|
Rogmax11, I completely understand your sentiments and your intentions on this topic. You and I are definitely on the same page when it comes to realism and a desire to see plausible results. Your posting about credible statistical results was probably the most eloquent and concise summary of such concerns that I've ever read. Programmers should print that posting and keep it near their workstations at all times.
Unfortunately, most people in the text-based simulation community view anything like what you posted as an affront to their sensibilities. They don't understand that the people who desire realism and are willing to analyze a product critically are the ones that drive the market and help make these games as good as they are, despite their imperfections. It's not a matter of unfounded criticism or a cynical attitude. It's a matter of continually pushing for and encouraging software development that will lead us to the best possible results.
But those who strive for better products are confronted with the same unthinking responses all the time. "If you want total realism, then just watch games on TV or simply pour through the box scores for a given season." That is nonsense. No one is asking for an exact duplication of statistical performance.
Think of it this way. If players hit 100 home runs in a season or batted for a .560 average in your simulations, even those of you who are more forgiving when it comes to realism would not stand for it. But what happened with this Mark Prior example is just as ridiculous. Based on those numbers, his winning percentage would be .843. That is 126 points higher than the highest winning percentage in modern baseball history, which belongs to Spud Chandler at .717. He finished his career in 1947. Prior's winning percentage during that stretch would also be 47 points higher than the highest winning percentage of all time, which was set by Al Spalding in the 1870's, when only underhanded pitching was allowed and some pitchers were starting nearly 70 games per season.
The point is not to duplicate seasons exactly as they occurred. It's to simulate fictional seasons with some semblance of realism. If we saw players getting 100 home runs every season, we'd hardly accept OOTP as a viable simulation. But seeing Mark Prior go 38-1 is no different than that. The highest win total since 1900 is 40, set by Ed Walsh in 1909. But he pitched in 66 games and started 48 of them to achieve that number! Yet Prior nearly matched his win total by simply going out and starting every fourth game. Ed Walsh pitched 464 innings to achieve 40 wins. Prior pitched only 341 innings to get 38 wins.
That is not within the realm of what is rational and reasonable to expect in a simulation. Keep the word 'simulation' in mind. A 'simulation' is defined as an imitation or as "a representation of the operation or features of one process or system through the use of another." It is "a theoretical account based on a similarity between the model and the phenomena that are to be explained." The closer that the imitation, representation, or theoretical account is to reality, the better the simulation. That's all we're looking to do. We want to keep closing that gap, until any major glitches and anomalies are either non-existent or extremely rare.
|