Here is our solution to online statistics manipulation.
As mentioned in my previous post, Problem: Statistics Manipulation, we have some players who manipulate their online statistics for solitaire games using a variety of methods. Equally, we have players complaining in the Goodsol Solitaire Forum about this practice and asking for something to be done. Worse, some players have stopped participating and reporting statistics due to the unreliability.
In an attempt to address this issue, and generally make our games more fun for our customers, we designed a new mode of gameplay called (for now) Climb Mode.
In Climb Mode, a user plays the deals for a particular game in strictly numerical order, starting with deal #1. This eliminates the ability to cherry-pick or replay deals to inflate victory percentages. At the same time, it assures that all players are compared on results from exactly the same deals. (This mode actually removes several features, such as Select Game, Previous Game, and New [random] Game.) Further, the game options are always fixed, so easier (or harder) variants are not available.
To address the question of which types of gameplay is preferred, we created a cumulative scoring system, where the score on a certain game is the total of all scores for deals played. This allows players (like me) to play for the maximum number of victories, hence scoring the most points per deal, to compete fairly with other players who play as many games as possible, rather than replaying each trying for a victory. Once one moves on from a game (by either winning or using Next Game), one cannot go back.
All games are timed in Climb Mode, and average time per deal is reported along with the total score. Ties in score are decided in favor of the player with the lower average time. Unlike in standard mode, the timer does not reset when restarting a deal, so one can replay as often as desired, but the extra attempts are reflected in the overall game time. (I have several deals in Free Klondike where I eventually won, but only after the timer maxed out at 99:59.9.)
Of course, the implementation of Climb Mode does not completely remove the possibility of cheating, as somebody who is truly determined could still potentially manipulate the results. However, we operate from the philosophy that anybody who goes through that much trouble to cheat in a friendly competition has bigger issues. In practice, since only paying customers are allowed to submit online scores, this has never been a problem, and the situation is certainly better than before.
Note that we decided to first implement the above changes in Pretty Good MahJongg rather than Pretty Good Solitaire. Now that the system has proven itself over the last few years in Pretty Good MahJongg, will Climb Mode be implemented in the next version of PGS? As the Magic 8 Ball says, “Signs point to yes.“