Analyzing your team performances
Or why regression to the mean is key
If I gave you a batter with a pretty good career record, but who is having a bad stretch, how quickly would you get rid of it? This is actually a pretty big question. One I already touched a bit in my previous article about rational decision-making processes. Too often, we’re subject to many psychological biases, and overweighting recent events is one of them. Ever heard of the Hot Hand Fallacy?
Very often in Astonishing Baseball, players tend to act based on recent performances instead of relying on data. What most players ignore is that players (and people in general) always regress toward the mean. The principle is very simple. For the sake of the example, imagine a player who has the same batting average for all his career: 0.280. Given this single point of data, you should expect him to bat 0.280, more or less. Problem is that, if that same player starts to bat 0.150 for a few games, things go badly. This player would probably be ditched by a lot of AB players, inventing all kinds of reasons. Things like luck (or lack or luck) or randomness are totally ignored.
Except that, again, a player is expected to regress to the mean. In that case, that same 0.280 player, even if he bats 0.150 for some games, will probably, over the long run, bat 0.280 again! Of course, this is only true in the long run, but it’s still probably the best thing to do to expect him to improve. Of course, things like age or injuries have an impact, but then it’s quickly difficult to adjust your expectations without solid maths.
Then, one huge difficulty in real baseball is to find that actual mean for a player. Using the player career stats is a good way to do it, and you can actually then adjust with other factors like park factors or age. That’s what most projection systems do!
But AB players are lucky, because the regression to the mean is even more important in Astonishing Baseball than in reality. And that actual mean is simply career stats! That same 0.280 player from before is expecting to bat 0.280 over the long run, for real. That’s how the game is made. Again, there are other factors, but it’s still a pretty good expectation to have. Note that it’s also the same for players who are over-performing the expectations
So next time you have a player who is not doing great for a while, don’t ditch him right away. You might want to give him more time to regress.
A bunch of other advices for AB players:
- Age, training and prospect potential are the most important factors to project a player true level
- Always take a look at career stats, they’re an important part of the game!
- If you have a player in a bad stretch, bench him for a few games. Yes, he’s expected to regress to the mean, but having some extra-energy can definitely help!