Realism vs Precision
Which is more important?
Realism vs Precision
This Delphi Forums thread caught my eye the other day.
It’s actually a pretty interesting topic — one that seems to come up again and again when it comes to sports sims. I’ve got a few thoughts that I thought I’d jot down here for the sake of discussion.
“Realism” is in the eye of the beholder. It’s hard to define what is and is not “realistic.” Some people are fine so long as the “right team” wins in the end. Some people want offensive statistics to fit within a certain range of real life results. Some people demand that certain historical anomalies are replicated, such as Jim Palmer never giving up grand slam home runs. In that sense, it’s difficult to determine just how “realistic” a game engine is or is not.
Games are about story telling. I don’t know of any replayers who play games simply for the sake of creating statistics. We don’t really care that much about whether a guy hit 35 home runs or stole 25 bases. The thing that keeps us coming back is the story that plays out in our head as we play through a season. In my experience, the more you focus on that story, the more enjoyable your project will be regardless of which game engine you use.
These games are games, first and foremost. The problem with precision-focused modifications is that they tend to clog up the flow of the game. In fact, one trend I’ve noticed over the years is that older replayers tend to move away from advanced games and towards basic games. At some point in time you realize that the extra dice rolls aren’t really adding much to the experience, and that you’d rather keep your project going than bog yourself down with a bunch of technical details. The games that have stood the test of time are games, first and foremost.
Precision can upset accuracy. This is because of a little thing called “sample size.” We most frequently run into this problem in the realm of lefty/righty split results. The vast majority of baseball players have such a low sample size of split at bats over the course of a single season that the results are more likely statistical noise than anything significant. Creating a game based on statistical noise can give you some bizarre results. I’ll talk more about this in a future post — but the truth is that lefty/righty splits tend to converge to historical norms once the sample size is high enough. In other words — you can get perfectly acceptable results using a generic split system rather than messing around with single season split totals.
Let me know what you think!
Spot on. The bit about returning to basic games as we get older resonates with me. I'm only 51, but we'll on my way to preferring simplicity over added precision.
When I play, I want the players to perform accurately, but not necessarily precisely. I want to feel like Im at the game, and that George Foster is a power threat, Pete Rose is likely to drive me nuts with clutch hits, and the Orioles pitching staff of the late 60's/early 70's gives up runs grudgingly, if at all.
I want to feel like Im at the game, and I want to do so without having to refer to endless chaarts or perform multiple mathematical calculations. Most of the time, a single should be a single, a homer should be a homer, and I shouldn't have to go past the first roll of the dice to find out.
Occasional extra rolls that add excitement, such as Strat-O-Matic ' robbing homeruns rule, add excitement to the game without bogging down the play.
Keep it simple, make it come out within a believable realm of possibility, and don't break my immersion with too many extra rolls and calculations, and you've got me.
I don't care about Lefty/Righty splits, but I do care about league totals. The sample size is large enough over 2,100 games (162 in AL and NL for 26 teams) to see how accurate a game is. Skeetersoft is better (more accurate) than APBA. APBA's Basic fielding is poor. Does anyone else keep individual fielding stats or just me? I want to see how accurate the game is and in some instances, it's pretty bad - both APBA and Skeetersoft. The SS's, 3B's and RF's (RF-1) get screwed. I think the creator's didn't care about individuals as much as Team's in that sense. Why don't the P's, C's and 1B's get any love individually. Keith Hernandez is just as valuable as Doug Ault at 1B if both there teams are in the same Fielding category (F1, 2 or 3). That's just dumb. I think Bill rigged the computer to make up for these things but still, cards and dice - it's very bad. Why replay? I'm doing the 1994 MLB replay now. C&D. #1 reason is that there were no playoffs that year and it would have been the first year with a Wild Card. #2 Montreal was supposed to be the best. Let's find out. #3 KG Jr and Matt Williams were on pace to break Maris's record of 61 HR's. Let's find out. #4 Gwynn had a chance to hit .400. He ended the year at .394. Let's find out if he can. Lot's of good reasons and answers to "What if's?"