]
Tier 4: 10 ticks (Messages/Digging/Acting) [Total Effect +10]

This can be applied to the 1:2:3:10:17 ratio as well to yield:

Tier 0: 0.25 tick x 4 (Basic) [Total Effect: +1]
Tier 1: 1 tick x 2 (Walk/Listen) [Total Effect: +2]
Tier 2: 3 ticks x 1 (Exercise/Play) [Total Effect +3]
Tier 3: 5 ticks x 2 (Packages/Exhibition) [Total Effect +10]
Tier 4: 17 ticks x 1 (Messages/Digging/Acting) [Total Effect +17]

I would like to also add that all of the above is guesswork and everything I'm doing comes down to a single methodology:

Goal: Maximize(STR+END)

• Baby: Our only options are Walk or Basic--since all of our mental work will be undone and the assumption is that you get at least a little more effectiveness out of directed care, the only viable choice to reach the goal is Walk.
• Young: Options are Walk, Exercise, Packages--since the arrows are bigger for Packages than for either Walk or Exercise, I'm assuming the plan is more effective than either Walk or Exercise at raising an individual statustic. Therefore, the only viable choice is Packages. There is, however, a caveat (see below).
• Adult: Our options are now Packages and Messages. This is the most experimental portion of the care plan. If Messages is more than twice as effective at raising STR as Packages is at raising both STR and END, then it can be chosen as a substitute for Packages. If it is equally effective (i.e. MESS[STR] = PACK[STR + END], then choosing it is counterproductive as it is highly detrimental to both the chocobo's affection and energy. If, on the other hand, it is not at least as effective, then choosing it is extremely counterproductive as it is very much more harmful to the chocobo's affection and energy.

Caveat: Parasite Worms
If, as has been suggested by Macht in another thread, Parasite Worms increase ΔStat values either on a temporary or permanent basis, and this effect is made significantly better when giving the worm during Receptive state, then it is beneficial to take a lighter care plan for a certain duration in order to achieve Receptive state and feed the worm at this time.

I'm currently working on these assumptions:

1. MESS[STR] > PACK[STR+END] at "Parent" affection.
2. The benefit of a Receptive feeding of a Parasite Worm is significantly greater than the benefit of a non-receptive feeding and this benefit exceeds the total value of statistical bonuses that could be achieved by the PACK care plan on those days that must be taken off in order to achieve Receptiveness.
3. That "Wants to be with you all the time." affection level makes the following true: MESS[STR] <= PACK[STR+END]

[Macht]

Re: Care Plan Arrow Sizes

Can agree with the fractional growth and all, but all I've got to say is who's going to care about a huge number growth if in the end all that the client is getting is a single byte. Or more simply saying since the client is only getting 1 byte why try to make it more complex then that?

You can get very complex but if means someone else can simplify your 10,000 point bases to the same equivalence in a more fitting and easier managed number range then why would someone bother with this large number set that the server might create. I for one can certain go out and start building out a huge probability that the server does, but seriously I don't care about that.

What's of more intrest to me is alright I feed a chocobo a carrot what will this translate into for the data the client is going to see. Does this mean my feeding the chocobo 3 carrots a day is only going demonstrate a change every 3 days, 2 days, 1 days? and what actions I do that would alter this.

As far as I've been able to build using the number equivalent range that the client seems to get that helps to make a more understandable responce that can also be proven by the data mining results.

[Sabaron]

Re: Care Plan Arrow Sizes

I'm curious to know if anyone has actually extracted concrete daily values for their own chocobos from the data sent over by the server.

I'm really only interested in relative magnitude of the care plan arrows in the SE Diagram, not discrete values (read above at my care plan assumptions).

I don't care what the numbers actually are either server or client side, only that LARGE arrows are at least 2x better than MEDIUM arrows otherwise the High-energy plans are counterproductive to my goal as stated, Max(STR+END).

The pixel width data suggests that it is possible that the BIG arrows could be at least 2x as powerful as the medium ones in that the ratio difference between MEDIUM and LARGE energy arrows is 10:17 which could be indicative of an efficiency increase and thus the ratio MEDIUM:LARGE stat arrows might actually be 10:20 or better. If it's better than 10:20, then I'm go for LARGE arrow care plans; if it's 10:20 or worse then I'm not keen on them (for this care plan at least).

A second notion revolves around the effects of affection on care plan effectiveness. At what level affection would the M:L ratio be better than 10:20 and when would it be equal/worse? I'm fixing M's value since I can maintain "Parent" during a continuous cycle of Package Delivery, but I cannot during "Messages".

[Macht]

Re: Care Plan Arrow Sizes

I'm curious to know if anyone has actually extracted concrete daily values for their own chocobos from the data sent over by the server.
I'm really only interested in relative magnitude of the care plan arrows, not discrete values (read above at my care plan assumptions).
I don't care what the numbers actually are either server or client side, only that LARGE arrows are at least 2x better than MEDIUM arrows otherwise the High-energy plans are counterproductive to my goal as stated, Max(STR+END).
The pixel width data suggests that it is possible that the BIG arrows could be at least 2x as powerful as the medium ones in that the ratio difference between MEDIUM and LARGE energy arrows is 10:17 which could be indicative of an efficiency increase and thus the ratio MEDIUM:LARGE stat arrows might actually be 10:20 or better. If it's better than 10:20, then I'm go for LARGE arrow care plans; if it's 10:20 or worse then I'm not keen on them (for this care plan at least).
A second notion revolves around the effects of affection on care plan effectiveness. At what level affection would the M:L ratio be better than 10:20 and when would it be equal/worse?

There was one guy I use to talk who was collecting that data out of the data mining he was doing. Showed me a couple days of it to demonstrate, been like 2 weeks since I last heard from him. Not sure if he's doing it or not, but the info he had showed did suggest that the care plans, foods, personal care effected those single byte numbers directly. Just how it did is still something I don't exactly know.

Yeah, the large care plans I'm still trying to figure out their effectiveness. I am noticing that it seems like the large arrows in order to be effective have conditions to them, just what the conditions are I'm still trying to figure out. I do know that feeding foods that related to the stat I was raising it definatly heped it raise a heck of a lot faster, but it seems without that food it's like using the large increases is useless it doesn't seem to budge at all and chocobo gets the failed CS a lot more without the proper food.

Leaves me to speculate that probably the largest arrow plans is only effecient to use if you have the right carrots for them.