Vision’s Incentive Systems: A Case Study

Every World of Warcraft raiding guild aspires to progress through raid content in a timely pace. Most members of the raiding corps focus on attaining gear and other goals, and so cannot be expected to directly satisfy the guild’s purpose. Thus, the guild must facilitate this indirectly by setting compatible goals, and then motivate raiding members to achieve them.

Human Resources (HR) departments of real-life companies handle these issues regularly, and the officerships of ingame guilds are no more exempt. I will introduce the theories associated with incentive systems, then apply them to analyse the incentive systems of my current guild, Vision of Frostmourne US.

Insights of psychology
Expectancy theory suggests that individuals are motivated to do satisfy the wishes of others if:

1. They are provided with a sufficient incentive/reward.
2. They can expect that their increased effort will lead to better performance. This is known as expectancy.
3. They can expect their improved performance to yield increased rewards and other favourable outcomes. This is known as instrumentality.
4. They can expect that consuming the reward will lead to improved satisfaction. This is known as valence.

If any of these conditions are not met, the incentive system is ineffective. The theory recognises two categories of rewards:

1. Extrinsic rewards, which are awarded to the individual by other people. Examples include payment and recognition.
2. Intrinsic rewards, which the individual realises personally. Examples include satisfaction and moral compatibility.

A given reward may either be desirable (e.g. payment) or an avoidance of something undesirable (e.g. moral compatibility).

Goal-setting theory suggests that:

• The more defined the goal, the more motivated the performing individual.
• The more the performing individual is involved in setting the goal, the less resentful they will be about attempting to meet it.
• Goals that are impossible to achieve discourage the individual from exerting the effort required for it.
• Goals that are too easy to achieve reduce the effort exerted and therefore the value of the individual’s contribution is diminished.
• Difficult but achievable goals maximise the value of an individual’s contribution without discouraging them altogether.

Agency theory, highly relevant to economics, covers the behaviour between a principal (who offers the incentive to work) and an agent (who has the ability to perform the work). It suggests that:

• The principal and agent have goals that are incompatible with each other; the principal wants work done of satisfactory quality, and the agent wants to exert the minimum effort required to earn a reward.
• The principal can motivate the agent to exert more effort by increasing the amount of incentive provided. However, by doing so, the principal loses the value of their own reward by the same amount.
• Both the principal and agent will do business if both of their rewards exceed their opportunity costs of not doing business. The principal can look for an alternative agent to complete their work, while the agent can look for an alternative principal to work for.
• The possibility exists that a given incentive will apparently be compatible with the principal's wishes yet encourages agent behaviour that is inconsistent with them.
• Information asymmetry between principal and agent occurs, where the agent knows information that the principal does not, in particular, whether the agent is actually meeting the principal’s expectations. The principal suffers a moral hazard, where the agent acts differently under different conditions, particularly whether they are performing work for the principal or not. As a result, the principal needs to monitor the agent’s work, which will cost them. (I have covered information asymmetry and moral hazard in detail in this blog post.)

The case study
Vision is a semi-hardcore raiding guild of mature-mannered players. Like most raiding guilds, they seek progression in current PvE content, though not at the fast pace of high-progression raiding guilds. They value players who attend raids regularly, are well prepared for them (with respect to both prior knowledge of encounters and consumables) and are loyal. They enjoy friendly banter both in guild chat and in the raid.

Vision’s loot policy:

• It is based on Dragon Kill Points (DKP) principles, where points (called DKP) are awarded to or subtracted from individual players under certain conditions. Accumulated points can be used as a claim on gear that drops during a raiding session.
• DKP is awarded based on both the duration the player stays in the raid and a subjective appraisal of individual effort. It may also be subtracted based on a subjective appraisal of poor effort.
• Limitations are placed on how much DKP a given player can stake as a claim on a given piece of gear. They may ‘bid’ a given amount within set bands, which varies on whether they are bidding for the main talent specialisation (main spec) or the off talent specialisation (offspec). The band for main spec bids is higher than that of the offspec bids. Bids that lie outside the respective allowed bands are ignored. The talent configuration is confirmed before the raid.
• The high bidder wins the piece of loot being offered. If there are multiple high bidders, the one with the highest DKP balance wins it.
• There is no maximum cap on DKP balances, and players are allowed negative DKP balances.
• If a player is substituted out on request of the raid leader, that player will be compensated in DKP.
• DKP erosion is not practised; it is deemed that the rest of the policy effectively handles issues that arise relating to it.
• The raid leader may decree a particular loot allocation “in very rare circumstances”.

Vision’s promotion policy:

• The non-official ranks are based on capability to contribute to raid sessions.
• 25-man raids are considered superior to 10-man raids.
• Officership is given to raid leaders, class leaders, recruitment officers and moderators (definitions according to this blog post).
• 10-man raid leaders are responsible for identifying, from the 10-man raids they run, candidates for the 25-man team.

The analysis
Vision’s purpose is to progress through current 25-man raid content in a moderate pace, and to provide a friendly social atmosphere. To facilitate the former, they set goals to perform first downs of planned bosses. Raiders are compensated extrinsically in gear and server presence; however some enjoy the experience and social setting and are rewarded intrinsically.

Items incompatible with expectancy theory:

• The raid leader and loot master are allowed significant influence on loot allocation decisions. The raid leader can override the bidding process, while the loot master can significantly influence the priority that different high bidders have. This somewhat reduces the expectation that exerting effort during a raiding session improves claim on loot, and therefore compromises instrumentality. As Vision is a social guild, the severity of this distrust diminishes with veteran members but not with new recruits.
• A significant portion of compensation from a raiding session is extrinsic gear. This scheme is fine when content is fresh, however, as more and more players take breaks late in a content patch cycle, the expected capability of the guild to complete a raiding session decreases. As a result, the vital expectation of compensation decreases, expectancy decays and players are discouraged from attending. This becomes a vicious cycle.
• The bidding bands may be highly limiting. If the maximum bid is set too low, there is less freedom to bid high. As a result, the expectation of a player to win a bidding session, and therefore the piece of gear, is low. Therefore, the instrumentality the bidding mechanism provides may be inadequate. This also discourages players from attending.

Items incompatible with goal setting theory:

• No goals are set for promotion to officership. As a result, the effort exerted, if any, to attain it is not well focused.

Items incompatible with agency theory:

• No reward is offered for pre-raid preparation. There is no positive incentive (for extrinsic-only players) to read boss strategies, nor any punishment for not applying consumables buffs. As a result, they cannot reasonably be expected to prepare and be well equipped to contribute to progression.
• The goals that are supposed to support the guild purpose are ill chosen. Rewarding for time spent in raid encourages attendance, but players are not motivated to make productive use of that time. In addition, new recruits may not trust the loot master’s appraisal of their effort. As a result, they may not direct much effort to progression, and a vicious cycle starts.

Every inquiry gives recommendations
Based on my analysis, I recommend that Vision:

• More clearly and objectively define the limits of the raid leader’s and loot master’s influence on DKP balances and loot allocation decisions.
• Weave additional activities and challenges into a raiding session when late in a content patch cycle. Examples include achievement runs, past content runs and conversation starters during trash clears. This should provide plentiful intrinsic motivation to attend.
• Extend the bidding bands, if not remove them altogether.
• Set criteria for officership that guild members are invited to attempt to meet.
• Reward pre-raid preparation. For example, applaud individual contributions to the guild’s boss strategies and threaten a player without a flask buff with a DKP penalty.
• Award DKP for boss downs and applaud heroic acts.
• Offer encouragement during moments of discouragement.

Originally posted here.