All these words spent on a superficial feature in a MMORPG

Now that I've put my chocobo out to pasture, I probably won't be writing any more about chocobo raising or racing unless another set of Official Races is introduced later. But I just want to go over a few observations I've made in the following areas as a reminder to myself in the future should I actually care about raising chocobos again.

Riding in the field

On FFXIWiki there is a chart mapping levels of a raised chocobo's strength to riding speed and endurance to maximum riding duration. It's generally accepted that raised chocobos at maximum strength are slightly slower than rental chocobos. (Way to provide positive incentives for raising chocobos!) Maximum riding time is also said to be between 15-45 minutes depending on your chocobo's endurance level, and it's also thought that Canter also raises maximum riding time (three more minutes?). This has intuitive appeal.

The other day, after finishing Hot & Cold I estimated my chocobo's maximum riding time to be about 37 minutes, which appears to be more than what could be predicted from the chart. But at least I can ride mine longer than I could a rental even though I never ride chocobos that long.

Of course, give it up for people who have the audacity to spout horseshit claims (present on the "Chocobo Raising" article on FFXIclopedia as late as July 30) that can't be verified or refuted easily, if at all. Seriously, riding speed is a function of discernment?

Free runs

Slightly less unimportant than the color of one's chocobo is how successful one's chocobo is in the free runs, yet I feel that these two things are often overemphasized from the occasional threads that I see pertaining to chocobos.

Sure, it'd be nice to get more than 28 chocobucks per eight races (in Windurst) since doing free runs is what a serious chocobo raiser is going to be doing the vast majority of the time. Yet are you really going to commit Speed Apples (or another item of your choice) to winning as many of these as you can, especially since you can trade only one Speed Apple at a time to the NPC? (Come on, SE, let's be allowed to stockpile items of the same kind for our convenience.)

Actually, it seems that before the June update people were able to beat the "super chocobos" (MeteorBrian, InvincibleLeg, StarOnion) on a regular basis.

For example, I've read that an SS/A/C/C chocobo (8/6/4/4) prior to the update was able to accumulate 36 chocobucks per cycle by beating the trio and StarOnion alone (twice) without items. Unfortunately, after the update, the outcomes have actually been variable and not in this chocobo's favor.

I've also read that an 8/8/8/1 chocobo beats the supers "about 50% of the time," while an 8/5/5/4 wins "about 30% of the time," in both cases without items. Presumably these observations were made after the June update.

Finally, an 8/5/5/5 chocobo (same as mine) before the update was able to beat the trio and MeteorBrian for 32/40 points in Bastok without items, but now is unable to beat any of them.

In my case, mine can beat StarOnion once in a while and can even come in second against the trio without items, but only occasionally. Perhaps it was a fluke, but when my chocobo was within the 8/6/4/5 range, it was able to beat InvincibleLeg and MeteorBrian at least once (but never beat the trio).

However, if it's true that even an 8/8/8/1 chocobo can't beat the supers regularly (low receptivity doesn't seem to be a big deal in free runs), then why even bother caring about free runs? It'd be much worthwhile to focus on winning the Crystal Stakes regularly instead.

The cost of Official Races

Yes, I finally completed the entire Pashhow Swamptrot series while I was finishing raising my chocobo, but I'm still not happy about the cost of entering these races.

You really get all your chocobucks from Hot & Cold? Bullshit.

At this point, I've grudgingly accepted the various elements of bad sportsmanship that are pervasive in chocobo racing. I suppose you could compare chocobo racing to Mario Kart. But in the Mario Kart games, you control your racer directly, racing outcomes are much less influenced by "random" factors, and you don't have to spend hours and gil doing busywork "free" runs to accumulate the chocobucks necessary to cover the entry fees for the higher-level races, because there are no entry fees. Winning lower-level races opens up access to higher-level races, and you don't have to pay entry fees. What a concept!

