One solitaire game problem can be online statistics manipulation.

[Editor’s note: This blog post was written more than a year ago, but only posted now.]

One of the common complaints on the Goodsol support forum is that the online statistics are skewed in some way. To be sure, there are some exploits that can be used in the flagship product, Pretty Good Solitaire, to manipulate the statistics in ones favor. In particular, one can select any deal (of 2147483647 available deals per game), so selecting known wins helps. Also, the starting position can be examined and, if no moves are made, abandoned without adversely affecting ones victory percentage. For the truly vain, unswayed by monotony, one can even play the same game repeatedly to boost their score.

Game secret: Due to the shuffling algorithm used, by design, the very last deal (i.e., game #2147483647) produces an unshuffled deck, which sometimes makes a victory easier. Veteran players who are really desperate to win a particular Solitaire game sometimes use this deal number in an attempt to get above 0%.

In addition to manipulating statistics via cherry-picking deals, there are a few other complaints. Some of the games have rules that make winning easier (or harder), so the statistics do not always compare like with like. Similarly, some players attempt to get high victory percentages, restarting a game repeatedly until it can be won (if possible), while others value the number of victories and play many more games to do that. Which is better: 100 straight victories in 100 games (100%), or 200 victories in 1000 games (20%)? That is a philosophical question.

We decided to seek a better approach for the online statistics to address some of the expressed concerns. My next blog post will discuss our solution to this issue.