This article is only a guide. Information expressed in a guide is usually more opinion than fact and should be taken as such. Guides are written by players, based upon their experiences, successes and mistakes, and are meant to aid other players. However, there may be differing opinions than those expressed in a guide.
Strategies and information in guides may not work for everyone.


Pumpkin Heads, Trick Staves

The following 4 Normal-Quality items are obtained randomly by trading sweets to a costumed NPC.

Horror Heads, Treat Staves

The following 4 High-Quality items are obtained by equipping the appropriate Normal-Quality item and then trading sweets to a costumed NPC.

  • Ensure that you only equip 1 NQ Event Item at a time, when trying to get the matching HQ item.
  • Note: If you equip a Trick staff, verify that you do not have a Grip equipped in your Sub slot.


The Harvest Festival is a collective and mutually beneficial effort by the players. It is in everyone's interest that sweets are sold efficiently and prizes are earned efficiently. Wasted sweets and gil simply disappear from our economy. Anything that we can do to optimize our strategies for sweets and prizes will benefit everyone.

First-Trade Theory

For a given player, his first trade of a given sweet on a Vana'diel day is more likely to earn a prize than successive trades.
This First-Trade Theory accords with the experiences of most players, but we do not have hard data demonstrating its superiority to spamming. However, we can point to anecdotes in which players traded 1,000 cookies to a NPC and received nothing. Spamming is unreliable.

How to Utilize the First-Trade Theory

Only trade one of each sweet on a given Vana'diel day. This will save you both money and time. This strategy is more reliable than spamming.


Trading stacks of a given sweet on the same Vana'diel day can sometimes produce results.
On Harvest Festival 2008 Guide , someone claimed to have received all four HQ items by trading 3 stacks of garlic crackers on a single day in Windurst. On the other hand, other people have claimed to have traded 1,000 cookies and received nothing. This would lead us to believe that spamming is not purely random. Factors such as the type of sweet, the particular NPC, or server trading activity may influence the effectiveness of spamming.

The author does not recommend spamming as a strategy, until more is known about its factors. Spamming contributes to waste of both the economy and time. Nevertheless, spamming is the easiest strategy, as there are often many stacks of a given sweet for sale next to a NPC.

Preference Theory

A particular NPC favors certain sweets and disfavors others.
If this theory is correct, then it is in everyone's interest that we discover which NPCs prefer which sweets.

Preference Theory Trial

To test the Preference Theory, select one NPC. Have a group of players try different sweets, focusing on that NPC. Document the results. This trial can be done across different servers, as the NPC should have the same preferences on each server.

Habituation Theory

A particular NPC habituates to (tires of) a particular sweet, if that sweet is repeatedly traded to him.
If this theory is correct, it explains the inconclusiveness of spamming. The more that you spam, the more that your chance of success decreases.

Habituation Theory Trial

To test the Habituation Theory, select one NPC. Have one or two players spam him. Then have them switch to different sweets. Document the results. This trial can only be done on a single server, as the NPC would have a habituation variable, such as Garlic_Crackers_Traded, that would reset after a period of time.

Familiarity Theory

A particular NPC becomes familiar with you after trading you a prize, and is unlikely to trade you another prize on the same Vana'diel day.
People seem to believe this, but data would be nice.

Familiarity Theory Trial

We only need the numbers to demonstrate that after someone gives you a prize, further efforts have less success than switching NPCs.

Stock Theory

A particular NPC has more of certain prizes 'in stock', and is more likely to give Pumpkins than Staves (or vice versa).
We have no opinions either way on this theory yet.

Stock Theory Trial

This trial can be done across different servers, as a particular NPC should have the same Pumpkin/Staff inclination on all servers.

One High-Quality Per Day Theory

You can only receive one high-quality Pumpkin/Staff per Vana'diel day.
This theory is easy to disprove. However, the weaker version is more interesting.
Once you receive a high-quality Pumpkin/Staff, your chances of receiving another high-quality on the same Vana'diel day decrease.
This weaker version is more difficult to demonstrate.


Please use the Discussion page to provide data about any particular theory. Thank you.