Yes, I am very pleased to have paid for the privilege of getting nickel-and-dimed by my Cardian competitors' AoE abilities and, as a further insult, having some insane taru observer (guess who) directly interfere with the outcome of the race. Yes, I find it amusing to compete against a cactuar that can use "Mad Dash" seven times in rapid succession (still, I won the Lethal Race my first try), and that the supposed creme de la creme of chocobo racing must use a Dream Apple (increases speed, stamina, and resistance to other chocobos' items) and be assisted by her fawning handmaiden in order to win (also won the Dream Race the first attempt).

So what if I got lucky on the last two races? Did I really have to pay at least 225 chocobucks total (more like 500 considering all the times I tried the Survival Race) for the honor?

Lower entry fees would make the costs more tolerable. Let us pay the entry fees in gil, even. But why chocobucks? If your chocobo simply can't win these races, I'm pretty sure getting trounced a bunch of times would be ample evidence of that fact (as my first "crappy" chocobo was), and you don't need the steep cost of chocobuck entry fees to dissuade you from persisting like an idiot. (Instead, hoard thousands and thousands of chocobucks to raise a chocobo that can win!)

I suppose that if another series of Official Races is introduced later, we should expect entry fees up to 100 chocobucks and I'll come to accept that grudgingly, too.

How sobering it is

I wish I had kept better notes on chocobo raising. I'm sure I wrote down when my chocobo's endurance finally fell to "better than average," but I can't find that scrap of paper at the moment.

I do know that endurance fell to "impressive" (below 192) after day 51, and that it fell to "average" (below 96) after day 62, so the endurance loss during that time should be around 96 points over 11 days, or about 8.73 points lost per day. Yet between day 43 and day 51 its endurance was in the "outstanding" range (192-223), which corresponds to a loss of about 32 points or 4 per day, which is actually the expected loss under "digging for treasure" according to ff11wiki.

I'm not sure how to explain the rapid loss of endurance between day 51 and day 62 unless DSC+8 from successful "digging for treasure" results in END-8 after the attribute cap is reached (thought to be 640 total points across STR/END/DSC/RCP). To put it another way, the care plan normally lowers endurance by 4 points, but when the attribute cap is in effect, endurance must be lowered by 4 more points to compensate for DSC+8 (otherwise the cap would be exceeded), for a total loss of END-8. The average of 8.73 points seems to be a result of my using strength training once during that time.

Note: When the attribute cap is in effect, raising an attribute with chocobuck training (strength in my case) leads to the reduction of the highest-level attribute that is not being raised (endurance). At least that's how it's thought to work.

If this is actually true, then this is convenient for raising an 8/7/4/4 chocobo at maximum strength without too much chocobuck expense. You could raise STR and END to the max with care plans, then use chocobuck training programs to raise receptivity to 94-95 points with 19 training sessions.

After nine successful days of "digging for treasure" (hope for no failures), you should then have 72 DSC ("a bit deficient"), about 219 END ("outstanding"), and exactly 640 total attribute points. Then after three more successful days of "digging for treasure," you should have 255/195/96/94 (640 pts).

You could then retire your chocobo and use one of the 100-chocobuck training tokens to raise RCP to 96 and lower STR to 253 accordingly. Then you could use another 100-chocobuck token to raise STR back to 255 and lower END to 193 accordingly.

If that plan actually works then you would have spent "only" 1,720 chocobucks to increase your chocobo's attributes directly, or as much as 490k gil and 41 hours of free runs.

In my case, not only did I spend 2,200 chocobucks to raise receptivity to "better than average," I also used 480 chocobucks worth of training programs (2 strength, 1 endurance, 2 discernment, 1 strength) between days 43 and 62 mostly out of impatience. And to top it off, I used 400 more chocobucks (1 endurance, 4 strength) to cap strength, leaving my chocobo at SS/B/B/B.

3,080 chocobucks. Somewhere between 657,000 and 880,000 gil. Over 73 hours of free runs. In the long run, hardly anything at all.

Well, I could have tried to rein in costs a little better, but with a minimum of 16 days required to raise discernment to 128+ with care plans (without any care plan failures), there most likely wouldn't have been enough time to finish and I would've had to resort to using chocobuck training anyway. (Of course, as it turned out, I finished care plan training after day 62 and discernment had just risen to "better than average," which was cutting it really close.)

