I know there has been a very lively discussion surrounding our Player Ratings. To a certain extent, some issues have been resolved. However, it seems there remains some general discomfort with our Player Ratings and a sizable number of people, inside and outside of FoolHQ, remain in support of Average Returns being incorporated into the player ratings structure. I have been doing quite a bit of thinking about Player Ratings lately and I wanted to offer my thoughts. I am currently a strong supporter of our Player Ratings. For starters, the tension our ratings create between score and accuracy expose players to a very productive thought process. Players must always wrestle with the following question: Is the relative 2/3 score gain worth the 1/3 accuracy loss? Or more generally: Is the relative gain of outperformance worth the risk of underperformance? This is a highly productive question that isolates the fundamental risk-reward tension behind every new investment. Each stock will have a different answer given its story. Depending on each company's financial statements, management, industry and valuation, each player must come to his/her own independent decision on whether or not the risk of loss is worth the potential gain. This remains the essential question of stockpicking – given everything I know about a company, is the reward worth the risk? Our player ratings expose this central question and force players to answer it seriously if they expect to perform well in the game. In encouraging constructive and thorough investment picks that carefully weigh risk versus reward, our player ratings generate intelligent picks and an intelligent community of picks. The Player Rating's goal is to efficiently generate productive community intelligence and the current algorithm achieves this goal quite nicely. Furthermore, a strong accuracy component reinforces a core tenent behind investing: only buy something for less than its intrinsic value. While this tenent is often associated with value investing, it truly applies to all investing methodologies; investors of every camp – value, growth, turn-arounds, asset plays, etc. - believe they are purchasing stocks worth more than the purchase price, albeit for different reasons. Accuracy emphasizes and reinforces this central tenent of investing. If you are truly buying a stock that is actually worth significantly greater than your purchase price, it will eventually increase in value and you will achieve full accuracy points. Accuracy spotlights direction of movement over scale of movement; if you are buying for less than intrinsic value, the stock should eventually move in the direction of your pick. A strong accuracy component encourages investors to stay reminded of this central, often-forgotten tenent of investing: only buy a stock if you are getting more in value than you what you pay up-front. Arguing for Player Rating to incorporate Average Returns misses the fundamental role of Player Ratings within CAPS. The purpose of Player Ratings is to compel players into producing the best possible community intelligence. The reason why portfolio games primarily flopped throughout the internet was that games based on the metric of average returns do not contribute to community intelligence…“So what this player happened to find this decade's best stock and turned his $100,000 into $20,000,000… what does that do for me now?” While arguably the most useful metric for an individual's portfolio, average returns do not contribute to productive community intelligence, the central role of player ratings. As I so often hear Todd Etter saying, “Why would a player be compelled to pick a stock he thinks will outperform the market if the pick can potentially bring down his stellar 50% average returns…” This remark is right on point. The average return component detracts from community intelligence by discouraging players from making picks that are potentially market-beating but would bring down the player's average return. Yet, we want players to make every pick that is market-beating because any stock that beats the market makes for a compelling investment. Doesn't it? It is instructive to extend this ratings structure to the extreme in order to see its real implications for the game. A CAPS Player Rating composed entirely of average return would compel players to find the fewest possible market-destroying investments. In the absolute extreme case, the top player would be the person who found the single best-performing stock. This person would have the highest average return since no stock would have performed better and his entire portfolio would consist only of this best performing stock. While an extreme example, this demonstrates the logic implicit within average return. Average return encourages players to find the single, highest-performing stocks. To a lesser extent, average return encourages players to create a very small portfolio of high-achieving stocks. After all, this is truly the ideal way to make the most money in your portfolio, without a doubt. However, small pick lists that only shoot for the stars will not produce a very smart community of investors. Such pick lists will omit all those picks that beat the market by a small margin, whereas we want all market-beating investments within our community. Our current player-rating algorithm encourages players to pick those aggressive high-achieving stocks, as well as all those stocks that will only slightly outperform the market. Therefore, our ratings generate intelligence around all potential market-beating investments. The key distinction again is the following: player ratings are devised so players are compelled to create pick lists that generate the most productive community intelligence. The current algorithms achieve this result because they force playes to carefully weigh risk-reward (accuracy-score) on a case-by-case basis and make any and all picks they believe will beat the market. An average return component detracts from community intelligence because it discourages players from making market-beating picks that could potentially subtract from their average return. In theory, CAPS was never meant to mimick real world portfolios. I think a large portion of the unrest stems from the fact that the appearance tempts one to associate CAPS pick lists with miniature portfolios. While this comparison is tempting, resist it, Fools! CAPS pick lists were never meant to mimick individual portfolios, in theory or in appearance. CAPS is the more foundational, bottom layer. It is a budding research platform for you to find your market-destroying investments. Once you find them, you can feel free to create your own tiny, aggressive CAPS-produced portfolios. Portfolio management and asset allocation are separate issues for which CAPS does not claim to provide great answers. CAPS is about providing you with a platform for generating the best market-beating investment ideas and the player ratings algorithm aims to flesh these ideas out of the community in the most efficient way. That said, portfolio metrics can be still be a useful front-end feature of CAPS. CAPS can spotlight the player's average returns as an additional metric for viewing by the community. This appears harm-free and adds another layer of value onto CAPS. I believe John Keeling is in full support of spotlighting average returns and other portfolio metrics somewhere on the CAPS page. However, back-end incorporation into the player ratings simply doesn't support the fundamental goal of generating the best possible community intelligence. To summarize, I do not think our current player ratings overweight accuracy. I think the tension between score and accuracy exposes players to productive risk-reward questions when considering each new investment. In the end, our ratings encourage players to patiently and deliberately pick every new market-beating stock. That sounds like an intelligent community to me! I am definitely open to further discussion and I have independent research I want to conduct on this issue before it is entirely resolved in my head. However, this is where I currently stand. I hope this helps clear up some confusion and people find these arguments persuasive.