The end of the road

After the day 54 progress report, I entered my chocobo in the "Survival Race" yet again, this time with "a bit deficient" discernment, and to my relief it finally eked out a victory after three attempts. On this successful run, it used a speed apple at the appropriate time during the final leg of the race (as opposed to the middle of the race the first time, when it finished second) on a final spurt directive, allowing it to pull ahead of Sakura at the end. This confirms to me the benefit of a sufficient level of discernment (when I actually get there).

However, it seems that due to the "random variability" that this skeleton crew of "developers" think is fucking cool in the form of "accidents" and cheap-fuck abilities and items used against other racers, the damn sheep actually wasn't in contention this time. Still, I finally managed to win with a still-developing chocobo--SS/A/D/B with Gallop--that had mediocre discernment and no Canter. Considering that these developers still hold the fucking conceit that imposing "Japan midnight" time restrictions everywhere (here, between non-repeat official races) serves a purpose other than to annoy the fuck out of players, I'm glad to get these piece-of-shit races out of the way when I can.

With that in mind I even tried the next race, the "Deadly Race"--similar setup again with mobs that can use their abilities twice (which is cheap) and yet another fucking annoying sheep--and also cleared it after three attempts.

I can't say that I'm ecstatic that I spent 195 chocobucks (about 55,714 gil worth of "free" runs for a total reward of 15,500 gil) for these two official races, but no one forced me to attempt these races with an unfinished chocobo, just as no one forced a group of players to fight Pandemonium Warden for 18 hours to the point of nausea. They knowingly accept the costs and risks of being part of the "vanguard" by fighting an unknown quantity. (Perhaps they should read up on the concept of "sunk cost" before trying again.) Who isn't surprised that the FFXI developers shit on our faces again and again and again and again and again, yet we continue to play? And I was so impatient about my chocobo's unexpectedly slow progress that I decided to roll the dice.

How slow does my chocobo go, anyway? From day 44 to day 55, my chocobo had been digging for treasure (care plan) twelve times, yet its discernment had yet to clear the "average" threshold. This indicates that the increase in discernment during this time was at best between 89 and 91, which does not jibe with claims that the most intensive plans raise an attribute one level in 3-4 days without failures. (In my impatience I used one session of training during this period.) This seems to agree with the idea that a chocobo's personality influences the effectiveness of care plans. It never took more than five days to raise either strength or discernment one level while "carrying packages."

Another explanation is that care plans become less effective as the chocobo gets older. A chocobo is in the prime of its life, so to speak, after the day 43 progress report, but I have never used anything other than "digging for treasure" after day 43. It's also possible that care plan results are constrained when the chocobo is at 640/640, which I finally thought to check today with Robino-Mobino.

Whatever the case, prospective chocobo raisers should consider these things should they strive for a certain attribute "profile." At this point I am very concerned that my chocobo's endurance won't "decay" fast enough so that I'll have to spend even more chocobucks adjusting strength, discernment, and receptivity in order to force my chocobo's endurance down. More chocobucks means more free runs. More free runs means more time. (Depending on how endurance pans out, I could have saved myself the cost if I had stopped endurance training around 192/255 instead of 224/255.)

Even so, at this point in my FFXI "experience," the time spent on free runs doesn't crowd out other things I could do in the game because there are no things I even want to do without having to cajole unmotivated people to do things like the Ashu Talif series of quests, or even some simple assaults to hoard points for Salvage. ZNMs can be spawned on demand, yet no one wants to do them. I am not so enamored with some rare/ex loot that will collect dust in mog storage. But neither do I care anymore about trying things that are off the beaten path (in other words, things that are off the path of least resistance).

I could go on and on, but it's a waste of time in the face of other people's lassitude. Other people can't squander my goodwill because there's nothing left to squander. As for soloing, that doesn't much appeal to me anymore. I will never step foot in Temenos again with the intention of soloing some Goblin Slaughtermen or doing anything else (other than zoning) now that I finally got a fourth Plaited Cord without having to do 10+ attempts between the third and the last, or wait 3+ months because some shitty LS is fixated on Apollyon to outfit with Homam shitty players with worse attitudes and personalities than mine. I don't care much to spend an hour obtaining access to Holy Cow just to be one-shot. Even if I were motivated to solo Genbu, the cost of Winterstone (650k seems to be a common list price if not a selling price) and Gem of the North (100k) means I should expect a net loss in exchange for the "fun" of kiting for hours and some shitty abjurations.

At this point I don't need to manufacture so-called fun through contrived challenges. Instead, I'll simply log on for minutes a day to feed and care for my chocobo and bide time before the next version update is revealed to be the barebones disappointment that I expect it to be, the anomaly of June 9 notwithstanding.

Heat without light

Hypothesis testing pertaining to game mechanics is mostly a waste of time because the kinds of questions asked are mostly a waste of time.

For example, on the BG forums I found a proposal to determine whether the critical hit rate bonus on Senjuinrikio increases the critical hit rate of the first hit of Blade: Jin, involving testing on "too weak" mobs so that the mob dies on the first hit of the weapon skill. Two sets of data, one using Mamushito +1 (DMG 38), the other using Senjuinrikio (also DMG 38), are to be produced.

Never mind how tedious it would be to carry out such an experiment. Never mind the widely accepted conventional wisdom that Senjuinrikio's crit bonus does affect Blade: Jin. Let's pretend a formal statistical test is actually worth using.

Then the concern here is what sample size (the same for each set of data) is "sufficient" enough to be able to detect the effect of Senjuinrikio on Blade: Jin? To answer that question, it might help to do some prospective power computations for a test of two independent proportions (under the Neyman-Pearson paradigm of hypothesis testing). In other words, given a sample size that is the same for each group, what is the probability of rejecting the null hypothesis (Senjuinrikio has no effect) when the null hypothesis is false (so that you are inclined instead to accept that Senji has an effect)?

When determining a "suitable" sample size, one practical concern is that the variance of a (known) binomial proportion is a function of both the actual value of the proportion and n, the number of trials, and that when the number of trials is fixed, the variance is maximized for p = .5.

So not only can you control the sample size for your experiment, you can also attempt to keep your "base" critical hit rate as far away from 50% as possible in your attempt to achieve adequate statistical power. If you want to try to keep your sample size low (relatively speaking), and you know people willing to be your bitch on demand, you could try to conduct these tests in Salvage and, hell, throw in some Stumbling Sandals, too. (I honestly don't know what kind of critical hit rates you can expect in Salvage, though.)

Assuming that the Senji does actually have an effect on Blade: Jin, power curves of this two-sample test (Fisher's exact test in this case) for various null proportions (base critical hit rates) show that the power is lower for higher null proportions:


Let's say you'll accept a power of .80 given a Type I error of .05 ("false positive"). If your base critical hit rate is 20%, you'll need a sample size of 639 for each sample proportion. But if your base critical hit rate is 5%, you'll need a sample size of 278.

Of course, you can always accept a higher Type I error in exchange for lower sample sizes for a given level of power:


Given a Type I error of .10, if your base critical hit rate is 20%, you'll need a sample size of 475 for each sample proportion to achieve a power of .80. But, if your base critical hit rate is 5%, you'll need a sample size of "just" 211.

Again, this is all under the assumption that Senji does have an effect on Blade: Jin in the way that we expect.

Let us suppose it is possible to achieve a base critical hit rate of 5% and, therefore, assume that the critical hit rate with Senji is 11%. Assuming this, here is some R code that estimates the power of the test of two proportions from 1,000 simulated experiments (as expected, Fisher's exact test will correctly reject the null hypothesis about 80% of the time), using n = 211 for each group:

p_value = rep(0,1000)
n = 211

for (i in 1:1000) {
  a = rbinom(1,n,.05)
  b = rbinom(1,n,.11)
  c_table = matrix(c(a,n-a,b,n-b),nrow=2)
  p_value[i] = fisher.test(c_table,alternative="less")$p.value

}

power = sum(p_value < .1)/1000
power

Do the math

Recently I fought Water Leaper (a pugil) in Sea Serpent Grotto for the Impulse Drive polearm weapon skill, and I took a gamble by fighting it near its spawn point, hoping that the nearby mobs wouldn't wander in my direction. Unfortunately, with about 1/5 of Water Leaper's HP to go, I eventually got aggro from a crab. I was still game to continue on since the crab was EP, but not long after I "reeled" in a nearby sahagin, which finally led to my demise. (It turns out that I-13 on the first map is a good place to fight it provided you sneak pull Water Leaper to that location after killing the sahagin that would otherwise block your way there.)

Since the fight itself is not interesting I'll instead use some probability to round out this filler post.

Each mob's attack round can be treated as an independent Bernoulli trial with some fixed probability of "success" (any result that does not result in my evading the attack). In this situation I could model an "effective evasion rate," or the probability that I evade all attacks after all mobs involved complete one attack round. (I'm not including parrying since its contribution is negligible.)

The pugil's hit rate, which I estimated at about .50 from a parse, is higher than that of the EP mobs. Still, independent events are involved, so my effective evasion rate with the pugil and the crab involved is estimated to be (.80)(.50) = .40. With all three mobs involved, my effective evasion rate is estimated to be (.80)(.80)(.50) = .32. 32% is not that much worse than 40%, is it?

Of course, a 32% "effective" evasion rate is misleading because even though the probability of getting hit at least once (or losing at least one shadow) after a series of attack rounds is .68, the probability of getting hit once (or losing one shadow) is .48, and the probability of getting hit two or more times (or losing 2-3 shadows) is an uncomfortable .20.

Barring a double attack, a pugil alone can only land one hit per attack round at worst. But the contribution of the other mobs is what allowed my shadows to be depleted so rapidly that there wasn't much of a chance to survive.

Here is a case where probability computations don't really belie a situation where it's obviously futile to take on three mobs at once, where one is IT and even though the others are EP. But still, one should be wary of all kinds of bullshit mathematical sophistries that are often left unchecked on FFXI-related forums. You might be able to call a thief on his BS claims that he was able to "solo" a bunch of IT mobs all at once, though.

Chocobo comments

After 2,200 chocobucks, about 628,572 gil, and at least 52 hours of senseless automated free race cutscenes (barriers to entry, right?), I've finally completed my chocobo's receptivity regimen and its current distribution of attributes stands at SS/S/F/B (or 8/7/1/5) STR/END/DSC/RCP. Now I just have to wait out the next 10 days of "digging for treasure" and hope I don't have to do much tweaking to achieve the desired SS/B/B/B profile.

As expected, it took 26 targeted chocobuck upgrades to improve receptivity to "better than average." This was accomplished using 14 training sessions (1,120 chocobucks or about 320,000 gil based on the expected value of 3.5 chocobucks earned per 1,000-gil free race entry fee) along with 12 training tokens (1,080 chocobucks or about 308,572 gil) set aside so that I wouldn't have to wait long to finish the training regimen while accumulating another 1,000 chocobucks to pay for the rest of the upgrades.

I did attempt the Survival Race again with this SS/S/F/B chocobo and, not unexpectedly, it placed 7th. I am not so irritated with the result as with the fact that these races have steep entry fees in gil cost and time cost. Now that's 120 chocobucks wasted on four attempts at the Survival Race, or about 35,000 gil and almost 3 hours of free racing.

Considering that the entire Pashhow Swamptrot series takes 225 total chocobucks to finish (about 64,286 gil and 5-6 hours of free racing) without repeated attempts, and the gil rewards add up to 48,000 gil, it's typical of SE to make gil rewards totally incommensurate with the wasted time and gil put forth. Sure, a Chocobo Pullus Torque, another shiny trophy, and a title signifying a difficult "accomplishment" are nice, but consider the cost of raising a chocobo capable of clearing the Swamptrot. What the hell do they expect chocobo racers to do for cheaper chocobucks, Hot and Cold?

I'm banking on my chocobo being able to rake in the gil in the Chocobo Circuit, though, so the gil costs ain't a thing. As far as the time spent accumulating chocobucks goes, I wouldn't excoriate people for being sensible about acquiring chocobucks with, ahem, unattended free races.